[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[Guile-commits] GNU Guile branch, master, updated. 1ee2c72eafaae5f91f4c8
|
From: |
Ludovic Courtès |
|
Subject: |
[Guile-commits] GNU Guile branch, master, updated. 1ee2c72eafaae5f91f4c899bc4b4853af5c16f28 |
|
Date: |
Thu, 28 May 2009 21:19:29 +0000 |
This is an automated email from the git hooks/post-receive script. It was
generated because a ref change was pushed to the repository containing
the project "GNU Guile".
http://git.savannah.gnu.org/cgit/guile.git/commit/?id=1ee2c72eafaae5f91f4c899bc4b4853af5c16f28
The branch, master has been updated
via 1ee2c72eafaae5f91f4c899bc4b4853af5c16f28 (commit)
via 24d56127bb0f07bcb477e2c73ccc3cac0c51ee73 (commit)
from 21346c4f5e30910e3950c40bc267bb4249973240 (commit)
Those revisions listed above that are new to this repository have
not appeared on any other notification email; so we list those
revisions in full, below.
- Log -----------------------------------------------------------------
commit 1ee2c72eafaae5f91f4c899bc4b4853af5c16f28
Author: Ludovic Courtès <address@hidden>
Date: Wed May 27 18:18:07 2009 +0200
Import R6RS bytevectors and I/O ports from Guile-R6RS-Libs 0.2.
* README: Document dependency on GNU libunistring.
* benchmark-suite/Makefile.am (SCM_BENCHMARKS): Add
`benchmark/bytevectors.bm'.
* configure.in: Make sure we have libunistring; update $LIBS.
* libguile.h: Include "bytevectors.h" and "r6rs-ports.h".
* libguile/Makefile.am (libguile_la_SOURCES): Add `bytevectors.c' and
`r6rs-ports.c'
(DOT_X_FILES): Add `bytevectors.x' and `r6rs-ports.x'.
(DOT_DOC_FILES): Add `bytevectors.doc' and `r6rs-ports.doc'.
(noinst_HEADERS): Add `ieee-754.h'.
(modinclude_HEADERS): Add `bytevectors.h' and `r6rs-ports.h'
* libguile/validate.h (SCM_VALIDATE_BYTEVECTOR): New macro.
* module/Makefile.am (SOURCES): Add $(RNRS_SOURCES).
(RNRS_SOURCES): New variable.
* test-suite/Makefile.am (SCM_TESTS): Add `bytevectors.test' and
`r6rs-ports.test'.
commit 24d56127bb0f07bcb477e2c73ccc3cac0c51ee73
Author: Ludovic Courtès <address@hidden>
Date: Wed May 27 16:50:40 2009 +0200
Use GNU libunistring and Gnulib modules needed by R6RS bytevectors and
ports.
* m4/gnulib-cache.m4 (gl_MODULES): Add `byteswap', `iconv_open-utf',
`libunistring', `striconveh', and `string'.
-----------------------------------------------------------------------
Summary of changes:
README | 6 +
benchmark-suite/Makefile.am | 1 +
benchmark-suite/benchmarks/bytevectors.bm | 99 ++
build-aux/config.rpath | 24 +-
configure.in | 7 +
lib/Makefile.am | 239 ++++-
lib/byteswap.in.h | 44 +
lib/c-ctype.c | 396 ++++++
lib/c-ctype.h | 295 +++++
lib/c-strcase.h | 55 +
lib/c-strcasecmp.c | 57 +
lib/c-strcaseeq.h | 184 +++
lib/c-strncasecmp.c | 57 +
lib/iconv.c | 450 +++++++
lib/iconv.in.h | 71 +
lib/iconv_close.c | 47 +
lib/iconv_open-aix.gperf | 44 +
lib/iconv_open-hpux.gperf | 56 +
lib/iconv_open-irix.gperf | 31 +
lib/iconv_open-osf.gperf | 50 +
lib/iconv_open.c | 172 +++
lib/iconveh.h | 41 +
lib/striconveh.c | 1251 ++++++++++++++++++
lib/striconveh.h | 120 ++
lib/string.in.h | 605 +++++++++
lib/unistr.h | 681 ++++++++++
lib/unistr/u8-mbtouc-aux.c | 158 +++
lib/unistr/u8-mbtouc-unsafe-aux.c | 168 +++
lib/unistr/u8-mbtouc-unsafe.c | 179 +++
lib/unistr/u8-mbtouc.c | 168 +++
lib/unistr/u8-mbtoucr.c | 285 +++++
lib/unistr/u8-prev.c | 93 ++
lib/unistr/u8-uctomb-aux.c | 69 +
lib/unistr/u8-uctomb.c | 88 ++
lib/unitypes.h | 26 +
libguile.h | 4 +-
libguile/Makefile.am | 26 +-
libguile/bytevectors.c | 1978 +++++++++++++++++++++++++++++
libguile/bytevectors.h | 133 ++
libguile/ieee-754.h | 90 ++
libguile/r6rs-ports.c | 1118 ++++++++++++++++
libguile/r6rs-ports.h | 43 +
libguile/validate.h | 5 +-
m4/byteswap.m4 | 18 +
m4/gnulib-cache.m4 | 7 +-
m4/gnulib-comp.m4 | 55 +
m4/iconv.m4 | 180 +++
m4/iconv_h.m4 | 34 +
m4/iconv_open.m4 | 237 ++++
m4/lib-ld.m4 | 110 ++
m4/lib-link.m4 | 761 +++++++++++
m4/lib-prefix.m4 | 224 ++++
m4/libunistring.m4 | 37 +
m4/string_h.m4 | 92 ++
module/Makefile.am | 7 +-
module/rnrs/bytevector.scm | 84 ++
module/rnrs/io/ports.scm | 111 ++
test-suite/Makefile.am | 2 +
test-suite/tests/bytevectors.test | 531 ++++++++
test-suite/tests/r6rs-ports.test | 455 +++++++
60 files changed, 12635 insertions(+), 24 deletions(-)
create mode 100644 benchmark-suite/benchmarks/bytevectors.bm
create mode 100644 lib/byteswap.in.h
create mode 100644 lib/c-ctype.c
create mode 100644 lib/c-ctype.h
create mode 100644 lib/c-strcase.h
create mode 100644 lib/c-strcasecmp.c
create mode 100644 lib/c-strcaseeq.h
create mode 100644 lib/c-strncasecmp.c
create mode 100644 lib/iconv.c
create mode 100644 lib/iconv.in.h
create mode 100644 lib/iconv_close.c
create mode 100644 lib/iconv_open-aix.gperf
create mode 100644 lib/iconv_open-hpux.gperf
create mode 100644 lib/iconv_open-irix.gperf
create mode 100644 lib/iconv_open-osf.gperf
create mode 100644 lib/iconv_open.c
create mode 100644 lib/iconveh.h
create mode 100644 lib/striconveh.c
create mode 100644 lib/striconveh.h
create mode 100644 lib/string.in.h
create mode 100644 lib/unistr.h
create mode 100644 lib/unistr/u8-mbtouc-aux.c
create mode 100644 lib/unistr/u8-mbtouc-unsafe-aux.c
create mode 100644 lib/unistr/u8-mbtouc-unsafe.c
create mode 100644 lib/unistr/u8-mbtouc.c
create mode 100644 lib/unistr/u8-mbtoucr.c
create mode 100644 lib/unistr/u8-prev.c
create mode 100644 lib/unistr/u8-uctomb-aux.c
create mode 100644 lib/unistr/u8-uctomb.c
create mode 100644 lib/unitypes.h
create mode 100644 libguile/bytevectors.c
create mode 100644 libguile/bytevectors.h
create mode 100644 libguile/ieee-754.h
create mode 100644 libguile/r6rs-ports.c
create mode 100644 libguile/r6rs-ports.h
create mode 100644 m4/byteswap.m4
create mode 100644 m4/iconv.m4
create mode 100644 m4/iconv_h.m4
create mode 100644 m4/iconv_open.m4
create mode 100644 m4/lib-ld.m4
create mode 100644 m4/lib-link.m4
create mode 100644 m4/lib-prefix.m4
create mode 100644 m4/libunistring.m4
create mode 100644 m4/string_h.m4
create mode 100644 module/rnrs/bytevector.scm
create mode 100644 module/rnrs/io/ports.scm
create mode 100644 test-suite/tests/bytevectors.test
create mode 100644 test-suite/tests/r6rs-ports.test
diff --git a/README b/README
index 9993fcf..4950229 100644
--- a/README
+++ b/README
@@ -61,6 +61,12 @@ Guile requires the following external packages:
libltdl is used for loading extensions at run-time. It is
available from http://www.gnu.org/software/libtool/
+ - GNU libunistring
+
+ libunistring is used for Unicode string operations, such as the
+ `utf*->string' procedures. It is available from
+ http://www.gnu.org/software/libunistring/ .
+
Special Instructions For Some Systems =====================================
diff --git a/benchmark-suite/Makefile.am b/benchmark-suite/Makefile.am
index e65e8bc..dcadd58 100644
--- a/benchmark-suite/Makefile.am
+++ b/benchmark-suite/Makefile.am
@@ -1,4 +1,5 @@
SCM_BENCHMARKS = benchmarks/0-reference.bm \
+ benchmarks/bytevectors.bm \
benchmarks/continuations.bm \
benchmarks/if.bm \
benchmarks/logand.bm \
diff --git a/benchmark-suite/benchmarks/bytevectors.bm
b/benchmark-suite/benchmarks/bytevectors.bm
new file mode 100644
index 0000000..9547a71
--- /dev/null
+++ b/benchmark-suite/benchmarks/bytevectors.bm
@@ -0,0 +1,99 @@
+;;; -*- mode: scheme; coding: latin-1; -*-
+;;; R6RS Byte Vectors.
+;;;
+;;; Copyright 2009 Ludovic Courtès <address@hidden>
+;;;
+;;;
+;;; This program is free software; you can redistribute it and/or modify
+;;; it under the terms of the GNU General Public License as published by
+;;; the Free Software Foundation; either version 2 of the License, or
+;;; (at your option) any later version.
+;;;
+;;; This program is distributed in the hope that it will be useful,
+;;; but WITHOUT ANY WARRANTY; without even the implied warranty of
+;;; MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+;;; GNU General Public License for more details.
+;;;
+;;; You should have received a copy of the GNU General Public License
+;;; along with this program; if not, write to the Free Software
+;;; Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
+
+(define-module (benchmarks bytevector)
+ :use-module (rnrs bytevector)
+ :use-module (srfi srfi-4)
+ :use-module (benchmark-suite lib))
+
+(define bv (make-bytevector 16384))
+
+(define %native-endianness
+ (native-endianness))
+
+(define %foreign-endianness
+ (if (eq? (native-endianness) (endianness little))
+ (endianness big)
+ (endianness little)))
+
+(define u8v (make-u8vector 16384))
+(define u16v (make-u16vector 8192))
+(define u32v (make-u32vector 4196))
+(define u64v (make-u64vector 2048))
+
+
+(with-benchmark-prefix "ref/set!"
+
+ (benchmark "bytevector-u8-ref" 1000000
+ (bytevector-u8-ref bv 0))
+
+ (benchmark "bytevector-u16-ref (foreign)" 1000000
+ (bytevector-u16-ref bv 0 %foreign-endianness))
+
+ (benchmark "bytevector-u16-ref (native)" 1000000
+ (bytevector-u16-ref bv 0 %native-endianness))
+
+ (benchmark "bytevector-u16-native-ref" 1000000
+ (bytevector-u16-native-ref bv 0))
+
+ (benchmark "bytevector-u32-ref (foreign)" 1000000
+ (bytevector-u32-ref bv 0 %foreign-endianness))
+
+ (benchmark "bytevector-u32-ref (native)" 1000000
+ (bytevector-u32-ref bv 0 %native-endianness))
+
+ (benchmark "bytevector-u32-native-ref" 1000000
+ (bytevector-u32-native-ref bv 0))
+
+ (benchmark "bytevector-u64-ref (foreign)" 1000000
+ (bytevector-u64-ref bv 0 %foreign-endianness))
+
+ (benchmark "bytevector-u64-ref (native)" 1000000
+ (bytevector-u64-ref bv 0 %native-endianness))
+
+ (benchmark "bytevector-u64-native-ref" 1000000
+ (bytevector-u16-native-ref bv 0)))
+
+
+(with-benchmark-prefix "lists"
+
+ (benchmark "bytevector->u8-list" 2000
+ (bytevector->u8-list bv))
+
+ (benchmark "bytevector->uint-list 16-bit" 2000
+ (bytevector->uint-list bv (native-endianness) 2))
+
+ (benchmark "bytevector->uint-list 64-bit" 2000
+ (bytevector->uint-list bv (native-endianness) 8)))
+
+
+(with-benchmark-prefix "SRFI-4" ;; for comparison
+
+ (benchmark "u8vector-ref" 1000000
+ (u8vector-ref u8v 0))
+
+ (benchmark "u16vector-ref" 1000000
+ (u16vector-ref u16v 0))
+
+ (benchmark "u32vector-ref" 1000000
+ (u32vector-ref u32v 0))
+
+ (benchmark "u64vector-ref" 1000000
+ (u64vector-ref u64v 0)))
diff --git a/build-aux/config.rpath b/build-aux/config.rpath
index 35f959b..85c2f20 100755
--- a/build-aux/config.rpath
+++ b/build-aux/config.rpath
@@ -47,7 +47,7 @@ for cc_temp in $CC""; do
done
cc_basename=`echo "$cc_temp" | sed -e 's%^.*/%%'`
-# Code taken from libtool.m4's AC_LIBTOOL_PROG_COMPILER_PIC.
+# Code taken from libtool.m4's _LT_COMPILER_PIC.
wl=
if test "$GCC" = yes; then
@@ -64,7 +64,7 @@ else
;;
esac
;;
- mingw* | cygwin* | pw32* | os2*)
+ mingw* | cygwin* | pw32* | os2* | cegcc*)
;;
hpux9* | hpux10* | hpux11*)
wl='-Wl,'
@@ -76,7 +76,13 @@ else
;;
linux* | k*bsd*-gnu)
case $cc_basename in
- icc* | ecc*)
+ ecc*)
+ wl='-Wl,'
+ ;;
+ icc* | ifort*)
+ wl='-Wl,'
+ ;;
+ lf95*)
wl='-Wl,'
;;
pgcc | pgf77 | pgf90)
@@ -124,7 +130,7 @@ else
esac
fi
-# Code taken from libtool.m4's AC_LIBTOOL_PROG_LD_SHLIBS.
+# Code taken from libtool.m4's _LT_LINKER_SHLIBS.
hardcode_libdir_flag_spec=
hardcode_libdir_separator=
@@ -132,7 +138,7 @@ hardcode_direct=no
hardcode_minus_L=no
case "$host_os" in
- cygwin* | mingw* | pw32*)
+ cygwin* | mingw* | pw32* | cegcc*)
# FIXME: the MSVC++ port hasn't been tested in a loooong time
# When not using gcc, we currently assume that we are using
# Microsoft Visual C++.
@@ -182,7 +188,7 @@ if test "$with_gnu_ld" = yes; then
ld_shlibs=no
fi
;;
- cygwin* | mingw* | pw32*)
+ cygwin* | mingw* | pw32* | cegcc*)
# hardcode_libdir_flag_spec is actually meaningless, as there is
# no search path for DLLs.
hardcode_libdir_flag_spec='-L$libdir'
@@ -326,7 +332,7 @@ else
;;
bsdi[45]*)
;;
- cygwin* | mingw* | pw32*)
+ cygwin* | mingw* | pw32* | cegcc*)
# When not using gcc, we currently assume that we are using
# Microsoft Visual C++.
# hardcode_libdir_flag_spec is actually meaningless, as there is
@@ -494,7 +500,7 @@ else
fi
# Check dynamic linker characteristics
-# Code taken from libtool.m4's AC_LIBTOOL_SYS_DYNAMIC_LINKER.
+# Code taken from libtool.m4's _LT_SYS_DYNAMIC_LINKER.
# Unlike libtool.m4, here we don't care about _all_ names of the library, but
# only about the one the linker finds when passed -lNAME. This is the last
# element of library_names_spec in libtool.m4, or possibly two of them if the
@@ -517,7 +523,7 @@ case "$host_os" in
bsdi[45]*)
library_names_spec='$libname$shrext'
;;
- cygwin* | mingw* | pw32*)
+ cygwin* | mingw* | pw32* | cegcc*)
shrext=.dll
library_names_spec='$libname.dll.a $libname.lib'
;;
diff --git a/configure.in b/configure.in
index 07c4766..6568e52 100644
--- a/configure.in
+++ b/configure.in
@@ -836,6 +836,13 @@ AC_LINK_IFELSE([AC_LANG_PROGRAM([[#include <gmp.h>]],
[],
[AC_MSG_ERROR([At least GNU MP 4.1 is required, see README])])
+dnl GNU libunistring tests.
+if test "x$LTLIBUNISTRING" != "x"; then
+ LIBS="$LTLIBUNISTRING $LIBS"
+else
+ AC_MSG_ERROR([GNU libunistring is required, please install it.])
+fi
+
dnl i18n tests
#AC_CHECK_HEADERS([libintl.h])
#AC_CHECK_FUNCS(gettext)
diff --git a/lib/Makefile.am b/lib/Makefile.am
index f321b0b..6f2f5c5 100644
--- a/lib/Makefile.am
+++ b/lib/Makefile.am
@@ -9,9 +9,9 @@
# the same distribution terms as the rest of that program.
#
# Generated by gnulib-tool.
-# Reproduce by: gnulib-tool --import --dir=. --lib=libgnu --source-base=lib
--m4-base=m4 --doc-base=doc --tests-base=tests --aux-dir=build-aux --lgpl
--libtool --macro-prefix=gl --no-vc-files alloca-opt autobuild count-one-bits
environ extensions flock fpieee full-read full-write lib-symbol-visibility
putenv stdlib strcase strftime
+# Reproduce by: gnulib-tool --import --dir=. --lib=libgnu --source-base=lib
--m4-base=m4 --doc-base=doc --tests-base=tests --aux-dir=build-aux --lgpl
--libtool --macro-prefix=gl --no-vc-files alloca-opt autobuild byteswap
count-one-bits environ extensions flock fpieee full-read full-write
iconv_open-utf lib-symbol-visibility libunistring putenv stdlib strcase
strftime striconveh string
-AUTOMAKE_OPTIONS = 1.5 gnits
+AUTOMAKE_OPTIONS = 1.5 gnits subdir-objects
SUBDIRS =
noinst_HEADERS =
@@ -54,6 +54,42 @@ EXTRA_DIST += alloca.in.h
## end gnulib module alloca-opt
+## begin gnulib module byteswap
+
+BUILT_SOURCES += $(BYTESWAP_H)
+
+# We need the following in order to create <byteswap.h> when the system
+# doesn't have one.
+byteswap.h: byteswap.in.h
+ { echo '/* DO NOT EDIT! GENERATED AUTOMATICALLY! */'; \
+ cat $(srcdir)/byteswap.in.h; \
+ } > address@hidden
+ mv -f address@hidden $@
+MOSTLYCLEANFILES += byteswap.h byteswap.h-t
+
+EXTRA_DIST += byteswap.in.h
+
+## end gnulib module byteswap
+
+## begin gnulib module c-ctype
+
+libgnu_la_SOURCES += c-ctype.h c-ctype.c
+
+## end gnulib module c-ctype
+
+## begin gnulib module c-strcase
+
+libgnu_la_SOURCES += c-strcase.h c-strcasecmp.c c-strncasecmp.c
+
+## end gnulib module c-strcase
+
+## begin gnulib module c-strcaseeq
+
+
+EXTRA_DIST += c-strcaseeq.h
+
+## end gnulib module c-strcaseeq
+
## begin gnulib module configmake
# Retrieve values of the variables through 'configure' followed by
@@ -143,6 +179,72 @@ libgnu_la_SOURCES += full-write.h full-write.c
## end gnulib module full-write
+## begin gnulib module gperf
+
+GPERF = gperf
+
+## end gnulib module gperf
+
+## begin gnulib module havelib
+
+
+EXTRA_DIST += $(top_srcdir)/build-aux/config.rpath
+
+## end gnulib module havelib
+
+## begin gnulib module iconv_open
+
+BUILT_SOURCES += $(ICONV_H)
+
+# We need the following in order to create <iconv.h> when the system
+# doesn't have one that works with the given compiler.
+iconv.h: iconv.in.h
+ rm -f address@hidden $@
+ { echo '/* DO NOT EDIT! GENERATED AUTOMATICALLY! */' && \
+ sed -e 's|@''INCLUDE_NEXT''@|$(INCLUDE_NEXT)|g' \
+ -e 's|@''PRAGMA_SYSTEM_HEADER''@|@PRAGMA_SYSTEM_HEADER@|g' \
+ -e 's|@''NEXT_ICONV_H''@|$(NEXT_ICONV_H)|g' \
+ -e 's|@''ICONV_CONST''@|$(ICONV_CONST)|g' \
+ -e 's|@''REPLACE_ICONV''@|$(REPLACE_ICONV)|g' \
+ -e 's|@''REPLACE_ICONV_OPEN''@|$(REPLACE_ICONV_OPEN)|g' \
+ -e 's|@''REPLACE_ICONV_UTF''@|$(REPLACE_ICONV_UTF)|g' \
+ < $(srcdir)/iconv.in.h; \
+ } > address@hidden
+ mv address@hidden $@
+MOSTLYCLEANFILES += iconv.h iconv.h-t
+
+iconv_open-aix.h: iconv_open-aix.gperf
+ $(GPERF) -m 10 $(srcdir)/iconv_open-aix.gperf >
$(srcdir)/iconv_open-aix.h-t
+ mv $(srcdir)/iconv_open-aix.h-t $(srcdir)/iconv_open-aix.h
+iconv_open-hpux.h: iconv_open-hpux.gperf
+ $(GPERF) -m 10 $(srcdir)/iconv_open-hpux.gperf >
$(srcdir)/iconv_open-hpux.h-t
+ mv $(srcdir)/iconv_open-hpux.h-t $(srcdir)/iconv_open-hpux.h
+iconv_open-irix.h: iconv_open-irix.gperf
+ $(GPERF) -m 10 $(srcdir)/iconv_open-irix.gperf >
$(srcdir)/iconv_open-irix.h-t
+ mv $(srcdir)/iconv_open-irix.h-t $(srcdir)/iconv_open-irix.h
+iconv_open-osf.h: iconv_open-osf.gperf
+ $(GPERF) -m 10 $(srcdir)/iconv_open-osf.gperf >
$(srcdir)/iconv_open-osf.h-t
+ mv $(srcdir)/iconv_open-osf.h-t $(srcdir)/iconv_open-osf.h
+BUILT_SOURCES += iconv_open-aix.h iconv_open-hpux.h iconv_open-irix.h
iconv_open-osf.h
+MOSTLYCLEANFILES += iconv_open-aix.h-t iconv_open-hpux.h-t
iconv_open-irix.h-t iconv_open-osf.h-t
+MAINTAINERCLEANFILES += iconv_open-aix.h iconv_open-hpux.h iconv_open-irix.h
iconv_open-osf.h
+EXTRA_DIST += iconv_open-aix.h iconv_open-hpux.h iconv_open-irix.h
iconv_open-osf.h
+
+EXTRA_DIST += iconv.in.h iconv_open-aix.gperf iconv_open-hpux.gperf
iconv_open-irix.gperf iconv_open-osf.gperf iconv_open.c
+
+EXTRA_libgnu_la_SOURCES += iconv_open.c
+
+## end gnulib module iconv_open
+
+## begin gnulib module iconv_open-utf
+
+
+EXTRA_DIST += iconv.c iconv_close.c
+
+EXTRA_libgnu_la_SOURCES += iconv.c iconv_close.c
+
+## end gnulib module iconv_open-utf
+
## begin gnulib module lib-symbol-visibility
# The value of $(CFLAG_VISIBILITY) needs to be added to the CFLAGS for the
@@ -442,6 +544,95 @@ EXTRA_libgnu_la_SOURCES += strftime.c
## end gnulib module strftime
+## begin gnulib module striconveh
+
+libgnu_la_SOURCES += striconveh.h striconveh.c
+if GL_COND_LIBTOOL
+libgnu_la_LDFLAGS += $(LTLIBICONV)
+endif
+
+EXTRA_DIST += iconveh.h
+
+## end gnulib module striconveh
+
+## begin gnulib module string
+
+BUILT_SOURCES += string.h
+
+# We need the following in order to create <string.h> when the system
+# doesn't have one that works with the given compiler.
+string.h: string.in.h
+ rm -f address@hidden $@
+ { echo '/* DO NOT EDIT! GENERATED AUTOMATICALLY! */' && \
+ sed -e 's|@''INCLUDE_NEXT''@|$(INCLUDE_NEXT)|g' \
+ -e 's|@''PRAGMA_SYSTEM_HEADER''@|@PRAGMA_SYSTEM_HEADER@|g' \
+ -e 's|@''NEXT_STRING_H''@|$(NEXT_STRING_H)|g' \
+ -e 's|@''GNULIB_MBSLEN''@|$(GNULIB_MBSLEN)|g' \
+ -e 's|@''GNULIB_MBSNLEN''@|$(GNULIB_MBSNLEN)|g' \
+ -e 's|@''GNULIB_MBSCHR''@|$(GNULIB_MBSCHR)|g' \
+ -e 's|@''GNULIB_MBSRCHR''@|$(GNULIB_MBSRCHR)|g' \
+ -e 's|@''GNULIB_MBSSTR''@|$(GNULIB_MBSSTR)|g' \
+ -e 's|@''GNULIB_MBSCASECMP''@|$(GNULIB_MBSCASECMP)|g' \
+ -e 's|@''GNULIB_MBSNCASECMP''@|$(GNULIB_MBSNCASECMP)|g' \
+ -e 's|@''GNULIB_MBSPCASECMP''@|$(GNULIB_MBSPCASECMP)|g' \
+ -e 's|@''GNULIB_MBSCASESTR''@|$(GNULIB_MBSCASESTR)|g' \
+ -e 's|@''GNULIB_MBSCSPN''@|$(GNULIB_MBSCSPN)|g' \
+ -e 's|@''GNULIB_MBSPBRK''@|$(GNULIB_MBSPBRK)|g' \
+ -e 's|@''GNULIB_MBSSPN''@|$(GNULIB_MBSSPN)|g' \
+ -e 's|@''GNULIB_MBSSEP''@|$(GNULIB_MBSSEP)|g' \
+ -e 's|@''GNULIB_MBSTOK_R''@|$(GNULIB_MBSTOK_R)|g' \
+ -e 's|@''GNULIB_MEMMEM''@|$(GNULIB_MEMMEM)|g' \
+ -e 's|@''GNULIB_MEMPCPY''@|$(GNULIB_MEMPCPY)|g' \
+ -e 's|@''GNULIB_MEMRCHR''@|$(GNULIB_MEMRCHR)|g' \
+ -e 's|@''GNULIB_RAWMEMCHR''@|$(GNULIB_RAWMEMCHR)|g' \
+ -e 's|@''GNULIB_STPCPY''@|$(GNULIB_STPCPY)|g' \
+ -e 's|@''GNULIB_STPNCPY''@|$(GNULIB_STPNCPY)|g' \
+ -e 's|@''GNULIB_STRCHRNUL''@|$(GNULIB_STRCHRNUL)|g' \
+ -e 's|@''GNULIB_STRDUP''@|$(GNULIB_STRDUP)|g' \
+ -e 's|@''GNULIB_STRNDUP''@|$(GNULIB_STRNDUP)|g' \
+ -e 's|@''GNULIB_STRNLEN''@|$(GNULIB_STRNLEN)|g' \
+ -e 's|@''GNULIB_STRPBRK''@|$(GNULIB_STRPBRK)|g' \
+ -e 's|@''GNULIB_STRSEP''@|$(GNULIB_STRSEP)|g' \
+ -e 's|@''GNULIB_STRSTR''@|$(GNULIB_STRSTR)|g' \
+ -e 's|@''GNULIB_STRCASESTR''@|$(GNULIB_STRCASESTR)|g' \
+ -e 's|@''GNULIB_STRTOK_R''@|$(GNULIB_STRTOK_R)|g' \
+ -e 's|@''GNULIB_STRERROR''@|$(GNULIB_STRERROR)|g' \
+ -e 's|@''GNULIB_STRSIGNAL''@|$(GNULIB_STRSIGNAL)|g' \
+ -e 's|@''GNULIB_STRVERSCMP''@|$(GNULIB_STRVERSCMP)|g' \
+ -e 's|@''HAVE_DECL_MEMMEM''@|$(HAVE_DECL_MEMMEM)|g' \
+ -e 's|@''HAVE_MEMPCPY''@|$(HAVE_MEMPCPY)|g' \
+ -e 's|@''HAVE_DECL_MEMRCHR''@|$(HAVE_DECL_MEMRCHR)|g' \
+ -e 's|@''HAVE_RAWMEMCHR''@|$(HAVE_RAWMEMCHR)|g' \
+ -e 's|@''HAVE_STPCPY''@|$(HAVE_STPCPY)|g' \
+ -e 's|@''HAVE_STPNCPY''@|$(HAVE_STPNCPY)|g' \
+ -e 's|@''HAVE_STRCHRNUL''@|$(HAVE_STRCHRNUL)|g' \
+ -e 's|@''HAVE_DECL_STRDUP''@|$(HAVE_DECL_STRDUP)|g' \
+ -e 's|@''HAVE_STRNDUP''@|$(HAVE_STRNDUP)|g' \
+ -e 's|@''HAVE_DECL_STRNDUP''@|$(HAVE_DECL_STRNDUP)|g' \
+ -e 's|@''HAVE_DECL_STRNLEN''@|$(HAVE_DECL_STRNLEN)|g' \
+ -e 's|@''HAVE_STRPBRK''@|$(HAVE_STRPBRK)|g' \
+ -e 's|@''HAVE_STRSEP''@|$(HAVE_STRSEP)|g' \
+ -e 's|@''HAVE_STRCASESTR''@|$(HAVE_STRCASESTR)|g' \
+ -e 's|@''HAVE_DECL_STRTOK_R''@|$(HAVE_DECL_STRTOK_R)|g' \
+ -e 's|@''HAVE_DECL_STRERROR''@|$(HAVE_DECL_STRERROR)|g' \
+ -e 's|@''HAVE_DECL_STRSIGNAL''@|$(HAVE_DECL_STRSIGNAL)|g' \
+ -e 's|@''HAVE_STRVERSCMP''@|$(HAVE_STRVERSCMP)|g' \
+ -e 's|@''REPLACE_MEMMEM''@|$(REPLACE_MEMMEM)|g' \
+ -e 's|@''REPLACE_STRCASESTR''@|$(REPLACE_STRCASESTR)|g' \
+ -e 's|@''REPLACE_STRDUP''@|$(REPLACE_STRDUP)|g' \
+ -e 's|@''REPLACE_STRSTR''@|$(REPLACE_STRSTR)|g' \
+ -e 's|@''REPLACE_STRERROR''@|$(REPLACE_STRERROR)|g' \
+ -e 's|@''REPLACE_STRSIGNAL''@|$(REPLACE_STRSIGNAL)|g' \
+ -e '/definition of GL_LINK_WARNING/r $(LINK_WARNING_H)' \
+ < $(srcdir)/string.in.h; \
+ } > address@hidden
+ mv address@hidden $@
+MOSTLYCLEANFILES += string.h string.h-t
+
+EXTRA_DIST += string.in.h
+
+## end gnulib module string
+
## begin gnulib module strings
BUILT_SOURCES += strings.h
@@ -598,6 +789,50 @@ EXTRA_DIST += unistd.in.h
## end gnulib module unistd
+## begin gnulib module unistr/base
+
+
+EXTRA_DIST += unistr.h
+
+## end gnulib module unistr/base
+
+## begin gnulib module unistr/u8-mbtouc
+
+libgnu_la_SOURCES += unistr/u8-mbtouc.c unistr/u8-mbtouc-aux.c
+
+## end gnulib module unistr/u8-mbtouc
+
+## begin gnulib module unistr/u8-mbtouc-unsafe
+
+libgnu_la_SOURCES += unistr/u8-mbtouc-unsafe.c unistr/u8-mbtouc-unsafe-aux.c
+
+## end gnulib module unistr/u8-mbtouc-unsafe
+
+## begin gnulib module unistr/u8-mbtoucr
+
+libgnu_la_SOURCES += unistr/u8-mbtoucr.c
+
+## end gnulib module unistr/u8-mbtoucr
+
+## begin gnulib module unistr/u8-prev
+
+libgnu_la_SOURCES += unistr/u8-prev.c
+
+## end gnulib module unistr/u8-prev
+
+## begin gnulib module unistr/u8-uctomb
+
+libgnu_la_SOURCES += unistr/u8-uctomb.c unistr/u8-uctomb-aux.c
+
+## end gnulib module unistr/u8-uctomb
+
+## begin gnulib module unitypes
+
+
+EXTRA_DIST += unitypes.h
+
+## end gnulib module unitypes
+
## begin gnulib module verify
libgnu_la_SOURCES += verify.h
diff --git a/lib/byteswap.in.h b/lib/byteswap.in.h
new file mode 100644
index 0000000..f03463d
--- /dev/null
+++ b/lib/byteswap.in.h
@@ -0,0 +1,44 @@
+/* byteswap.h - Byte swapping
+ Copyright (C) 2005, 2007 Free Software Foundation, Inc.
+ Written by Oskar Liljeblad <address@hidden>, 2005.
+
+ This program is free software: you can redistribute it and/or modify
+ it under the terms of the GNU Lesser General Public License as published by
+ the Free Software Foundation; either version 3 of the License, or
+ (at your option) any later version.
+
+ This program is distributed in the hope that it will be useful,
+ but WITHOUT ANY WARRANTY; without even the implied warranty of
+ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ GNU Lesser General Public License for more details.
+
+ You should have received a copy of the GNU Lesser General Public License
+ along with this program. If not, see <http://www.gnu.org/licenses/>. */
+
+#ifndef _GL_BYTESWAP_H
+#define _GL_BYTESWAP_H
+
+/* Given an unsigned 16-bit argument X, return the value corresponding to
+ X with reversed byte order. */
+#define bswap_16(x) ((((x) & 0x00FF) << 8) | \
+ (((x) & 0xFF00) >> 8))
+
+/* Given an unsigned 32-bit argument X, return the value corresponding to
+ X with reversed byte order. */
+#define bswap_32(x) ((((x) & 0x000000FF) << 24) | \
+ (((x) & 0x0000FF00) << 8) | \
+ (((x) & 0x00FF0000) >> 8) | \
+ (((x) & 0xFF000000) >> 24))
+
+/* Given an unsigned 64-bit argument X, return the value corresponding to
+ X with reversed byte order. */
+#define bswap_64(x) ((((x) & 0x00000000000000FFULL) << 56) | \
+ (((x) & 0x000000000000FF00ULL) << 40) | \
+ (((x) & 0x0000000000FF0000ULL) << 24) | \
+ (((x) & 0x00000000FF000000ULL) << 8) | \
+ (((x) & 0x000000FF00000000ULL) >> 8) | \
+ (((x) & 0x0000FF0000000000ULL) >> 24) | \
+ (((x) & 0x00FF000000000000ULL) >> 40) | \
+ (((x) & 0xFF00000000000000ULL) >> 56))
+
+#endif /* _GL_BYTESWAP_H */
diff --git a/lib/c-ctype.c b/lib/c-ctype.c
new file mode 100644
index 0000000..e36a513
--- /dev/null
+++ b/lib/c-ctype.c
@@ -0,0 +1,396 @@
+/* Character handling in C locale.
+
+ Copyright 2000-2003, 2006 Free Software Foundation, Inc.
+
+This program is free software; you can redistribute it and/or modify
+it under the terms of the GNU Lesser General Public License as published by
+the Free Software Foundation; either version 2 of the License, or
+(at your option) any later version.
+
+This program is distributed in the hope that it will be useful,
+but WITHOUT ANY WARRANTY; without even the implied warranty of
+MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+GNU Lesser General Public License for more details.
+
+You should have received a copy of the GNU Lesser General Public License
+along with this program; if not, write to the Free Software Foundation,
+Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */
+
+#include <config.h>
+
+/* Specification. */
+#define NO_C_CTYPE_MACROS
+#include "c-ctype.h"
+
+/* The function isascii is not locale dependent. Its use in EBCDIC is
+ questionable. */
+bool
+c_isascii (int c)
+{
+ return (c >= 0x00 && c <= 0x7f);
+}
+
+bool
+c_isalnum (int c)
+{
+#if C_CTYPE_CONSECUTIVE_DIGITS \
+ && C_CTYPE_CONSECUTIVE_UPPERCASE && C_CTYPE_CONSECUTIVE_LOWERCASE
+#if C_CTYPE_ASCII
+ return ((c >= '0' && c <= '9')
+ || ((c & ~0x20) >= 'A' && (c & ~0x20) <= 'Z'));
+#else
+ return ((c >= '0' && c <= '9')
+ || (c >= 'A' && c <= 'Z')
+ || (c >= 'a' && c <= 'z'));
+#endif
+#else
+ switch (c)
+ {
+ case '0': case '1': case '2': case '3': case '4': case '5':
+ case '6': case '7': case '8': case '9':
+ case 'A': case 'B': case 'C': case 'D': case 'E': case 'F':
+ case 'G': case 'H': case 'I': case 'J': case 'K': case 'L':
+ case 'M': case 'N': case 'O': case 'P': case 'Q': case 'R':
+ case 'S': case 'T': case 'U': case 'V': case 'W': case 'X':
+ case 'Y': case 'Z':
+ case 'a': case 'b': case 'c': case 'd': case 'e': case 'f':
+ case 'g': case 'h': case 'i': case 'j': case 'k': case 'l':
+ case 'm': case 'n': case 'o': case 'p': case 'q': case 'r':
+ case 's': case 't': case 'u': case 'v': case 'w': case 'x':
+ case 'y': case 'z':
+ return 1;
+ default:
+ return 0;
+ }
+#endif
+}
+
+bool
+c_isalpha (int c)
+{
+#if C_CTYPE_CONSECUTIVE_UPPERCASE && C_CTYPE_CONSECUTIVE_LOWERCASE
+#if C_CTYPE_ASCII
+ return ((c & ~0x20) >= 'A' && (c & ~0x20) <= 'Z');
+#else
+ return ((c >= 'A' && c <= 'Z') || (c >= 'a' && c <= 'z'));
+#endif
+#else
+ switch (c)
+ {
+ case 'A': case 'B': case 'C': case 'D': case 'E': case 'F':
+ case 'G': case 'H': case 'I': case 'J': case 'K': case 'L':
+ case 'M': case 'N': case 'O': case 'P': case 'Q': case 'R':
+ case 'S': case 'T': case 'U': case 'V': case 'W': case 'X':
+ case 'Y': case 'Z':
+ case 'a': case 'b': case 'c': case 'd': case 'e': case 'f':
+ case 'g': case 'h': case 'i': case 'j': case 'k': case 'l':
+ case 'm': case 'n': case 'o': case 'p': case 'q': case 'r':
+ case 's': case 't': case 'u': case 'v': case 'w': case 'x':
+ case 'y': case 'z':
+ return 1;
+ default:
+ return 0;
+ }
+#endif
+}
+
+bool
+c_isblank (int c)
+{
+ return (c == ' ' || c == '\t');
+}
+
+bool
+c_iscntrl (int c)
+{
+#if C_CTYPE_ASCII
+ return ((c & ~0x1f) == 0 || c == 0x7f);
+#else
+ switch (c)
+ {
+ case ' ': case '!': case '"': case '#': case '$': case '%':
+ case '&': case '\'': case '(': case ')': case '*': case '+':
+ case ',': case '-': case '.': case '/':
+ case '0': case '1': case '2': case '3': case '4': case '5':
+ case '6': case '7': case '8': case '9':
+ case ':': case ';': case '<': case '=': case '>': case '?':
+ case '@':
+ case 'A': case 'B': case 'C': case 'D': case 'E': case 'F':
+ case 'G': case 'H': case 'I': case 'J': case 'K': case 'L':
+ case 'M': case 'N': case 'O': case 'P': case 'Q': case 'R':
+ case 'S': case 'T': case 'U': case 'V': case 'W': case 'X':
+ case 'Y': case 'Z':
+ case '[': case '\\': case ']': case '^': case '_': case '`':
+ case 'a': case 'b': case 'c': case 'd': case 'e': case 'f':
+ case 'g': case 'h': case 'i': case 'j': case 'k': case 'l':
+ case 'm': case 'n': case 'o': case 'p': case 'q': case 'r':
+ case 's': case 't': case 'u': case 'v': case 'w': case 'x':
+ case 'y': case 'z':
+ case '{': case '|': case '}': case '~':
+ return 0;
+ default:
+ return 1;
+ }
+#endif
+}
+
+bool
+c_isdigit (int c)
+{
+#if C_CTYPE_CONSECUTIVE_DIGITS
+ return (c >= '0' && c <= '9');
+#else
+ switch (c)
+ {
+ case '0': case '1': case '2': case '3': case '4': case '5':
+ case '6': case '7': case '8': case '9':
+ return 1;
+ default:
+ return 0;
+ }
+#endif
+}
+
+bool
+c_islower (int c)
+{
+#if C_CTYPE_CONSECUTIVE_LOWERCASE
+ return (c >= 'a' && c <= 'z');
+#else
+ switch (c)
+ {
+ case 'a': case 'b': case 'c': case 'd': case 'e': case 'f':
+ case 'g': case 'h': case 'i': case 'j': case 'k': case 'l':
+ case 'm': case 'n': case 'o': case 'p': case 'q': case 'r':
+ case 's': case 't': case 'u': case 'v': case 'w': case 'x':
+ case 'y': case 'z':
+ return 1;
+ default:
+ return 0;
+ }
+#endif
+}
+
+bool
+c_isgraph (int c)
+{
+#if C_CTYPE_ASCII
+ return (c >= '!' && c <= '~');
+#else
+ switch (c)
+ {
+ case '!': case '"': case '#': case '$': case '%': case '&':
+ case '\'': case '(': case ')': case '*': case '+': case ',':
+ case '-': case '.': case '/':
+ case '0': case '1': case '2': case '3': case '4': case '5':
+ case '6': case '7': case '8': case '9':
+ case ':': case ';': case '<': case '=': case '>': case '?':
+ case '@':
+ case 'A': case 'B': case 'C': case 'D': case 'E': case 'F':
+ case 'G': case 'H': case 'I': case 'J': case 'K': case 'L':
+ case 'M': case 'N': case 'O': case 'P': case 'Q': case 'R':
+ case 'S': case 'T': case 'U': case 'V': case 'W': case 'X':
+ case 'Y': case 'Z':
+ case '[': case '\\': case ']': case '^': case '_': case '`':
+ case 'a': case 'b': case 'c': case 'd': case 'e': case 'f':
+ case 'g': case 'h': case 'i': case 'j': case 'k': case 'l':
+ case 'm': case 'n': case 'o': case 'p': case 'q': case 'r':
+ case 's': case 't': case 'u': case 'v': case 'w': case 'x':
+ case 'y': case 'z':
+ case '{': case '|': case '}': case '~':
+ return 1;
+ default:
+ return 0;
+ }
+#endif
+}
+
+bool
+c_isprint (int c)
+{
+#if C_CTYPE_ASCII
+ return (c >= ' ' && c <= '~');
+#else
+ switch (c)
+ {
+ case ' ': case '!': case '"': case '#': case '$': case '%':
+ case '&': case '\'': case '(': case ')': case '*': case '+':
+ case ',': case '-': case '.': case '/':
+ case '0': case '1': case '2': case '3': case '4': case '5':
+ case '6': case '7': case '8': case '9':
+ case ':': case ';': case '<': case '=': case '>': case '?':
+ case '@':
+ case 'A': case 'B': case 'C': case 'D': case 'E': case 'F':
+ case 'G': case 'H': case 'I': case 'J': case 'K': case 'L':
+ case 'M': case 'N': case 'O': case 'P': case 'Q': case 'R':
+ case 'S': case 'T': case 'U': case 'V': case 'W': case 'X':
+ case 'Y': case 'Z':
+ case '[': case '\\': case ']': case '^': case '_': case '`':
+ case 'a': case 'b': case 'c': case 'd': case 'e': case 'f':
+ case 'g': case 'h': case 'i': case 'j': case 'k': case 'l':
+ case 'm': case 'n': case 'o': case 'p': case 'q': case 'r':
+ case 's': case 't': case 'u': case 'v': case 'w': case 'x':
+ case 'y': case 'z':
+ case '{': case '|': case '}': case '~':
+ return 1;
+ default:
+ return 0;
+ }
+#endif
+}
+
+bool
+c_ispunct (int c)
+{
+#if C_CTYPE_ASCII
+ return ((c >= '!' && c <= '~')
+ && !((c >= '0' && c <= '9')
+ || ((c & ~0x20) >= 'A' && (c & ~0x20) <= 'Z')));
+#else
+ switch (c)
+ {
+ case '!': case '"': case '#': case '$': case '%': case '&':
+ case '\'': case '(': case ')': case '*': case '+': case ',':
+ case '-': case '.': case '/':
+ case ':': case ';': case '<': case '=': case '>': case '?':
+ case '@':
+ case '[': case '\\': case ']': case '^': case '_': case '`':
+ case '{': case '|': case '}': case '~':
+ return 1;
+ default:
+ return 0;
+ }
+#endif
+}
+
+bool
+c_isspace (int c)
+{
+ return (c == ' ' || c == '\t'
+ || c == '\n' || c == '\v' || c == '\f' || c == '\r');
+}
+
+bool
+c_isupper (int c)
+{
+#if C_CTYPE_CONSECUTIVE_UPPERCASE
+ return (c >= 'A' && c <= 'Z');
+#else
+ switch (c)
+ {
+ case 'A': case 'B': case 'C': case 'D': case 'E': case 'F':
+ case 'G': case 'H': case 'I': case 'J': case 'K': case 'L':
+ case 'M': case 'N': case 'O': case 'P': case 'Q': case 'R':
+ case 'S': case 'T': case 'U': case 'V': case 'W': case 'X':
+ case 'Y': case 'Z':
+ return 1;
+ default:
+ return 0;
+ }
+#endif
+}
+
+bool
+c_isxdigit (int c)
+{
+#if C_CTYPE_CONSECUTIVE_DIGITS \
+ && C_CTYPE_CONSECUTIVE_UPPERCASE && C_CTYPE_CONSECUTIVE_LOWERCASE
+#if C_CTYPE_ASCII
+ return ((c >= '0' && c <= '9')
+ || ((c & ~0x20) >= 'A' && (c & ~0x20) <= 'F'));
+#else
+ return ((c >= '0' && c <= '9')
+ || (c >= 'A' && c <= 'F')
+ || (c >= 'a' && c <= 'f'));
+#endif
+#else
+ switch (c)
+ {
+ case '0': case '1': case '2': case '3': case '4': case '5':
+ case '6': case '7': case '8': case '9':
+ case 'A': case 'B': case 'C': case 'D': case 'E': case 'F':
+ case 'a': case 'b': case 'c': case 'd': case 'e': case 'f':
+ return 1;
+ default:
+ return 0;
+ }
+#endif
+}
+
+int
+c_tolower (int c)
+{
+#if C_CTYPE_CONSECUTIVE_UPPERCASE && C_CTYPE_CONSECUTIVE_LOWERCASE
+ return (c >= 'A' && c <= 'Z' ? c - 'A' + 'a' : c);
+#else
+ switch (c)
+ {
+ case 'A': return 'a';
+ case 'B': return 'b';
+ case 'C': return 'c';
+ case 'D': return 'd';
+ case 'E': return 'e';
+ case 'F': return 'f';
+ case 'G': return 'g';
+ case 'H': return 'h';
+ case 'I': return 'i';
+ case 'J': return 'j';
+ case 'K': return 'k';
+ case 'L': return 'l';
+ case 'M': return 'm';
+ case 'N': return 'n';
+ case 'O': return 'o';
+ case 'P': return 'p';
+ case 'Q': return 'q';
+ case 'R': return 'r';
+ case 'S': return 's';
+ case 'T': return 't';
+ case 'U': return 'u';
+ case 'V': return 'v';
+ case 'W': return 'w';
+ case 'X': return 'x';
+ case 'Y': return 'y';
+ case 'Z': return 'z';
+ default: return c;
+ }
+#endif
+}
+
+int
+c_toupper (int c)
+{
+#if C_CTYPE_CONSECUTIVE_UPPERCASE && C_CTYPE_CONSECUTIVE_LOWERCASE
+ return (c >= 'a' && c <= 'z' ? c - 'a' + 'A' : c);
+#else
+ switch (c)
+ {
+ case 'a': return 'A';
+ case 'b': return 'B';
+ case 'c': return 'C';
+ case 'd': return 'D';
+ case 'e': return 'E';
+ case 'f': return 'F';
+ case 'g': return 'G';
+ case 'h': return 'H';
+ case 'i': return 'I';
+ case 'j': return 'J';
+ case 'k': return 'K';
+ case 'l': return 'L';
+ case 'm': return 'M';
+ case 'n': return 'N';
+ case 'o': return 'O';
+ case 'p': return 'P';
+ case 'q': return 'Q';
+ case 'r': return 'R';
+ case 's': return 'S';
+ case 't': return 'T';
+ case 'u': return 'U';
+ case 'v': return 'V';
+ case 'w': return 'W';
+ case 'x': return 'X';
+ case 'y': return 'Y';
+ case 'z': return 'Z';
+ default: return c;
+ }
+#endif
+}
diff --git a/lib/c-ctype.h b/lib/c-ctype.h
new file mode 100644
index 0000000..d7b067e
--- /dev/null
+++ b/lib/c-ctype.h
@@ -0,0 +1,295 @@
+/* Character handling in C locale.
+
+ These functions work like the corresponding functions in <ctype.h>,
+ except that they have the C (POSIX) locale hardwired, whereas the
+ <ctype.h> functions' behaviour depends on the current locale set via
+ setlocale.
+
+ Copyright (C) 2000-2003, 2006, 2008 Free Software Foundation, Inc.
+
+This program is free software; you can redistribute it and/or modify
+it under the terms of the GNU Lesser General Public License as published by
+the Free Software Foundation; either version 2 of the License, or
+(at your option) any later version.
+
+This program is distributed in the hope that it will be useful,
+but WITHOUT ANY WARRANTY; without even the implied warranty of
+MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+GNU Lesser General Public License for more details.
+
+You should have received a copy of the GNU Lesser General Public License
+along with this program; if not, write to the Free Software Foundation,
+Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */
+
+#ifndef C_CTYPE_H
+#define C_CTYPE_H
+
+#include <stdbool.h>
+
+
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+
+/* The functions defined in this file assume the "C" locale and a character
+ set without diacritics (ASCII-US or EBCDIC-US or something like that).
+ Even if the "C" locale on a particular system is an extension of the ASCII
+ character set (like on BeOS, where it is UTF-8, or on AmigaOS, where it
+ is ISO-8859-1), the functions in this file recognize only the ASCII
+ characters. */
+
+
+/* Check whether the ASCII optimizations apply. */
+
+/* ANSI C89 (and ISO C99 5.2.1.3 too) already guarantees that
+ '0', '1', ..., '9' have consecutive integer values. */
+#define C_CTYPE_CONSECUTIVE_DIGITS 1
+
+#if ('A' <= 'Z') \
+ && ('A' + 1 == 'B') && ('B' + 1 == 'C') && ('C' + 1 == 'D') \
+ && ('D' + 1 == 'E') && ('E' + 1 == 'F') && ('F' + 1 == 'G') \
+ && ('G' + 1 == 'H') && ('H' + 1 == 'I') && ('I' + 1 == 'J') \
+ && ('J' + 1 == 'K') && ('K' + 1 == 'L') && ('L' + 1 == 'M') \
+ && ('M' + 1 == 'N') && ('N' + 1 == 'O') && ('O' + 1 == 'P') \
+ && ('P' + 1 == 'Q') && ('Q' + 1 == 'R') && ('R' + 1 == 'S') \
+ && ('S' + 1 == 'T') && ('T' + 1 == 'U') && ('U' + 1 == 'V') \
+ && ('V' + 1 == 'W') && ('W' + 1 == 'X') && ('X' + 1 == 'Y') \
+ && ('Y' + 1 == 'Z')
+#define C_CTYPE_CONSECUTIVE_UPPERCASE 1
+#endif
+
+#if ('a' <= 'z') \
+ && ('a' + 1 == 'b') && ('b' + 1 == 'c') && ('c' + 1 == 'd') \
+ && ('d' + 1 == 'e') && ('e' + 1 == 'f') && ('f' + 1 == 'g') \
+ && ('g' + 1 == 'h') && ('h' + 1 == 'i') && ('i' + 1 == 'j') \
+ && ('j' + 1 == 'k') && ('k' + 1 == 'l') && ('l' + 1 == 'm') \
+ && ('m' + 1 == 'n') && ('n' + 1 == 'o') && ('o' + 1 == 'p') \
+ && ('p' + 1 == 'q') && ('q' + 1 == 'r') && ('r' + 1 == 's') \
+ && ('s' + 1 == 't') && ('t' + 1 == 'u') && ('u' + 1 == 'v') \
+ && ('v' + 1 == 'w') && ('w' + 1 == 'x') && ('x' + 1 == 'y') \
+ && ('y' + 1 == 'z')
+#define C_CTYPE_CONSECUTIVE_LOWERCASE 1
+#endif
+
+#if (' ' == 32) && ('!' == 33) && ('"' == 34) && ('#' == 35) \
+ && ('%' == 37) && ('&' == 38) && ('\'' == 39) && ('(' == 40) \
+ && (')' == 41) && ('*' == 42) && ('+' == 43) && (',' == 44) \
+ && ('-' == 45) && ('.' == 46) && ('/' == 47) && ('0' == 48) \
+ && ('1' == 49) && ('2' == 50) && ('3' == 51) && ('4' == 52) \
+ && ('5' == 53) && ('6' == 54) && ('7' == 55) && ('8' == 56) \
+ && ('9' == 57) && (':' == 58) && (';' == 59) && ('<' == 60) \
+ && ('=' == 61) && ('>' == 62) && ('?' == 63) && ('A' == 65) \
+ && ('B' == 66) && ('C' == 67) && ('D' == 68) && ('E' == 69) \
+ && ('F' == 70) && ('G' == 71) && ('H' == 72) && ('I' == 73) \
+ && ('J' == 74) && ('K' == 75) && ('L' == 76) && ('M' == 77) \
+ && ('N' == 78) && ('O' == 79) && ('P' == 80) && ('Q' == 81) \
+ && ('R' == 82) && ('S' == 83) && ('T' == 84) && ('U' == 85) \
+ && ('V' == 86) && ('W' == 87) && ('X' == 88) && ('Y' == 89) \
+ && ('Z' == 90) && ('[' == 91) && ('\\' == 92) && (']' == 93) \
+ && ('^' == 94) && ('_' == 95) && ('a' == 97) && ('b' == 98) \
+ && ('c' == 99) && ('d' == 100) && ('e' == 101) && ('f' == 102) \
+ && ('g' == 103) && ('h' == 104) && ('i' == 105) && ('j' == 106) \
+ && ('k' == 107) && ('l' == 108) && ('m' == 109) && ('n' == 110) \
+ && ('o' == 111) && ('p' == 112) && ('q' == 113) && ('r' == 114) \
+ && ('s' == 115) && ('t' == 116) && ('u' == 117) && ('v' == 118) \
+ && ('w' == 119) && ('x' == 120) && ('y' == 121) && ('z' == 122) \
+ && ('{' == 123) && ('|' == 124) && ('}' == 125) && ('~' == 126)
+/* The character set is ASCII or one of its variants or extensions, not EBCDIC.
+ Testing the value of '\n' and '\r' is not relevant. */
+#define C_CTYPE_ASCII 1
+#endif
+
+
+/* Function declarations. */
+
+/* Unlike the functions in <ctype.h>, which require an argument in the range
+ of the 'unsigned char' type, the functions here operate on values that are
+ in the 'unsigned char' range or in the 'char' range. In other words,
+ when you have a 'char' value, you need to cast it before using it as
+ argument to a <ctype.h> function:
+
+ const char *s = ...;
+ if (isalpha ((unsigned char) *s)) ...
+
+ but you don't need to cast it for the functions defined in this file:
+
+ const char *s = ...;
+ if (c_isalpha (*s)) ...
+ */
+
+extern bool c_isascii (int c); /* not locale dependent */
+
+extern bool c_isalnum (int c);
+extern bool c_isalpha (int c);
+extern bool c_isblank (int c);
+extern bool c_iscntrl (int c);
+extern bool c_isdigit (int c);
+extern bool c_islower (int c);
+extern bool c_isgraph (int c);
+extern bool c_isprint (int c);
+extern bool c_ispunct (int c);
+extern bool c_isspace (int c);
+extern bool c_isupper (int c);
+extern bool c_isxdigit (int c);
+
+extern int c_tolower (int c);
+extern int c_toupper (int c);
+
+
+#if defined __GNUC__ && defined __OPTIMIZE__ && !defined __OPTIMIZE_SIZE__ &&
!defined NO_C_CTYPE_MACROS
+
+/* ASCII optimizations. */
+
+#undef c_isascii
+#define c_isascii(c) \
+ ({ int __c = (c); \
+ (__c >= 0x00 && __c <= 0x7f); \
+ })
+
+#if C_CTYPE_CONSECUTIVE_DIGITS \
+ && C_CTYPE_CONSECUTIVE_UPPERCASE && C_CTYPE_CONSECUTIVE_LOWERCASE
+#if C_CTYPE_ASCII
+#undef c_isalnum
+#define c_isalnum(c) \
+ ({ int __c = (c); \
+ ((__c >= '0' && __c <= '9') \
+ || ((__c & ~0x20) >= 'A' && (__c & ~0x20) <= 'Z')); \
+ })
+#else
+#undef c_isalnum
+#define c_isalnum(c) \
+ ({ int __c = (c); \
+ ((__c >= '0' && __c <= '9') \
+ || (__c >= 'A' && __c <= 'Z') \
+ || (__c >= 'a' && __c <= 'z')); \
+ })
+#endif
+#endif
+
+#if C_CTYPE_CONSECUTIVE_UPPERCASE && C_CTYPE_CONSECUTIVE_LOWERCASE
+#if C_CTYPE_ASCII
+#undef c_isalpha
+#define c_isalpha(c) \
+ ({ int __c = (c); \
+ ((__c & ~0x20) >= 'A' && (__c & ~0x20) <= 'Z'); \
+ })
+#else
+#undef c_isalpha
+#define c_isalpha(c) \
+ ({ int __c = (c); \
+ ((__c >= 'A' && __c <= 'Z') || (__c >= 'a' && __c <= 'z')); \
+ })
+#endif
+#endif
+
+#undef c_isblank
+#define c_isblank(c) \
+ ({ int __c = (c); \
+ (__c == ' ' || __c == '\t'); \
+ })
+
+#if C_CTYPE_ASCII
+#undef c_iscntrl
+#define c_iscntrl(c) \
+ ({ int __c = (c); \
+ ((__c & ~0x1f) == 0 || __c == 0x7f); \
+ })
+#endif
+
+#if C_CTYPE_CONSECUTIVE_DIGITS
+#undef c_isdigit
+#define c_isdigit(c) \
+ ({ int __c = (c); \
+ (__c >= '0' && __c <= '9'); \
+ })
+#endif
+
+#if C_CTYPE_CONSECUTIVE_LOWERCASE
+#undef c_islower
+#define c_islower(c) \
+ ({ int __c = (c); \
+ (__c >= 'a' && __c <= 'z'); \
+ })
+#endif
+
+#if C_CTYPE_ASCII
+#undef c_isgraph
+#define c_isgraph(c) \
+ ({ int __c = (c); \
+ (__c >= '!' && __c <= '~'); \
+ })
+#endif
+
+#if C_CTYPE_ASCII
+#undef c_isprint
+#define c_isprint(c) \
+ ({ int __c = (c); \
+ (__c >= ' ' && __c <= '~'); \
+ })
+#endif
+
+#if C_CTYPE_ASCII
+#undef c_ispunct
+#define c_ispunct(c) \
+ ({ int _c = (c); \
+ (c_isgraph (_c) && ! c_isalnum (_c)); \
+ })
+#endif
+
+#undef c_isspace
+#define c_isspace(c) \
+ ({ int __c = (c); \
+ (__c == ' ' || __c == '\t' \
+ || __c == '\n' || __c == '\v' || __c == '\f' || __c == '\r'); \
+ })
+
+#if C_CTYPE_CONSECUTIVE_UPPERCASE
+#undef c_isupper
+#define c_isupper(c) \
+ ({ int __c = (c); \
+ (__c >= 'A' && __c <= 'Z'); \
+ })
+#endif
+
+#if C_CTYPE_CONSECUTIVE_DIGITS \
+ && C_CTYPE_CONSECUTIVE_UPPERCASE && C_CTYPE_CONSECUTIVE_LOWERCASE
+#if C_CTYPE_ASCII
+#undef c_isxdigit
+#define c_isxdigit(c) \
+ ({ int __c = (c); \
+ ((__c >= '0' && __c <= '9') \
+ || ((__c & ~0x20) >= 'A' && (__c & ~0x20) <= 'F')); \
+ })
+#else
+#undef c_isxdigit
+#define c_isxdigit(c) \
+ ({ int __c = (c); \
+ ((__c >= '0' && __c <= '9') \
+ || (__c >= 'A' && __c <= 'F') \
+ || (__c >= 'a' && __c <= 'f')); \
+ })
+#endif
+#endif
+
+#if C_CTYPE_CONSECUTIVE_UPPERCASE && C_CTYPE_CONSECUTIVE_LOWERCASE
+#undef c_tolower
+#define c_tolower(c) \
+ ({ int __c = (c); \
+ (__c >= 'A' && __c <= 'Z' ? __c - 'A' + 'a' : __c); \
+ })
+#undef c_toupper
+#define c_toupper(c) \
+ ({ int __c = (c); \
+ (__c >= 'a' && __c <= 'z' ? __c - 'a' + 'A' : __c); \
+ })
+#endif
+
+#endif /* optimizing for speed */
+
+
+#ifdef __cplusplus
+}
+#endif
+
+#endif /* C_CTYPE_H */
diff --git a/lib/c-strcase.h b/lib/c-strcase.h
new file mode 100644
index 0000000..714a3c6
--- /dev/null
+++ b/lib/c-strcase.h
@@ -0,0 +1,55 @@
+/* Case-insensitive string comparison functions in C locale.
+ Copyright (C) 1995-1996, 2001, 2003, 2005 Free Software Foundation, Inc.
+
+ This program is free software; you can redistribute it and/or modify
+ it under the terms of the GNU Lesser General Public License as published by
+ the Free Software Foundation; either version 2, or (at your option)
+ any later version.
+
+ This program is distributed in the hope that it will be useful,
+ but WITHOUT ANY WARRANTY; without even the implied warranty of
+ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ GNU Lesser General Public License for more details.
+
+ You should have received a copy of the GNU Lesser General Public License
+ along with this program; if not, write to the Free Software Foundation,
+ Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */
+
+#ifndef C_STRCASE_H
+#define C_STRCASE_H
+
+#include <stddef.h>
+
+
+/* The functions defined in this file assume the "C" locale and a character
+ set without diacritics (ASCII-US or EBCDIC-US or something like that).
+ Even if the "C" locale on a particular system is an extension of the ASCII
+ character set (like on BeOS, where it is UTF-8, or on AmigaOS, where it
+ is ISO-8859-1), the functions in this file recognize only the ASCII
+ characters. More precisely, one of the string arguments must be an ASCII
+ string; the other one can also contain non-ASCII characters (but then
+ the comparison result will be nonzero). */
+
+
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+
+/* Compare strings S1 and S2, ignoring case, returning less than, equal to or
+ greater than zero if S1 is lexicographically less than, equal to or greater
+ than S2. */
+extern int c_strcasecmp (const char *s1, const char *s2);
+
+/* Compare no more than N characters of strings S1 and S2, ignoring case,
+ returning less than, equal to or greater than zero if S1 is
+ lexicographically less than, equal to or greater than S2. */
+extern int c_strncasecmp (const char *s1, const char *s2, size_t n);
+
+
+#ifdef __cplusplus
+}
+#endif
+
+
+#endif /* C_STRCASE_H */
diff --git a/lib/c-strcasecmp.c b/lib/c-strcasecmp.c
new file mode 100644
index 0000000..a523898
--- /dev/null
+++ b/lib/c-strcasecmp.c
@@ -0,0 +1,57 @@
+/* c-strcasecmp.c -- case insensitive string comparator in C locale
+ Copyright (C) 1998-1999, 2005-2006 Free Software Foundation, Inc.
+
+ This program is free software; you can redistribute it and/or modify
+ it under the terms of the GNU Lesser General Public License as published by
+ the Free Software Foundation; either version 2, or (at your option)
+ any later version.
+
+ This program is distributed in the hope that it will be useful,
+ but WITHOUT ANY WARRANTY; without even the implied warranty of
+ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ GNU Lesser General Public License for more details.
+
+ You should have received a copy of the GNU Lesser General Public License
+ along with this program; if not, write to the Free Software Foundation,
+ Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */
+
+#include <config.h>
+
+/* Specification. */
+#include "c-strcase.h"
+
+#include <limits.h>
+
+#include "c-ctype.h"
+
+int
+c_strcasecmp (const char *s1, const char *s2)
+{
+ register const unsigned char *p1 = (const unsigned char *) s1;
+ register const unsigned char *p2 = (const unsigned char *) s2;
+ unsigned char c1, c2;
+
+ if (p1 == p2)
+ return 0;
+
+ do
+ {
+ c1 = c_tolower (*p1);
+ c2 = c_tolower (*p2);
+
+ if (c1 == '\0')
+ break;
+
+ ++p1;
+ ++p2;
+ }
+ while (c1 == c2);
+
+ if (UCHAR_MAX <= INT_MAX)
+ return c1 - c2;
+ else
+ /* On machines where 'char' and 'int' are types of the same size, the
+ difference of two 'unsigned char' values - including the sign bit -
+ doesn't fit in an 'int'. */
+ return (c1 > c2 ? 1 : c1 < c2 ? -1 : 0);
+}
diff --git a/lib/c-strcaseeq.h b/lib/c-strcaseeq.h
new file mode 100644
index 0000000..cd29b66
--- /dev/null
+++ b/lib/c-strcaseeq.h
@@ -0,0 +1,184 @@
+/* Optimized case-insensitive string comparison in C locale.
+ Copyright (C) 2001-2002, 2007 Free Software Foundation, Inc.
+
+ This program is free software: you can redistribute it and/or modify it
+ under the terms of the GNU Lesser General Public License as published
+ by the Free Software Foundation; either version 3 of the License, or
+ (at your option) any later version.
+
+ This program is distributed in the hope that it will be useful,
+ but WITHOUT ANY WARRANTY; without even the implied warranty of
+ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+ Lesser General Public License for more details.
+
+ You should have received a copy of the GNU Lesser General Public License
+ along with this program. If not, see <http://www.gnu.org/licenses/>. */
+
+/* Written by Bruno Haible <address@hidden>. */
+
+#include "c-strcase.h"
+#include "c-ctype.h"
+
+/* STRCASEEQ allows to optimize string comparison with a small literal string.
+ STRCASEEQ (s, "UTF-8", 'U','T','F','-','8',0,0,0,0)
+ is semantically equivalent to
+ c_strcasecmp (s, "UTF-8") == 0
+ just faster. */
+
+/* Help GCC to generate good code for string comparisons with
+ immediate strings. */
+#if defined (__GNUC__) && defined (__OPTIMIZE__)
+
+/* Case insensitive comparison of ASCII characters. */
+# if C_CTYPE_ASCII
+# define CASEEQ(other,upper) \
+ (c_isupper (upper) ? ((other) & ~0x20) == (upper) : (other) == (upper))
+# elif C_CTYPE_CONSECUTIVE_UPPERCASE && C_CTYPE_CONSECUTIVE_LOWERCASE
+# define CASEEQ(other,upper) \
+ (c_isupper (upper) ? (other) == (upper) || (other) == (upper) - 'A' + 'a'
: (other) == (upper))
+# else
+# define CASEEQ(other,upper) \
+ (c_toupper (other) == (upper))
+# endif
+
+static inline int
+strcaseeq9 (const char *s1, const char *s2)
+{
+ return c_strcasecmp (s1 + 9, s2 + 9) == 0;
+}
+
+static inline int
+strcaseeq8 (const char *s1, const char *s2, char s28)
+{
+ if (CASEEQ (s1[8], s28))
+ {
+ if (s28 == 0)
+ return 1;
+ else
+ return strcaseeq9 (s1, s2);
+ }
+ else
+ return 0;
+}
+
+static inline int
+strcaseeq7 (const char *s1, const char *s2, char s27, char s28)
+{
+ if (CASEEQ (s1[7], s27))
+ {
+ if (s27 == 0)
+ return 1;
+ else
+ return strcaseeq8 (s1, s2, s28);
+ }
+ else
+ return 0;
+}
+
+static inline int
+strcaseeq6 (const char *s1, const char *s2, char s26, char s27, char s28)
+{
+ if (CASEEQ (s1[6], s26))
+ {
+ if (s26 == 0)
+ return 1;
+ else
+ return strcaseeq7 (s1, s2, s27, s28);
+ }
+ else
+ return 0;
+}
+
+static inline int
+strcaseeq5 (const char *s1, const char *s2, char s25, char s26, char s27, char
s28)
+{
+ if (CASEEQ (s1[5], s25))
+ {
+ if (s25 == 0)
+ return 1;
+ else
+ return strcaseeq6 (s1, s2, s26, s27, s28);
+ }
+ else
+ return 0;
+}
+
+static inline int
+strcaseeq4 (const char *s1, const char *s2, char s24, char s25, char s26, char
s27, char s28)
+{
+ if (CASEEQ (s1[4], s24))
+ {
+ if (s24 == 0)
+ return 1;
+ else
+ return strcaseeq5 (s1, s2, s25, s26, s27, s28);
+ }
+ else
+ return 0;
+}
+
+static inline int
+strcaseeq3 (const char *s1, const char *s2, char s23, char s24, char s25, char
s26, char s27, char s28)
+{
+ if (CASEEQ (s1[3], s23))
+ {
+ if (s23 == 0)
+ return 1;
+ else
+ return strcaseeq4 (s1, s2, s24, s25, s26, s27, s28);
+ }
+ else
+ return 0;
+}
+
+static inline int
+strcaseeq2 (const char *s1, const char *s2, char s22, char s23, char s24, char
s25, char s26, char s27, char s28)
+{
+ if (CASEEQ (s1[2], s22))
+ {
+ if (s22 == 0)
+ return 1;
+ else
+ return strcaseeq3 (s1, s2, s23, s24, s25, s26, s27, s28);
+ }
+ else
+ return 0;
+}
+
+static inline int
+strcaseeq1 (const char *s1, const char *s2, char s21, char s22, char s23, char
s24, char s25, char s26, char s27, char s28)
+{
+ if (CASEEQ (s1[1], s21))
+ {
+ if (s21 == 0)
+ return 1;
+ else
+ return strcaseeq2 (s1, s2, s22, s23, s24, s25, s26, s27, s28);
+ }
+ else
+ return 0;
+}
+
+static inline int
+strcaseeq0 (const char *s1, const char *s2, char s20, char s21, char s22, char
s23, char s24, char s25, char s26, char s27, char s28)
+{
+ if (CASEEQ (s1[0], s20))
+ {
+ if (s20 == 0)
+ return 1;
+ else
+ return strcaseeq1 (s1, s2, s21, s22, s23, s24, s25, s26, s27, s28);
+ }
+ else
+ return 0;
+}
+
+#define STRCASEEQ(s1,s2,s20,s21,s22,s23,s24,s25,s26,s27,s28) \
+ strcaseeq0 (s1, s2, s20, s21, s22, s23, s24, s25, s26, s27, s28)
+
+#else
+
+#define STRCASEEQ(s1,s2,s20,s21,s22,s23,s24,s25,s26,s27,s28) \
+ (c_strcasecmp (s1, s2) == 0)
+
+#endif
diff --git a/lib/c-strncasecmp.c b/lib/c-strncasecmp.c
new file mode 100644
index 0000000..c1496ca
--- /dev/null
+++ b/lib/c-strncasecmp.c
@@ -0,0 +1,57 @@
+/* c-strncasecmp.c -- case insensitive string comparator in C locale
+ Copyright (C) 1998-1999, 2005-2006 Free Software Foundation, Inc.
+
+ This program is free software; you can redistribute it and/or modify
+ it under the terms of the GNU Lesser General Public License as published by
+ the Free Software Foundation; either version 2, or (at your option)
+ any later version.
+
+ This program is distributed in the hope that it will be useful,
+ but WITHOUT ANY WARRANTY; without even the implied warranty of
+ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ GNU Lesser General Public License for more details.
+
+ You should have received a copy of the GNU Lesser General Public License
+ along with this program; if not, write to the Free Software Foundation,
+ Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */
+
+#include <config.h>
+
+/* Specification. */
+#include "c-strcase.h"
+
+#include <limits.h>
+
+#include "c-ctype.h"
+
+int
+c_strncasecmp (const char *s1, const char *s2, size_t n)
+{
+ register const unsigned char *p1 = (const unsigned char *) s1;
+ register const unsigned char *p2 = (const unsigned char *) s2;
+ unsigned char c1, c2;
+
+ if (p1 == p2 || n == 0)
+ return 0;
+
+ do
+ {
+ c1 = c_tolower (*p1);
+ c2 = c_tolower (*p2);
+
+ if (--n == 0 || c1 == '\0')
+ break;
+
+ ++p1;
+ ++p2;
+ }
+ while (c1 == c2);
+
+ if (UCHAR_MAX <= INT_MAX)
+ return c1 - c2;
+ else
+ /* On machines where 'char' and 'int' are types of the same size, the
+ difference of two 'unsigned char' values - including the sign bit -
+ doesn't fit in an 'int'. */
+ return (c1 > c2 ? 1 : c1 < c2 ? -1 : 0);
+}
diff --git a/lib/iconv.c b/lib/iconv.c
new file mode 100644
index 0000000..56a84c4
--- /dev/null
+++ b/lib/iconv.c
@@ -0,0 +1,450 @@
+/* Character set conversion.
+ Copyright (C) 1999-2001, 2007 Free Software Foundation, Inc.
+
+ This program is free software; you can redistribute it and/or modify
+ it under the terms of the GNU Lesser General Public License as published by
+ the Free Software Foundation; either version 2, or (at your option)
+ any later version.
+
+ This program is distributed in the hope that it will be useful,
+ but WITHOUT ANY WARRANTY; without even the implied warranty of
+ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ GNU Lesser General Public License for more details.
+
+ You should have received a copy of the GNU Lesser General Public License
along
+ with this program; if not, write to the Free Software Foundation,
+ Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */
+
+#include <config.h>
+
+/* Specification. */
+#include <iconv.h>
+
+#include <stddef.h>
+
+#if REPLACE_ICONV_UTF
+# include <errno.h>
+# include <stdint.h>
+# include <stdlib.h>
+# include "unistr.h"
+# ifndef uintptr_t
+# define uintptr_t unsigned long
+# endif
+#endif
+
+#if REPLACE_ICONV_UTF
+
+/* UTF-{16,32}{BE,LE} converters taken from GNU libiconv 1.11. */
+
+/* Return code if invalid. (xxx_mbtowc) */
+# define RET_ILSEQ -1
+/* Return code if no bytes were read. (xxx_mbtowc) */
+# define RET_TOOFEW -2
+
+/* Return code if invalid. (xxx_wctomb) */
+# define RET_ILUNI -1
+/* Return code if output buffer is too small. (xxx_wctomb, xxx_reset) */
+# define RET_TOOSMALL -2
+
+/*
+ * UTF-16BE
+ */
+
+/* Specification: RFC 2781 */
+
+static int
+utf16be_mbtowc (ucs4_t *pwc, const unsigned char *s, size_t n)
+{
+ if (n >= 2)
+ {
+ ucs4_t wc = (s[0] << 8) + s[1];
+ if (wc >= 0xd800 && wc < 0xdc00)
+ {
+ if (n >= 4)
+ {
+ ucs4_t wc2 = (s[2] << 8) + s[3];
+ if (!(wc2 >= 0xdc00 && wc2 < 0xe000))
+ return RET_ILSEQ;
+ *pwc = 0x10000 + ((wc - 0xd800) << 10) + (wc2 - 0xdc00);
+ return 4;
+ }
+ }
+ else if (wc >= 0xdc00 && wc < 0xe000)
+ {
+ return RET_ILSEQ;
+ }
+ else
+ {
+ *pwc = wc;
+ return 2;
+ }
+ }
+ return RET_TOOFEW;
+}
+
+static int
+utf16be_wctomb (unsigned char *r, ucs4_t wc, size_t n)
+{
+ if (!(wc >= 0xd800 && wc < 0xe000))
+ {
+ if (wc < 0x10000)
+ {
+ if (n >= 2)
+ {
+ r[0] = (unsigned char) (wc >> 8);
+ r[1] = (unsigned char) wc;
+ return 2;
+ }
+ else
+ return RET_TOOSMALL;
+ }
+ else if (wc < 0x110000)
+ {
+ if (n >= 4)
+ {
+ ucs4_t wc1 = 0xd800 + ((wc - 0x10000) >> 10);
+ ucs4_t wc2 = 0xdc00 + ((wc - 0x10000) & 0x3ff);
+ r[0] = (unsigned char) (wc1 >> 8);
+ r[1] = (unsigned char) wc1;
+ r[2] = (unsigned char) (wc2 >> 8);
+ r[3] = (unsigned char) wc2;
+ return 4;
+ }
+ else
+ return RET_TOOSMALL;
+ }
+ }
+ return RET_ILUNI;
+}
+
+/*
+ * UTF-16LE
+ */
+
+/* Specification: RFC 2781 */
+
+static int
+utf16le_mbtowc (ucs4_t *pwc, const unsigned char *s, size_t n)
+{
+ if (n >= 2)
+ {
+ ucs4_t wc = s[0] + (s[1] << 8);
+ if (wc >= 0xd800 && wc < 0xdc00)
+ {
+ if (n >= 4)
+ {
+ ucs4_t wc2 = s[2] + (s[3] << 8);
+ if (!(wc2 >= 0xdc00 && wc2 < 0xe000))
+ return RET_ILSEQ;
+ *pwc = 0x10000 + ((wc - 0xd800) << 10) + (wc2 - 0xdc00);
+ return 4;
+ }
+ }
+ else if (wc >= 0xdc00 && wc < 0xe000)
+ {
+ return RET_ILSEQ;
+ }
+ else
+ {
+ *pwc = wc;
+ return 2;
+ }
+ }
+ return RET_TOOFEW;
+}
+
+static int
+utf16le_wctomb (unsigned char *r, ucs4_t wc, size_t n)
+{
+ if (!(wc >= 0xd800 && wc < 0xe000))
+ {
+ if (wc < 0x10000)
+ {
+ if (n >= 2)
+ {
+ r[0] = (unsigned char) wc;
+ r[1] = (unsigned char) (wc >> 8);
+ return 2;
+ }
+ else
+ return RET_TOOSMALL;
+ }
+ else if (wc < 0x110000)
+ {
+ if (n >= 4)
+ {
+ ucs4_t wc1 = 0xd800 + ((wc - 0x10000) >> 10);
+ ucs4_t wc2 = 0xdc00 + ((wc - 0x10000) & 0x3ff);
+ r[0] = (unsigned char) wc1;
+ r[1] = (unsigned char) (wc1 >> 8);
+ r[2] = (unsigned char) wc2;
+ r[3] = (unsigned char) (wc2 >> 8);
+ return 4;
+ }
+ else
+ return RET_TOOSMALL;
+ }
+ }
+ return RET_ILUNI;
+}
+
+/*
+ * UTF-32BE
+ */
+
+/* Specification: Unicode 3.1 Standard Annex #19 */
+
+static int
+utf32be_mbtowc (ucs4_t *pwc, const unsigned char *s, size_t n)
+{
+ if (n >= 4)
+ {
+ ucs4_t wc = (s[0] << 24) + (s[1] << 16) + (s[2] << 8) + s[3];
+ if (wc < 0x110000 && !(wc >= 0xd800 && wc < 0xe000))
+ {
+ *pwc = wc;
+ return 4;
+ }
+ else
+ return RET_ILSEQ;
+ }
+ return RET_TOOFEW;
+}
+
+static int
+utf32be_wctomb (unsigned char *r, ucs4_t wc, size_t n)
+{
+ if (wc < 0x110000 && !(wc >= 0xd800 && wc < 0xe000))
+ {
+ if (n >= 4)
+ {
+ r[0] = 0;
+ r[1] = (unsigned char) (wc >> 16);
+ r[2] = (unsigned char) (wc >> 8);
+ r[3] = (unsigned char) wc;
+ return 4;
+ }
+ else
+ return RET_TOOSMALL;
+ }
+ return RET_ILUNI;
+}
+
+/*
+ * UTF-32LE
+ */
+
+/* Specification: Unicode 3.1 Standard Annex #19 */
+
+static int
+utf32le_mbtowc (ucs4_t *pwc, const unsigned char *s, size_t n)
+{
+ if (n >= 4)
+ {
+ ucs4_t wc = s[0] + (s[1] << 8) + (s[2] << 16) + (s[3] << 24);
+ if (wc < 0x110000 && !(wc >= 0xd800 && wc < 0xe000))
+ {
+ *pwc = wc;
+ return 4;
+ }
+ else
+ return RET_ILSEQ;
+ }
+ return RET_TOOFEW;
+}
+
+static int
+utf32le_wctomb (unsigned char *r, ucs4_t wc, size_t n)
+{
+ if (wc < 0x110000 && !(wc >= 0xd800 && wc < 0xe000))
+ {
+ if (n >= 4)
+ {
+ r[0] = (unsigned char) wc;
+ r[1] = (unsigned char) (wc >> 8);
+ r[2] = (unsigned char) (wc >> 16);
+ r[3] = 0;
+ return 4;
+ }
+ else
+ return RET_TOOSMALL;
+ }
+ return RET_ILUNI;
+}
+
+#endif
+
+size_t
+rpl_iconv (iconv_t cd,
+ ICONV_CONST char **inbuf, size_t *inbytesleft,
+ char **outbuf, size_t *outbytesleft)
+#undef iconv
+{
+#if REPLACE_ICONV_UTF
+ switch ((uintptr_t) cd)
+ {
+ {
+ int (*xxx_wctomb) (unsigned char *, ucs4_t, size_t);
+
+ case (uintptr_t) _ICONV_UTF8_UTF16BE:
+ xxx_wctomb = utf16be_wctomb;
+ goto loop_from_utf8;
+ case (uintptr_t) _ICONV_UTF8_UTF16LE:
+ xxx_wctomb = utf16le_wctomb;
+ goto loop_from_utf8;
+ case (uintptr_t) _ICONV_UTF8_UTF32BE:
+ xxx_wctomb = utf32be_wctomb;
+ goto loop_from_utf8;
+ case (uintptr_t) _ICONV_UTF8_UTF32LE:
+ xxx_wctomb = utf32le_wctomb;
+ goto loop_from_utf8;
+
+ loop_from_utf8:
+ if (inbuf == NULL || *inbuf == NULL)
+ return 0;
+ {
+ ICONV_CONST char *inptr = *inbuf;
+ size_t inleft = *inbytesleft;
+ char *outptr = *outbuf;
+ size_t outleft = *outbytesleft;
+ size_t res = 0;
+ while (inleft > 0)
+ {
+ ucs4_t uc;
+ int m = u8_mbtoucr (&uc, (const uint8_t *) inptr, inleft);
+ if (m <= 0)
+ {
+ if (m == -1)
+ {
+ errno = EILSEQ;
+ res = (size_t)(-1);
+ break;
+ }
+ if (m == -2)
+ {
+ errno = EINVAL;
+ res = (size_t)(-1);
+ break;
+ }
+ abort ();
+ }
+ else
+ {
+ int n = xxx_wctomb ((uint8_t *) outptr, uc, outleft);
+ if (n < 0)
+ {
+ if (n == RET_ILUNI)
+ {
+ errno = EILSEQ;
+ res = (size_t)(-1);
+ break;
+ }
+ if (n == RET_TOOSMALL)
+ {
+ errno = E2BIG;
+ res = (size_t)(-1);
+ break;
+ }
+ abort ();
+ }
+ else
+ {
+ inptr += m;
+ inleft -= m;
+ outptr += n;
+ outleft -= n;
+ }
+ }
+ }
+ *inbuf = inptr;
+ *inbytesleft = inleft;
+ *outbuf = outptr;
+ *outbytesleft = outleft;
+ return res;
+ }
+ }
+
+ {
+ int (*xxx_mbtowc) (ucs4_t *, const unsigned char *, size_t);
+
+ case (uintptr_t) _ICONV_UTF16BE_UTF8:
+ xxx_mbtowc = utf16be_mbtowc;
+ goto loop_to_utf8;
+ case (uintptr_t) _ICONV_UTF16LE_UTF8:
+ xxx_mbtowc = utf16le_mbtowc;
+ goto loop_to_utf8;
+ case (uintptr_t) _ICONV_UTF32BE_UTF8:
+ xxx_mbtowc = utf32be_mbtowc;
+ goto loop_to_utf8;
+ case (uintptr_t) _ICONV_UTF32LE_UTF8:
+ xxx_mbtowc = utf32le_mbtowc;
+ goto loop_to_utf8;
+
+ loop_to_utf8:
+ if (inbuf == NULL || *inbuf == NULL)
+ return 0;
+ {
+ ICONV_CONST char *inptr = *inbuf;
+ size_t inleft = *inbytesleft;
+ char *outptr = *outbuf;
+ size_t outleft = *outbytesleft;
+ size_t res = 0;
+ while (inleft > 0)
+ {
+ ucs4_t uc;
+ int m = xxx_mbtowc (&uc, (const uint8_t *) inptr, inleft);
+ if (m <= 0)
+ {
+ if (m == RET_ILSEQ)
+ {
+ errno = EILSEQ;
+ res = (size_t)(-1);
+ break;
+ }
+ if (m == RET_TOOFEW)
+ {
+ errno = EINVAL;
+ res = (size_t)(-1);
+ break;
+ }
+ abort ();
+ }
+ else
+ {
+ int n = u8_uctomb ((uint8_t *) outptr, uc, outleft);
+ if (n < 0)
+ {
+ if (n == -1)
+ {
+ errno = EILSEQ;
+ res = (size_t)(-1);
+ break;
+ }
+ if (n == -2)
+ {
+ errno = E2BIG;
+ res = (size_t)(-1);
+ break;
+ }
+ abort ();
+ }
+ else
+ {
+ inptr += m;
+ inleft -= m;
+ outptr += n;
+ outleft -= n;
+ }
+ }
+ }
+ *inbuf = inptr;
+ *inbytesleft = inleft;
+ *outbuf = outptr;
+ *outbytesleft = outleft;
+ return res;
+ }
+ }
+ }
+#endif
+ return iconv (cd, inbuf, inbytesleft, outbuf, outbytesleft);
+}
diff --git a/lib/iconv.in.h b/lib/iconv.in.h
new file mode 100644
index 0000000..915dce2
--- /dev/null
+++ b/lib/iconv.in.h
@@ -0,0 +1,71 @@
+/* A GNU-like <iconv.h>.
+
+ Copyright (C) 2007-2008 Free Software Foundation, Inc.
+
+ This program is free software; you can redistribute it and/or modify
+ it under the terms of the GNU Lesser General Public License as published by
+ the Free Software Foundation; either version 2, or (at your option)
+ any later version.
+
+ This program is distributed in the hope that it will be useful,
+ but WITHOUT ANY WARRANTY; without even the implied warranty of
+ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ GNU Lesser General Public License for more details.
+
+ You should have received a copy of the GNU Lesser General Public License
+ along with this program; if not, write to the Free Software Foundation,
+ Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */
+
+#ifndef _GL_ICONV_H
+
+#if __GNUC__ >= 3
address@hidden@
+#endif
+
+/* The include_next requires a split double-inclusion guard. */
address@hidden@ @NEXT_ICONV_H@
+
+#ifndef _GL_ICONV_H
+#define _GL_ICONV_H
+
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+
+#if @REPLACE_ICONV_OPEN@
+/* An iconv_open wrapper that supports the IANA standardized encoding names
+ ("ISO-8859-1" etc.) as far as possible. */
+# define iconv_open rpl_iconv_open
+extern iconv_t iconv_open (const char *tocode, const char *fromcode);
+#endif
+
+#if @REPLACE_ICONV_UTF@
+/* Special constants for supporting UTF-{16,32}{BE,LE} encodings.
+ Not public. */
+# define _ICONV_UTF8_UTF16BE (iconv_t)(-161)
+# define _ICONV_UTF8_UTF16LE (iconv_t)(-162)
+# define _ICONV_UTF8_UTF32BE (iconv_t)(-163)
+# define _ICONV_UTF8_UTF32LE (iconv_t)(-164)
+# define _ICONV_UTF16BE_UTF8 (iconv_t)(-165)
+# define _ICONV_UTF16LE_UTF8 (iconv_t)(-166)
+# define _ICONV_UTF32BE_UTF8 (iconv_t)(-167)
+# define _ICONV_UTF32LE_UTF8 (iconv_t)(-168)
+#endif
+
+#if @REPLACE_ICONV@
+# define iconv rpl_iconv
+extern size_t iconv (iconv_t cd,
+ @ICONV_CONST@ char **inbuf, size_t *inbytesleft,
+ char **outbuf, size_t *outbytesleft);
+# define iconv_close rpl_iconv_close
+extern int iconv_close (iconv_t cd);
+#endif
+
+
+#ifdef __cplusplus
+}
+#endif
+
+#endif /* _GL_ICONV_H */
+#endif /* _GL_ICONV_H */
diff --git a/lib/iconv_close.c b/lib/iconv_close.c
new file mode 100644
index 0000000..3680412
--- /dev/null
+++ b/lib/iconv_close.c
@@ -0,0 +1,47 @@
+/* Character set conversion.
+ Copyright (C) 2007 Free Software Foundation, Inc.
+
+ This program is free software; you can redistribute it and/or modify
+ it under the terms of the GNU Lesser General Public License as published by
+ the Free Software Foundation; either version 2, or (at your option)
+ any later version.
+
+ This program is distributed in the hope that it will be useful,
+ but WITHOUT ANY WARRANTY; without even the implied warranty of
+ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ GNU Lesser General Public License for more details.
+
+ You should have received a copy of the GNU Lesser General Public License
along
+ with this program; if not, write to the Free Software Foundation,
+ Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */
+
+#include <config.h>
+
+/* Specification. */
+#include <iconv.h>
+
+#include <stdint.h>
+#ifndef uintptr_t
+# define uintptr_t unsigned long
+#endif
+
+int
+rpl_iconv_close (iconv_t cd)
+#undef iconv_close
+{
+#if REPLACE_ICONV_UTF
+ switch ((uintptr_t) cd)
+ {
+ case (uintptr_t) _ICONV_UTF8_UTF16BE:
+ case (uintptr_t) _ICONV_UTF8_UTF16LE:
+ case (uintptr_t) _ICONV_UTF8_UTF32BE:
+ case (uintptr_t) _ICONV_UTF8_UTF32LE:
+ case (uintptr_t) _ICONV_UTF16BE_UTF8:
+ case (uintptr_t) _ICONV_UTF16LE_UTF8:
+ case (uintptr_t) _ICONV_UTF32BE_UTF8:
+ case (uintptr_t) _ICONV_UTF32LE_UTF8:
+ return 0;
+ }
+#endif
+ return iconv_close (cd);
+}
diff --git a/lib/iconv_open-aix.gperf b/lib/iconv_open-aix.gperf
new file mode 100644
index 0000000..6782b99
--- /dev/null
+++ b/lib/iconv_open-aix.gperf
@@ -0,0 +1,44 @@
+struct mapping { int standard_name; const char vendor_name[10 + 1]; };
+%struct-type
+%language=ANSI-C
+%define slot-name standard_name
+%define hash-function-name mapping_hash
+%define lookup-function-name mapping_lookup
+%readonly-tables
+%global-table
+%define word-array-name mappings
+%pic
+%%
+# On AIX 5.1, look in /usr/lib/nls/loc/uconvTable.
+ISO-8859-1, "ISO8859-1"
+ISO-8859-2, "ISO8859-2"
+ISO-8859-3, "ISO8859-3"
+ISO-8859-4, "ISO8859-4"
+ISO-8859-5, "ISO8859-5"
+ISO-8859-6, "ISO8859-6"
+ISO-8859-7, "ISO8859-7"
+ISO-8859-8, "ISO8859-8"
+ISO-8859-9, "ISO8859-9"
+ISO-8859-15, "ISO8859-15"
+CP437, "IBM-437"
+CP850, "IBM-850"
+CP852, "IBM-852"
+CP856, "IBM-856"
+CP857, "IBM-857"
+CP861, "IBM-861"
+CP865, "IBM-865"
+CP869, "IBM-869"
+ISO-8859-13, "IBM-921"
+CP922, "IBM-922"
+CP932, "IBM-932"
+CP943, "IBM-943"
+CP1046, "IBM-1046"
+CP1124, "IBM-1124"
+CP1125, "IBM-1125"
+CP1129, "IBM-1129"
+CP1252, "IBM-1252"
+GB2312, "IBM-eucCN"
+EUC-JP, "IBM-eucJP"
+EUC-KR, "IBM-eucKR"
+EUC-TW, "IBM-eucTW"
+BIG5, "big5"
diff --git a/lib/iconv_open-hpux.gperf b/lib/iconv_open-hpux.gperf
new file mode 100644
index 0000000..5a35c83
--- /dev/null
+++ b/lib/iconv_open-hpux.gperf
@@ -0,0 +1,56 @@
+struct mapping { int standard_name; const char vendor_name[9 + 1]; };
+%struct-type
+%language=ANSI-C
+%define slot-name standard_name
+%define hash-function-name mapping_hash
+%define lookup-function-name mapping_lookup
+%readonly-tables
+%global-table
+%define word-array-name mappings
+%pic
+%%
+# On HP-UX 11.11, look in /usr/lib/nls/iconv.
+ISO-8859-1, "iso88591"
+ISO-8859-2, "iso88592"
+ISO-8859-5, "iso88595"
+ISO-8859-6, "iso88596"
+ISO-8859-7, "iso88597"
+ISO-8859-8, "iso88598"
+ISO-8859-9, "iso88599"
+ISO-8859-15, "iso885915"
+CP437, "cp437"
+CP775, "cp775"
+CP850, "cp850"
+CP852, "cp852"
+CP855, "cp855"
+CP857, "cp857"
+CP861, "cp861"
+CP862, "cp862"
+CP864, "cp864"
+CP865, "cp865"
+CP866, "cp866"
+CP869, "cp869"
+CP874, "cp874"
+CP1250, "cp1250"
+CP1251, "cp1251"
+CP1252, "cp1252"
+CP1253, "cp1253"
+CP1254, "cp1254"
+CP1255, "cp1255"
+CP1256, "cp1256"
+CP1257, "cp1257"
+CP1258, "cp1258"
+HP-ROMAN8, "roman8"
+HP-ARABIC8, "arabic8"
+HP-GREEK8, "greek8"
+HP-HEBREW8, "hebrew8"
+HP-TURKISH8, "turkish8"
+HP-KANA8, "kana8"
+TIS-620, "tis620"
+GB2312, "hp15CN"
+EUC-JP, "eucJP"
+EUC-KR, "eucKR"
+EUC-TW, "eucTW"
+BIG5, "big5"
+SHIFT_JIS, "sjis"
+UTF-8, "utf8"
diff --git a/lib/iconv_open-irix.gperf b/lib/iconv_open-irix.gperf
new file mode 100644
index 0000000..3672a80
--- /dev/null
+++ b/lib/iconv_open-irix.gperf
@@ -0,0 +1,31 @@
+struct mapping { int standard_name; const char vendor_name[10 + 1]; };
+%struct-type
+%language=ANSI-C
+%define slot-name standard_name
+%define hash-function-name mapping_hash
+%define lookup-function-name mapping_lookup
+%readonly-tables
+%global-table
+%define word-array-name mappings
+%pic
+%%
+# On IRIX 6.5, look in /usr/lib/iconv and /usr/lib/international/encodings.
+ISO-8859-1, "ISO8859-1"
+ISO-8859-2, "ISO8859-2"
+ISO-8859-3, "ISO8859-3"
+ISO-8859-4, "ISO8859-4"
+ISO-8859-5, "ISO8859-5"
+ISO-8859-6, "ISO8859-6"
+ISO-8859-7, "ISO8859-7"
+ISO-8859-8, "ISO8859-8"
+ISO-8859-9, "ISO8859-9"
+ISO-8859-15, "ISO8859-15"
+KOI8-R, "KOI8"
+CP855, "DOS855"
+CP1251, "WIN1251"
+GB2312, "eucCN"
+EUC-JP, "eucJP"
+EUC-KR, "eucKR"
+EUC-TW, "eucTW"
+SHIFT_JIS, "sjis"
+TIS-620, "TIS620"
diff --git a/lib/iconv_open-osf.gperf b/lib/iconv_open-osf.gperf
new file mode 100644
index 0000000..f468ff6
--- /dev/null
+++ b/lib/iconv_open-osf.gperf
@@ -0,0 +1,50 @@
+struct mapping { int standard_name; const char vendor_name[10 + 1]; };
+%struct-type
+%language=ANSI-C
+%define slot-name standard_name
+%define hash-function-name mapping_hash
+%define lookup-function-name mapping_lookup
+%readonly-tables
+%global-table
+%define word-array-name mappings
+%pic
+%%
+# On OSF/1 5.1, look in /usr/lib/nls/loc/iconv.
+ISO-8859-1, "ISO8859-1"
+ISO-8859-2, "ISO8859-2"
+ISO-8859-3, "ISO8859-3"
+ISO-8859-4, "ISO8859-4"
+ISO-8859-5, "ISO8859-5"
+ISO-8859-6, "ISO8859-6"
+ISO-8859-7, "ISO8859-7"
+ISO-8859-8, "ISO8859-8"
+ISO-8859-9, "ISO8859-9"
+ISO-8859-15, "ISO8859-15"
+CP437, "cp437"
+CP775, "cp775"
+CP850, "cp850"
+CP852, "cp852"
+CP855, "cp855"
+CP857, "cp857"
+CP861, "cp861"
+CP862, "cp862"
+CP865, "cp865"
+CP866, "cp866"
+CP869, "cp869"
+CP874, "cp874"
+CP949, "KSC5601"
+CP1250, "cp1250"
+CP1251, "cp1251"
+CP1252, "cp1252"
+CP1253, "cp1253"
+CP1254, "cp1254"
+CP1255, "cp1255"
+CP1256, "cp1256"
+CP1257, "cp1257"
+CP1258, "cp1258"
+EUC-JP, "eucJP"
+EUC-KR, "eucKR"
+EUC-TW, "eucTW"
+BIG5, "big5"
+SHIFT_JIS, "SJIS"
+TIS-620, "TACTIS"
diff --git a/lib/iconv_open.c b/lib/iconv_open.c
new file mode 100644
index 0000000..3d873ac
--- /dev/null
+++ b/lib/iconv_open.c
@@ -0,0 +1,172 @@
+/* Character set conversion.
+ Copyright (C) 2007 Free Software Foundation, Inc.
+
+ This program is free software; you can redistribute it and/or modify
+ it under the terms of the GNU Lesser General Public License as published by
+ the Free Software Foundation; either version 2, or (at your option)
+ any later version.
+
+ This program is distributed in the hope that it will be useful,
+ but WITHOUT ANY WARRANTY; without even the implied warranty of
+ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ GNU Lesser General Public License for more details.
+
+ You should have received a copy of the GNU Lesser General Public License
along
+ with this program; if not, write to the Free Software Foundation,
+ Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */
+
+#include <config.h>
+
+/* Specification. */
+#include <iconv.h>
+
+#include <errno.h>
+#include <string.h>
+#include "c-ctype.h"
+#include "c-strcase.h"
+
+#define SIZEOF(a) (sizeof(a) / sizeof(a[0]))
+
+/* Namespace cleanliness. */
+#define mapping_lookup rpl_iconv_open_mapping_lookup
+
+/* The macro ICONV_FLAVOR is defined to one of these or undefined. */
+
+#define ICONV_FLAVOR_AIX "iconv_open-aix.h"
+#define ICONV_FLAVOR_HPUX "iconv_open-hpux.h"
+#define ICONV_FLAVOR_IRIX "iconv_open-irix.h"
+#define ICONV_FLAVOR_OSF "iconv_open-osf.h"
+
+#ifdef ICONV_FLAVOR
+# include ICONV_FLAVOR
+#endif
+
+iconv_t
+rpl_iconv_open (const char *tocode, const char *fromcode)
+#undef iconv_open
+{
+ char fromcode_upper[32];
+ char tocode_upper[32];
+ char *fromcode_upper_end;
+ char *tocode_upper_end;
+
+#if REPLACE_ICONV_UTF
+ /* Special handling of conversion between UTF-8 and UTF-{16,32}{BE,LE}.
+ Do this here, before calling the real iconv_open(), because OSF/1 5.1
+ iconv() to these encoding inserts a BOM, which is wrong.
+ We do not need to handle conversion between arbitrary encodings and
+ UTF-{16,32}{BE,LE}, because the 'striconveh' module implements two-step
+ conversion throough UTF-8.
+ The _ICONV_* constants are chosen to be disjoint from any iconv_t
+ returned by the system's iconv_open() functions. Recall that iconv_t
+ is a scalar type. */
+ if (c_toupper (fromcode[0]) == 'U'
+ && c_toupper (fromcode[1]) == 'T'
+ && c_toupper (fromcode[2]) == 'F'
+ && fromcode[3] == '-')
+ {
+ if (c_toupper (tocode[0]) == 'U'
+ && c_toupper (tocode[1]) == 'T'
+ && c_toupper (tocode[2]) == 'F'
+ && tocode[3] == '-')
+ {
+ if (strcmp (fromcode + 4, "8") == 0)
+ {
+ if (c_strcasecmp (tocode + 4, "16BE") == 0)
+ return _ICONV_UTF8_UTF16BE;
+ if (c_strcasecmp (tocode + 4, "16LE") == 0)
+ return _ICONV_UTF8_UTF16LE;
+ if (c_strcasecmp (tocode + 4, "32BE") == 0)
+ return _ICONV_UTF8_UTF32BE;
+ if (c_strcasecmp (tocode + 4, "32LE") == 0)
+ return _ICONV_UTF8_UTF32LE;
+ }
+ else if (strcmp (tocode + 4, "8") == 0)
+ {
+ if (c_strcasecmp (fromcode + 4, "16BE") == 0)
+ return _ICONV_UTF16BE_UTF8;
+ if (c_strcasecmp (fromcode + 4, "16LE") == 0)
+ return _ICONV_UTF16LE_UTF8;
+ if (c_strcasecmp (fromcode + 4, "32BE") == 0)
+ return _ICONV_UTF32BE_UTF8;
+ if (c_strcasecmp (fromcode + 4, "32LE") == 0)
+ return _ICONV_UTF32LE_UTF8;
+ }
+ }
+ }
+#endif
+
+ /* Do *not* add special support for 8-bit encodings like ASCII or ISO-8859-1
+ here. This would lead to programs that work in some locales (such as the
+ "C" or "en_US" locales) but do not work in East Asian locales. It is
+ better if programmers make their programs depend on GNU libiconv (except
+ on glibc systems), e.g. by using the AM_ICONV macro and documenting the
+ dependency in an INSTALL or DEPENDENCIES file. */
+
+ /* Try with the original names first.
+ This covers the case when fromcode or tocode is a lowercase encoding name
+ that is understood by the system's iconv_open but not listed in our
+ mappings table. */
+ {
+ iconv_t cd = iconv_open (tocode, fromcode);
+ if (cd != (iconv_t)(-1))
+ return cd;
+ }
+
+ /* Convert the encodings to upper case, because
+ 1. in the arguments of iconv_open() on AIX, HP-UX, and OSF/1 the case
+ matters,
+ 2. it makes searching in the table faster. */
+ {
+ const char *p = fromcode;
+ char *q = fromcode_upper;
+ while ((*q = c_toupper (*p)) != '\0')
+ {
+ p++;
+ q++;
+ if (q == &fromcode_upper[SIZEOF (fromcode_upper)])
+ {
+ errno = EINVAL;
+ return (iconv_t)(-1);
+ }
+ }
+ fromcode_upper_end = q;
+ }
+
+ {
+ const char *p = tocode;
+ char *q = tocode_upper;
+ while ((*q = c_toupper (*p)) != '\0')
+ {
+ p++;
+ q++;
+ if (q == &tocode_upper[SIZEOF (tocode_upper)])
+ {
+ errno = EINVAL;
+ return (iconv_t)(-1);
+ }
+ }
+ tocode_upper_end = q;
+ }
+
+#ifdef ICONV_FLAVOR
+ /* Apply the mappings. */
+ {
+ const struct mapping *m =
+ mapping_lookup (fromcode_upper, fromcode_upper_end - fromcode_upper);
+
+ fromcode = (m != NULL ? m->vendor_name : fromcode_upper);
+ }
+ {
+ const struct mapping *m =
+ mapping_lookup (tocode_upper, tocode_upper_end - tocode_upper);
+
+ tocode = (m != NULL ? m->vendor_name : tocode_upper);
+ }
+#else
+ fromcode = fromcode_upper;
+ tocode = tocode_upper;
+#endif
+
+ return iconv_open (tocode, fromcode);
+}
diff --git a/lib/iconveh.h b/lib/iconveh.h
new file mode 100644
index 0000000..06cda52
--- /dev/null
+++ b/lib/iconveh.h
@@ -0,0 +1,41 @@
+/* Character set conversion handler type.
+ Copyright (C) 2001-2007, 2009 Free Software Foundation, Inc.
+ Written by Bruno Haible.
+
+ This program is free software: you can redistribute it and/or modify
+ it under the terms of the GNU Lesser General Public License as published by
+ the Free Software Foundation; either version 3 of the License, or
+ (at your option) any later version.
+
+ This program is distributed in the hope that it will be useful,
+ but WITHOUT ANY WARRANTY; without even the implied warranty of
+ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ GNU Lesser General Public License for more details.
+
+ You should have received a copy of the GNU Lesser General Public License
+ along with this program. If not, see <http://www.gnu.org/licenses/>. */
+
+#ifndef _ICONVEH_H
+#define _ICONVEH_H
+
+
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+
+/* Handling of unconvertible characters. */
+enum iconv_ilseq_handler
+{
+ iconveh_error, /* return and set errno = EILSEQ */
+ iconveh_question_mark, /* use one '?' per unconvertible character */
+ iconveh_escape_sequence /* use escape sequence \uxxxx or \Uxxxxxxxx */
+};
+
+
+#ifdef __cplusplus
+}
+#endif
+
+
+#endif /* _ICONVEH_H */
diff --git a/lib/striconveh.c b/lib/striconveh.c
new file mode 100644
index 0000000..b39a01f
--- /dev/null
+++ b/lib/striconveh.c
@@ -0,0 +1,1251 @@
+/* Character set conversion with error handling.
+ Copyright (C) 2001-2008 Free Software Foundation, Inc.
+ Written by Bruno Haible and Simon Josefsson.
+
+ This program is free software: you can redistribute it and/or modify
+ it under the terms of the GNU Lesser General Public License as published by
+ the Free Software Foundation; either version 3 of the License, or
+ (at your option) any later version.
+
+ This program is distributed in the hope that it will be useful,
+ but WITHOUT ANY WARRANTY; without even the implied warranty of
+ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ GNU Lesser General Public License for more details.
+
+ You should have received a copy of the GNU Lesser General Public License
+ along with this program. If not, see <http://www.gnu.org/licenses/>. */
+
+#include <config.h>
+
+/* Specification. */
+#include "striconveh.h"
+
+#include <errno.h>
+#include <stdbool.h>
+#include <stdlib.h>
+#include <string.h>
+
+#if HAVE_ICONV
+# include <iconv.h>
+# include "unistr.h"
+#endif
+
+#include "c-strcase.h"
+#include "c-strcaseeq.h"
+
+#ifndef SIZE_MAX
+# define SIZE_MAX ((size_t) -1)
+#endif
+
+
+#if HAVE_ICONV
+
+/* The caller must provide CD, CD1, CD2, not just CD, because when a conversion
+ error occurs, we may have to determine the Unicode representation of the
+ inconvertible character. */
+
+/* iconv_carefully is like iconv, except that it stops as soon as it encounters
+ a conversion error, and it returns in *INCREMENTED a boolean telling whether
+ it has incremented the input pointers past the error location. */
+# if !defined _LIBICONV_VERSION && !defined __GLIBC__
+/* Irix iconv() inserts a NUL byte if it cannot convert.
+ NetBSD iconv() inserts a question mark if it cannot convert.
+ Only GNU libiconv and GNU libc are known to prefer to fail rather
+ than doing a lossy conversion. */
+static size_t
+iconv_carefully (iconv_t cd,
+ const char **inbuf, size_t *inbytesleft,
+ char **outbuf, size_t *outbytesleft,
+ bool *incremented)
+{
+ const char *inptr = *inbuf;
+ const char *inptr_end = inptr + *inbytesleft;
+ char *outptr = *outbuf;
+ size_t outsize = *outbytesleft;
+ const char *inptr_before;
+ size_t res;
+
+ do
+ {
+ size_t insize;
+
+ inptr_before = inptr;
+ res = (size_t)(-1);
+
+ for (insize = 1; inptr + insize <= inptr_end; insize++)
+ {
+ res = iconv (cd,
+ (ICONV_CONST char **) &inptr, &insize,
+ &outptr, &outsize);
+ if (!(res == (size_t)(-1) && errno == EINVAL))
+ break;
+ /* iconv can eat up a shift sequence but give EINVAL while attempting
+ to convert the first character. E.g. libiconv does this. */
+ if (inptr > inptr_before)
+ {
+ res = 0;
+ break;
+ }
+ }
+
+ if (res == 0)
+ {
+ *outbuf = outptr;
+ *outbytesleft = outsize;
+ }
+ }
+ while (res == 0 && inptr < inptr_end);
+
+ *inbuf = inptr;
+ *inbytesleft = inptr_end - inptr;
+ if (res != (size_t)(-1) && res > 0)
+ {
+ /* iconv() has already incremented INPTR. We cannot go back to a
+ previous INPTR, otherwise the state inside CD would become invalid,
+ if FROM_CODESET is a stateful encoding. So, tell the caller that
+ *INBUF has already been incremented. */
+ *incremented = (inptr > inptr_before);
+ errno = EILSEQ;
+ return (size_t)(-1);
+ }
+ else
+ {
+ *incremented = false;
+ return res;
+ }
+}
+# else
+# define iconv_carefully(cd, inbuf, inbytesleft, outbuf, outbytesleft,
incremented) \
+ (*(incremented) = false, \
+ iconv (cd, (ICONV_CONST char **) (inbuf), inbytesleft, outbuf,
outbytesleft))
+# endif
+
+/* iconv_carefully_1 is like iconv_carefully, except that it stops after
+ converting one character or one shift sequence. */
+static size_t
+iconv_carefully_1 (iconv_t cd,
+ const char **inbuf, size_t *inbytesleft,
+ char **outbuf, size_t *outbytesleft,
+ bool *incremented)
+{
+ const char *inptr_before = *inbuf;
+ const char *inptr = inptr_before;
+ const char *inptr_end = inptr_before + *inbytesleft;
+ char *outptr = *outbuf;
+ size_t outsize = *outbytesleft;
+ size_t res = (size_t)(-1);
+ size_t insize;
+
+ for (insize = 1; inptr_before + insize <= inptr_end; insize++)
+ {
+ inptr = inptr_before;
+ res = iconv (cd,
+ (ICONV_CONST char **) &inptr, &insize,
+ &outptr, &outsize);
+ if (!(res == (size_t)(-1) && errno == EINVAL))
+ break;
+ /* iconv can eat up a shift sequence but give EINVAL while attempting
+ to convert the first character. E.g. libiconv does this. */
+ if (inptr > inptr_before)
+ {
+ res = 0;
+ break;
+ }
+ }
+
+ *inbuf = inptr;
+ *inbytesleft = inptr_end - inptr;
+# if !defined _LIBICONV_VERSION && !defined __GLIBC__
+ /* Irix iconv() inserts a NUL byte if it cannot convert.
+ NetBSD iconv() inserts a question mark if it cannot convert.
+ Only GNU libiconv and GNU libc are known to prefer to fail rather
+ than doing a lossy conversion. */
+ if (res != (size_t)(-1) && res > 0)
+ {
+ /* iconv() has already incremented INPTR. We cannot go back to a
+ previous INPTR, otherwise the state inside CD would become invalid,
+ if FROM_CODESET is a stateful encoding. So, tell the caller that
+ *INBUF has already been incremented. */
+ *incremented = (inptr > inptr_before);
+ errno = EILSEQ;
+ return (size_t)(-1);
+ }
+# endif
+
+ if (res != (size_t)(-1))
+ {
+ *outbuf = outptr;
+ *outbytesleft = outsize;
+ }
+ *incremented = false;
+ return res;
+}
+
+/* utf8conv_carefully is like iconv, except that
+ - it converts from UTF-8 to UTF-8,
+ - it stops as soon as it encounters a conversion error, and it returns
+ in *INCREMENTED a boolean telling whether it has incremented the input
+ pointers past the error location,
+ - if one_character_only is true, it stops after converting one
+ character. */
+static size_t
+utf8conv_carefully (bool one_character_only,
+ const char **inbuf, size_t *inbytesleft,
+ char **outbuf, size_t *outbytesleft,
+ bool *incremented)
+{
+ const char *inptr = *inbuf;
+ size_t insize = *inbytesleft;
+ char *outptr = *outbuf;
+ size_t outsize = *outbytesleft;
+ size_t res;
+
+ res = 0;
+ do
+ {
+ ucs4_t uc;
+ int n;
+ int m;
+
+ n = u8_mbtoucr (&uc, (const uint8_t *) inptr, insize);
+ if (n < 0)
+ {
+ errno = (n == -2 ? EINVAL : EILSEQ);
+ n = u8_mbtouc (&uc, (const uint8_t *) inptr, insize);
+ inptr += n;
+ insize -= n;
+ res = (size_t)(-1);
+ *incremented = true;
+ break;
+ }
+ if (outsize == 0)
+ {
+ errno = E2BIG;
+ res = (size_t)(-1);
+ *incremented = false;
+ break;
+ }
+ m = u8_uctomb ((uint8_t *) outptr, uc, outsize);
+ if (m == -2)
+ {
+ errno = E2BIG;
+ res = (size_t)(-1);
+ *incremented = false;
+ break;
+ }
+ inptr += n;
+ insize -= n;
+ if (m == -1)
+ {
+ errno = EILSEQ;
+ res = (size_t)(-1);
+ *incremented = true;
+ break;
+ }
+ outptr += m;
+ outsize -= m;
+ }
+ while (!one_character_only && insize > 0);
+
+ *inbuf = inptr;
+ *inbytesleft = insize;
+ *outbuf = outptr;
+ *outbytesleft = outsize;
+ return res;
+}
+
+static int
+mem_cd_iconveh_internal (const char *src, size_t srclen,
+ iconv_t cd, iconv_t cd1, iconv_t cd2,
+ enum iconv_ilseq_handler handler,
+ size_t extra_alloc,
+ size_t *offsets,
+ char **resultp, size_t *lengthp)
+{
+ /* When a conversion error occurs, we cannot start using CD1 and CD2 at
+ this point: FROM_CODESET may be a stateful encoding like ISO-2022-KR.
+ Instead, we have to start afresh from the beginning of SRC. */
+ /* Use a temporary buffer, so that for small strings, a single malloc()
+ call will be sufficient. */
+# define tmpbufsize 4096
+ /* The alignment is needed when converting e.g. to glibc's WCHAR_T or
+ libiconv's UCS-4-INTERNAL encoding. */
+ union { unsigned int align; char buf[tmpbufsize]; } tmp;
+# define tmpbuf tmp.buf
+
+ char *initial_result;
+ char *result;
+ size_t allocated;
+ size_t length;
+ size_t last_length = (size_t)(-1); /* only needed if offsets != NULL */
+
+ if (*resultp != NULL && *lengthp >= sizeof (tmpbuf))
+ {
+ initial_result = *resultp;
+ allocated = *lengthp;
+ }
+ else
+ {
+ initial_result = tmpbuf;
+ allocated = sizeof (tmpbuf);
+ }
+ result = initial_result;
+
+ /* Test whether a direct conversion is possible at all. */
+ if (cd == (iconv_t)(-1))
+ goto indirectly;
+
+ if (offsets != NULL)
+ {
+ size_t i;
+
+ for (i = 0; i < srclen; i++)
+ offsets[i] = (size_t)(-1);
+
+ last_length = (size_t)(-1);
+ }
+ length = 0;
+
+ /* First, try a direct conversion, and see whether a conversion error
+ occurs at all. */
+ {
+ const char *inptr = src;
+ size_t insize = srclen;
+
+ /* Avoid glibc-2.1 bug and Solaris 2.7-2.9 bug. */
+# if defined _LIBICONV_VERSION \
+ || !((__GLIBC__ - 0 == 2 && __GLIBC_MINOR__ - 0 <= 1) || defined __sun)
+ /* Set to the initial state. */
+ iconv (cd, NULL, NULL, NULL, NULL);
+# endif
+
+ while (insize > 0)
+ {
+ char *outptr = result + length;
+ size_t outsize = allocated - extra_alloc - length;
+ bool incremented;
+ size_t res;
+ bool grow;
+
+ if (offsets != NULL)
+ {
+ if (length != last_length) /* ensure that offset[] be increasing */
+ {
+ offsets[inptr - src] = length;
+ last_length = length;
+ }
+ res = iconv_carefully_1 (cd,
+ &inptr, &insize,
+ &outptr, &outsize,
+ &incremented);
+ }
+ else
+ /* Use iconv_carefully instead of iconv here, because:
+ - If TO_CODESET is UTF-8, we can do the error handling in this
+ loop, no need for a second loop,
+ - With iconv() implementations other than GNU libiconv and GNU
+ libc, if we use iconv() in a big swoop, checking for an E2BIG
+ return, we lose the number of irreversible conversions. */
+ res = iconv_carefully (cd,
+ &inptr, &insize,
+ &outptr, &outsize,
+ &incremented);
+
+ length = outptr - result;
+ grow = (length + extra_alloc > allocated / 2);
+ if (res == (size_t)(-1))
+ {
+ if (errno == E2BIG)
+ grow = true;
+ else if (errno == EINVAL)
+ break;
+ else if (errno == EILSEQ && handler != iconveh_error)
+ {
+ if (cd2 == (iconv_t)(-1))
+ {
+ /* TO_CODESET is UTF-8. */
+ /* Error handling can produce up to 1 byte of output. */
+ if (length + 1 + extra_alloc > allocated)
+ {
+ char *memory;
+
+ allocated = 2 * allocated;
+ if (length + 1 + extra_alloc > allocated)
+ abort ();
+ if (result == initial_result)
+ memory = (char *) malloc (allocated);
+ else
+ memory = (char *) realloc (result, allocated);
+ if (memory == NULL)
+ {
+ if (result != initial_result)
+ free (result);
+ errno = ENOMEM;
+ return -1;
+ }
+ if (result == initial_result)
+ memcpy (memory, initial_result, length);
+ result = memory;
+ grow = false;
+ }
+ /* The input is invalid in FROM_CODESET. Eat up one byte
+ and emit a question mark. */
+ if (!incremented)
+ {
+ if (insize == 0)
+ abort ();
+ inptr++;
+ insize--;
+ }
+ result[length] = '?';
+ length++;
+ }
+ else
+ goto indirectly;
+ }
+ else
+ {
+ if (result != initial_result)
+ {
+ int saved_errno = errno;
+ free (result);
+ errno = saved_errno;
+ }
+ return -1;
+ }
+ }
+ if (insize == 0)
+ break;
+ if (grow)
+ {
+ char *memory;
+
+ allocated = 2 * allocated;
+ if (result == initial_result)
+ memory = (char *) malloc (allocated);
+ else
+ memory = (char *) realloc (result, allocated);
+ if (memory == NULL)
+ {
+ if (result != initial_result)
+ free (result);
+ errno = ENOMEM;
+ return -1;
+ }
+ if (result == initial_result)
+ memcpy (memory, initial_result, length);
+ result = memory;
+ }
+ }
+ }
+
+ /* Now get the conversion state back to the initial state.
+ But avoid glibc-2.1 bug and Solaris 2.7 bug. */
+#if defined _LIBICONV_VERSION \
+ || !((__GLIBC__ == 2 && __GLIBC_MINOR__ <= 1) || defined __sun)
+ for (;;)
+ {
+ char *outptr = result + length;
+ size_t outsize = allocated - extra_alloc - length;
+ size_t res;
+
+ res = iconv (cd, NULL, NULL, &outptr, &outsize);
+ length = outptr - result;
+ if (res == (size_t)(-1))
+ {
+ if (errno == E2BIG)
+ {
+ char *memory;
+
+ allocated = 2 * allocated;
+ if (result == initial_result)
+ memory = (char *) malloc (allocated);
+ else
+ memory = (char *) realloc (result, allocated);
+ if (memory == NULL)
+ {
+ if (result != initial_result)
+ free (result);
+ errno = ENOMEM;
+ return -1;
+ }
+ if (result == initial_result)
+ memcpy (memory, initial_result, length);
+ result = memory;
+ }
+ else
+ {
+ if (result != initial_result)
+ {
+ int saved_errno = errno;
+ free (result);
+ errno = saved_errno;
+ }
+ return -1;
+ }
+ }
+ else
+ break;
+ }
+#endif
+
+ /* The direct conversion succeeded. */
+ goto done;
+
+ indirectly:
+ /* The direct conversion failed.
+ Use a conversion through UTF-8. */
+ if (offsets != NULL)
+ {
+ size_t i;
+
+ for (i = 0; i < srclen; i++)
+ offsets[i] = (size_t)(-1);
+
+ last_length = (size_t)(-1);
+ }
+ length = 0;
+ {
+ const bool slowly = (offsets != NULL || handler == iconveh_error);
+# define utf8bufsize 4096 /* may also be smaller or larger than tmpbufsize */
+ char utf8buf[utf8bufsize + 1];
+ size_t utf8len = 0;
+ const char *in1ptr = src;
+ size_t in1size = srclen;
+ bool do_final_flush1 = true;
+ bool do_final_flush2 = true;
+
+ /* Avoid glibc-2.1 bug and Solaris 2.7-2.9 bug. */
+# if defined _LIBICONV_VERSION \
+ || !((__GLIBC__ - 0 == 2 && __GLIBC_MINOR__ - 0 <= 1) || defined __sun)
+ /* Set to the initial state. */
+ if (cd1 != (iconv_t)(-1))
+ iconv (cd1, NULL, NULL, NULL, NULL);
+ if (cd2 != (iconv_t)(-1))
+ iconv (cd2, NULL, NULL, NULL, NULL);
+# endif
+
+ while (in1size > 0 || do_final_flush1 || utf8len > 0 || do_final_flush2)
+ {
+ char *out1ptr = utf8buf + utf8len;
+ size_t out1size = utf8bufsize - utf8len;
+ bool incremented1;
+ size_t res1;
+ int errno1;
+
+ /* Conversion step 1: from FROM_CODESET to UTF-8. */
+ if (in1size > 0)
+ {
+ if (offsets != NULL
+ && length != last_length) /* ensure that offset[] be increasing
*/
+ {
+ offsets[in1ptr - src] = length;
+ last_length = length;
+ }
+ if (cd1 != (iconv_t)(-1))
+ {
+ if (slowly)
+ res1 = iconv_carefully_1 (cd1,
+ &in1ptr, &in1size,
+ &out1ptr, &out1size,
+ &incremented1);
+ else
+ res1 = iconv_carefully (cd1,
+ &in1ptr, &in1size,
+ &out1ptr, &out1size,
+ &incremented1);
+ }
+ else
+ {
+ /* FROM_CODESET is UTF-8. */
+ res1 = utf8conv_carefully (slowly,
+ &in1ptr, &in1size,
+ &out1ptr, &out1size,
+ &incremented1);
+ }
+ }
+ else if (do_final_flush1)
+ {
+ /* Now get the conversion state of CD1 back to the initial state.
+ But avoid glibc-2.1 bug and Solaris 2.7 bug. */
+# if defined _LIBICONV_VERSION \
+ || !((__GLIBC__ == 2 && __GLIBC_MINOR__ <= 1) || defined __sun)
+ if (cd1 != (iconv_t)(-1))
+ res1 = iconv (cd1, NULL, NULL, &out1ptr, &out1size);
+ else
+# endif
+ res1 = 0;
+ do_final_flush1 = false;
+ incremented1 = true;
+ }
+ else
+ {
+ res1 = 0;
+ incremented1 = true;
+ }
+ if (res1 == (size_t)(-1)
+ && !(errno == E2BIG || errno == EINVAL || errno == EILSEQ))
+ {
+ if (result != initial_result)
+ {
+ int saved_errno = errno;
+ free (result);
+ errno = saved_errno;
+ }
+ return -1;
+ }
+ if (res1 == (size_t)(-1)
+ && errno == EILSEQ && handler != iconveh_error)
+ {
+ /* The input is invalid in FROM_CODESET. Eat up one byte and
+ emit a question mark. Room for the question mark was allocated
+ at the end of utf8buf. */
+ if (!incremented1)
+ {
+ if (in1size == 0)
+ abort ();
+ in1ptr++;
+ in1size--;
+ }
+ utf8buf[utf8len++] = '?';
+ }
+ errno1 = errno;
+ utf8len = out1ptr - utf8buf;
+
+ if (offsets != NULL
+ || in1size == 0
+ || utf8len > utf8bufsize / 2
+ || (res1 == (size_t)(-1) && errno1 == E2BIG))
+ {
+ /* Conversion step 2: from UTF-8 to TO_CODESET. */
+ const char *in2ptr = utf8buf;
+ size_t in2size = utf8len;
+
+ while (in2size > 0
+ || (in1size == 0 && !do_final_flush1 && do_final_flush2))
+ {
+ char *out2ptr = result + length;
+ size_t out2size = allocated - extra_alloc - length;
+ bool incremented2;
+ size_t res2;
+ bool grow;
+
+ if (in2size > 0)
+ {
+ if (cd2 != (iconv_t)(-1))
+ res2 = iconv_carefully (cd2,
+ &in2ptr, &in2size,
+ &out2ptr, &out2size,
+ &incremented2);
+ else
+ /* TO_CODESET is UTF-8. */
+ res2 = utf8conv_carefully (false,
+ &in2ptr, &in2size,
+ &out2ptr, &out2size,
+ &incremented2);
+ }
+ else /* in1size == 0 && !do_final_flush1
+ && in2size == 0 && do_final_flush2 */
+ {
+ /* Now get the conversion state of CD1 back to the initial
+ state. But avoid glibc-2.1 bug and Solaris 2.7 bug. */
+# if defined _LIBICONV_VERSION \
+ || !((__GLIBC__ == 2 && __GLIBC_MINOR__ <= 1) || defined __sun)
+ if (cd2 != (iconv_t)(-1))
+ res2 = iconv (cd2, NULL, NULL, &out2ptr, &out2size);
+ else
+# endif
+ res2 = 0;
+ do_final_flush2 = false;
+ incremented2 = true;
+ }
+
+ length = out2ptr - result;
+ grow = (length + extra_alloc > allocated / 2);
+ if (res2 == (size_t)(-1))
+ {
+ if (errno == E2BIG)
+ grow = true;
+ else if (errno == EINVAL)
+ break;
+ else if (errno == EILSEQ && handler != iconveh_error)
+ {
+ /* Error handling can produce up to 10 bytes of ASCII
+ output. But TO_CODESET may be UCS-2, UTF-16 or
+ UCS-4, so use CD2 here as well. */
+ char scratchbuf[10];
+ size_t scratchlen;
+ ucs4_t uc;
+ const char *inptr;
+ size_t insize;
+ size_t res;
+
+ if (incremented2)
+ {
+ if (u8_prev (&uc, (const uint8_t *) in2ptr,
+ (const uint8_t *) utf8buf)
+ == NULL)
+ abort ();
+ }
+ else
+ {
+ int n;
+ if (in2size == 0)
+ abort ();
+ n = u8_mbtouc_unsafe (&uc, (const uint8_t *) in2ptr,
+ in2size);
+ in2ptr += n;
+ in2size -= n;
+ }
+
+ if (handler == iconveh_escape_sequence)
+ {
+ static char hex[16] = "0123456789ABCDEF";
+ scratchlen = 0;
+ scratchbuf[scratchlen++] = '\\';
+ if (uc < 0x10000)
+ scratchbuf[scratchlen++] = 'u';
+ else
+ {
+ scratchbuf[scratchlen++] = 'U';
+ scratchbuf[scratchlen++] = hex[(uc>>28) & 15];
+ scratchbuf[scratchlen++] = hex[(uc>>24) & 15];
+ scratchbuf[scratchlen++] = hex[(uc>>20) & 15];
+ scratchbuf[scratchlen++] = hex[(uc>>16) & 15];
+ }
+ scratchbuf[scratchlen++] = hex[(uc>>12) & 15];
+ scratchbuf[scratchlen++] = hex[(uc>>8) & 15];
+ scratchbuf[scratchlen++] = hex[(uc>>4) & 15];
+ scratchbuf[scratchlen++] = hex[uc & 15];
+ }
+ else
+ {
+ scratchbuf[0] = '?';
+ scratchlen = 1;
+ }
+
+ inptr = scratchbuf;
+ insize = scratchlen;
+ if (cd2 != (iconv_t)(-1))
+ res = iconv (cd2,
+ (ICONV_CONST char **) &inptr, &insize,
+ &out2ptr, &out2size);
+ else
+ {
+ /* TO_CODESET is UTF-8. */
+ if (out2size >= insize)
+ {
+ memcpy (out2ptr, inptr, insize);
+ out2ptr += insize;
+ out2size -= insize;
+ inptr += insize;
+ insize = 0;
+ res = 0;
+ }
+ else
+ {
+ errno = E2BIG;
+ res = (size_t)(-1);
+ }
+ }
+ length = out2ptr - result;
+ if (res == (size_t)(-1) && errno == E2BIG)
+ {
+ char *memory;
+
+ allocated = 2 * allocated;
+ if (length + 1 + extra_alloc > allocated)
+ abort ();
+ if (result == initial_result)
+ memory = (char *) malloc (allocated);
+ else
+ memory = (char *) realloc (result, allocated);
+ if (memory == NULL)
+ {
+ if (result != initial_result)
+ free (result);
+ errno = ENOMEM;
+ return -1;
+ }
+ if (result == initial_result)
+ memcpy (memory, initial_result, length);
+ result = memory;
+ grow = false;
+
+ out2ptr = result + length;
+ out2size = allocated - extra_alloc - length;
+ if (cd2 != (iconv_t)(-1))
+ res = iconv (cd2,
+ (ICONV_CONST char **) &inptr,
+ &insize,
+ &out2ptr, &out2size);
+ else
+ {
+ /* TO_CODESET is UTF-8. */
+ if (!(out2size >= insize))
+ abort ();
+ memcpy (out2ptr, inptr, insize);
+ out2ptr += insize;
+ out2size -= insize;
+ inptr += insize;
+ insize = 0;
+ res = 0;
+ }
+ length = out2ptr - result;
+ }
+# if !defined _LIBICONV_VERSION && !defined __GLIBC__
+ /* Irix iconv() inserts a NUL byte if it cannot convert.
+ NetBSD iconv() inserts a question mark if it cannot
+ convert.
+ Only GNU libiconv and GNU libc are known to prefer
+ to fail rather than doing a lossy conversion. */
+ if (res != (size_t)(-1) && res > 0)
+ {
+ errno = EILSEQ;
+ res = (size_t)(-1);
+ }
+# endif
+ if (res == (size_t)(-1))
+ {
+ /* Failure converting the ASCII replacement. */
+ if (result != initial_result)
+ {
+ int saved_errno = errno;
+ free (result);
+ errno = saved_errno;
+ }
+ return -1;
+ }
+ }
+ else
+ {
+ if (result != initial_result)
+ {
+ int saved_errno = errno;
+ free (result);
+ errno = saved_errno;
+ }
+ return -1;
+ }
+ }
+ if (!(in2size > 0
+ || (in1size == 0 && !do_final_flush1 && do_final_flush2)))
+ break;
+ if (grow)
+ {
+ char *memory;
+
+ allocated = 2 * allocated;
+ if (result == initial_result)
+ memory = (char *) malloc (allocated);
+ else
+ memory = (char *) realloc (result, allocated);
+ if (memory == NULL)
+ {
+ if (result != initial_result)
+ free (result);
+ errno = ENOMEM;
+ return -1;
+ }
+ if (result == initial_result)
+ memcpy (memory, initial_result, length);
+ result = memory;
+ }
+ }
+
+ /* Move the remaining bytes to the beginning of utf8buf. */
+ if (in2size > 0)
+ memmove (utf8buf, in2ptr, in2size);
+ utf8len = in2size;
+ }
+
+ if (res1 == (size_t)(-1))
+ {
+ if (errno1 == EINVAL)
+ in1size = 0;
+ else if (errno1 == EILSEQ)
+ {
+ if (result != initial_result)
+ free (result);
+ errno = errno1;
+ return -1;
+ }
+ }
+ }
+# undef utf8bufsize
+ }
+
+ done:
+ /* Now the final memory allocation. */
+ if (result == tmpbuf)
+ {
+ size_t memsize = length + extra_alloc;
+ char *memory;
+
+ memory = (char *) malloc (memsize > 0 ? memsize : 1);
+ if (memory != NULL)
+ {
+ memcpy (memory, tmpbuf, length);
+ result = memory;
+ }
+ else
+ {
+ errno = ENOMEM;
+ return -1;
+ }
+ }
+ else if (result != *resultp && length + extra_alloc < allocated)
+ {
+ /* Shrink the allocated memory if possible. */
+ size_t memsize = length + extra_alloc;
+ char *memory;
+
+ memory = (char *) realloc (result, memsize > 0 ? memsize : 1);
+ if (memory != NULL)
+ result = memory;
+ }
+ *resultp = result;
+ *lengthp = length;
+ return 0;
+# undef tmpbuf
+# undef tmpbufsize
+}
+
+int
+mem_cd_iconveh (const char *src, size_t srclen,
+ iconv_t cd, iconv_t cd1, iconv_t cd2,
+ enum iconv_ilseq_handler handler,
+ size_t *offsets,
+ char **resultp, size_t *lengthp)
+{
+ return mem_cd_iconveh_internal (src, srclen, cd, cd1, cd2, handler, 0,
+ offsets, resultp, lengthp);
+}
+
+char *
+str_cd_iconveh (const char *src,
+ iconv_t cd, iconv_t cd1, iconv_t cd2,
+ enum iconv_ilseq_handler handler)
+{
+ /* For most encodings, a trailing NUL byte in the input will be converted
+ to a trailing NUL byte in the output. But not for UTF-7. So that this
+ function is usable for UTF-7, we have to exclude the NUL byte from the
+ conversion and add it by hand afterwards. */
+ char *result = NULL;
+ size_t length = 0;
+ int retval = mem_cd_iconveh_internal (src, strlen (src),
+ cd, cd1, cd2, handler, 1, NULL,
+ &result, &length);
+
+ if (retval < 0)
+ {
+ if (result != NULL)
+ {
+ int saved_errno = errno;
+ free (result);
+ errno = saved_errno;
+ }
+ return NULL;
+ }
+
+ /* Add the terminating NUL byte. */
+ result[length] = '\0';
+
+ return result;
+}
+
+#endif
+
+int
+mem_iconveh (const char *src, size_t srclen,
+ const char *from_codeset, const char *to_codeset,
+ enum iconv_ilseq_handler handler,
+ size_t *offsets,
+ char **resultp, size_t *lengthp)
+{
+ if (srclen == 0)
+ {
+ /* Nothing to convert. */
+ *lengthp = 0;
+ return 0;
+ }
+ else if (offsets == NULL && c_strcasecmp (from_codeset, to_codeset) == 0)
+ {
+ char *result;
+
+ if (*resultp != NULL && *lengthp >= srclen)
+ result = *resultp;
+ else
+ {
+ result = (char *) malloc (srclen);
+ if (result == NULL)
+ {
+ errno = ENOMEM;
+ return -1;
+ }
+ }
+ memcpy (result, src, srclen);
+ *resultp = result;
+ *lengthp = srclen;
+ return 0;
+ }
+ else
+ {
+#if HAVE_ICONV
+ iconv_t cd;
+ iconv_t cd1;
+ iconv_t cd2;
+ char *result;
+ size_t length;
+ int retval;
+
+ /* Avoid glibc-2.1 bug with EUC-KR. */
+# if (__GLIBC__ - 0 == 2 && __GLIBC_MINOR__ - 0 <= 1) && !defined
_LIBICONV_VERSION
+ if (c_strcasecmp (from_codeset, "EUC-KR") == 0
+ || c_strcasecmp (to_codeset, "EUC-KR") == 0)
+ {
+ errno = EINVAL;
+ return -1;
+ }
+# endif
+
+ cd = iconv_open (to_codeset, from_codeset);
+
+ if (STRCASEEQ (from_codeset, "UTF-8", 'U','T','F','-','8',0,0,0,0))
+ cd1 = (iconv_t)(-1);
+ else
+ {
+ cd1 = iconv_open ("UTF-8", from_codeset);
+ if (cd1 == (iconv_t)(-1))
+ {
+ int saved_errno = errno;
+ if (cd != (iconv_t)(-1))
+ iconv_close (cd);
+ errno = saved_errno;
+ return -1;
+ }
+ }
+
+ if (STRCASEEQ (to_codeset, "UTF-8", 'U','T','F','-','8',0,0,0,0)
+# if (__GLIBC__ == 2 && __GLIBC_MINOR__ >= 2) || __GLIBC__ > 2 ||
_LIBICONV_VERSION >= 0x0105
+ || c_strcasecmp (to_codeset, "UTF-8//TRANSLIT") == 0
+# endif
+ )
+ cd2 = (iconv_t)(-1);
+ else
+ {
+ cd2 = iconv_open (to_codeset, "UTF-8");
+ if (cd2 == (iconv_t)(-1))
+ {
+ int saved_errno = errno;
+ if (cd1 != (iconv_t)(-1))
+ iconv_close (cd1);
+ if (cd != (iconv_t)(-1))
+ iconv_close (cd);
+ errno = saved_errno;
+ return -1;
+ }
+ }
+
+ result = *resultp;
+ length = *lengthp;
+ retval = mem_cd_iconveh (src, srclen, cd, cd1, cd2, handler, offsets,
+ &result, &length);
+
+ if (retval < 0)
+ {
+ /* Close cd, cd1, cd2, but preserve the errno from str_cd_iconv. */
+ int saved_errno = errno;
+ if (cd2 != (iconv_t)(-1))
+ iconv_close (cd2);
+ if (cd1 != (iconv_t)(-1))
+ iconv_close (cd1);
+ if (cd != (iconv_t)(-1))
+ iconv_close (cd);
+ errno = saved_errno;
+ }
+ else
+ {
+ if (cd2 != (iconv_t)(-1) && iconv_close (cd2) < 0)
+ {
+ /* Return -1, but free the allocated memory, and while doing
+ that, preserve the errno from iconv_close. */
+ int saved_errno = errno;
+ if (cd1 != (iconv_t)(-1))
+ iconv_close (cd1);
+ if (cd != (iconv_t)(-1))
+ iconv_close (cd);
+ if (result != *resultp && result != NULL)
+ free (result);
+ errno = saved_errno;
+ return -1;
+ }
+ if (cd1 != (iconv_t)(-1) && iconv_close (cd1) < 0)
+ {
+ /* Return -1, but free the allocated memory, and while doing
+ that, preserve the errno from iconv_close. */
+ int saved_errno = errno;
+ if (cd != (iconv_t)(-1))
+ iconv_close (cd);
+ if (result != *resultp && result != NULL)
+ free (result);
+ errno = saved_errno;
+ return -1;
+ }
+ if (cd != (iconv_t)(-1) && iconv_close (cd) < 0)
+ {
+ /* Return -1, but free the allocated memory, and while doing
+ that, preserve the errno from iconv_close. */
+ int saved_errno = errno;
+ if (result != *resultp && result != NULL)
+ free (result);
+ errno = saved_errno;
+ return -1;
+ }
+ *resultp = result;
+ *lengthp = length;
+ }
+ return retval;
+#else
+ /* This is a different error code than if iconv_open existed but didn't
+ support from_codeset and to_codeset, so that the caller can emit
+ an error message such as
+ "iconv() is not supported. Installing GNU libiconv and
+ then reinstalling this package would fix this." */
+ errno = ENOSYS;
+ return -1;
+#endif
+ }
+}
+
+char *
+str_iconveh (const char *src,
+ const char *from_codeset, const char *to_codeset,
+ enum iconv_ilseq_handler handler)
+{
+ if (*src == '\0' || c_strcasecmp (from_codeset, to_codeset) == 0)
+ {
+ char *result = strdup (src);
+
+ if (result == NULL)
+ errno = ENOMEM;
+ return result;
+ }
+ else
+ {
+#if HAVE_ICONV
+ iconv_t cd;
+ iconv_t cd1;
+ iconv_t cd2;
+ char *result;
+
+ /* Avoid glibc-2.1 bug with EUC-KR. */
+# if (__GLIBC__ - 0 == 2 && __GLIBC_MINOR__ - 0 <= 1) && !defined
_LIBICONV_VERSION
+ if (c_strcasecmp (from_codeset, "EUC-KR") == 0
+ || c_strcasecmp (to_codeset, "EUC-KR") == 0)
+ {
+ errno = EINVAL;
+ return NULL;
+ }
+# endif
+
+ cd = iconv_open (to_codeset, from_codeset);
+
+ if (STRCASEEQ (from_codeset, "UTF-8", 'U','T','F','-','8',0,0,0,0))
+ cd1 = (iconv_t)(-1);
+ else
+ {
+ cd1 = iconv_open ("UTF-8", from_codeset);
+ if (cd1 == (iconv_t)(-1))
+ {
+ int saved_errno = errno;
+ if (cd != (iconv_t)(-1))
+ iconv_close (cd);
+ errno = saved_errno;
+ return NULL;
+ }
+ }
+
+ if (STRCASEEQ (to_codeset, "UTF-8", 'U','T','F','-','8',0,0,0,0)
+# if (__GLIBC__ == 2 && __GLIBC_MINOR__ >= 2) || __GLIBC__ > 2 ||
_LIBICONV_VERSION >= 0x0105
+ || c_strcasecmp (to_codeset, "UTF-8//TRANSLIT") == 0
+# endif
+ )
+ cd2 = (iconv_t)(-1);
+ else
+ {
+ cd2 = iconv_open (to_codeset, "UTF-8");
+ if (cd2 == (iconv_t)(-1))
+ {
+ int saved_errno = errno;
+ if (cd1 != (iconv_t)(-1))
+ iconv_close (cd1);
+ if (cd != (iconv_t)(-1))
+ iconv_close (cd);
+ errno = saved_errno;
+ return NULL;
+ }
+ }
+
+ result = str_cd_iconveh (src, cd, cd1, cd2, handler);
+
+ if (result == NULL)
+ {
+ /* Close cd, cd1, cd2, but preserve the errno from str_cd_iconv. */
+ int saved_errno = errno;
+ if (cd2 != (iconv_t)(-1))
+ iconv_close (cd2);
+ if (cd1 != (iconv_t)(-1))
+ iconv_close (cd1);
+ if (cd != (iconv_t)(-1))
+ iconv_close (cd);
+ errno = saved_errno;
+ }
+ else
+ {
+ if (cd2 != (iconv_t)(-1) && iconv_close (cd2) < 0)
+ {
+ /* Return NULL, but free the allocated memory, and while doing
+ that, preserve the errno from iconv_close. */
+ int saved_errno = errno;
+ if (cd1 != (iconv_t)(-1))
+ iconv_close (cd1);
+ if (cd != (iconv_t)(-1))
+ iconv_close (cd);
+ free (result);
+ errno = saved_errno;
+ return NULL;
+ }
+ if (cd1 != (iconv_t)(-1) && iconv_close (cd1) < 0)
+ {
+ /* Return NULL, but free the allocated memory, and while doing
+ that, preserve the errno from iconv_close. */
+ int saved_errno = errno;
+ if (cd != (iconv_t)(-1))
+ iconv_close (cd);
+ free (result);
+ errno = saved_errno;
+ return NULL;
+ }
+ if (cd != (iconv_t)(-1) && iconv_close (cd) < 0)
+ {
+ /* Return NULL, but free the allocated memory, and while doing
+ that, preserve the errno from iconv_close. */
+ int saved_errno = errno;
+ free (result);
+ errno = saved_errno;
+ return NULL;
+ }
+ }
+ return result;
+#else
+ /* This is a different error code than if iconv_open existed but didn't
+ support from_codeset and to_codeset, so that the caller can emit
+ an error message such as
+ "iconv() is not supported. Installing GNU libiconv and
+ then reinstalling this package would fix this." */
+ errno = ENOSYS;
+ return NULL;
+#endif
+ }
+}
diff --git a/lib/striconveh.h b/lib/striconveh.h
new file mode 100644
index 0000000..98b4d0c
--- /dev/null
+++ b/lib/striconveh.h
@@ -0,0 +1,120 @@
+/* Character set conversion with error handling.
+ Copyright (C) 2001-2007, 2009 Free Software Foundation, Inc.
+ Written by Bruno Haible and Simon Josefsson.
+
+ This program is free software: you can redistribute it and/or modify
+ it under the terms of the GNU Lesser General Public License as published by
+ the Free Software Foundation; either version 3 of the License, or
+ (at your option) any later version.
+
+ This program is distributed in the hope that it will be useful,
+ but WITHOUT ANY WARRANTY; without even the implied warranty of
+ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ GNU Lesser General Public License for more details.
+
+ You should have received a copy of the GNU Lesser General Public License
+ along with this program. If not, see <http://www.gnu.org/licenses/>. */
+
+#ifndef _STRICONVEH_H
+#define _STRICONVEH_H
+
+#include <stddef.h>
+#if HAVE_ICONV
+#include <iconv.h>
+#endif
+
+#include "iconveh.h"
+
+
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+
+#if HAVE_ICONV
+
+/* Convert an entire string from one encoding to another, using iconv.
+ The original string is at [SRC,...,SRC+SRCLEN-1].
+ CD is the conversion descriptor from FROMCODE to TOCODE, or (iconv_t)(-1) if
+ the system does not support a direct conversion from FROMCODE to TOCODE.
+ CD1 is the conversion descriptor from FROM_CODESET to UTF-8 (or
+ (iconv_t)(-1) if FROM_CODESET is UTF-8).
+ CD2 is the conversion descriptor from UTF-8 to TO_CODESET (or (iconv_t)(-1)
+ if TO_CODESET is UTF-8).
+ If OFFSETS is not NULL, it should point to an array of SRCLEN integers; this
+ array is filled with offsets into the result, i.e. the character starting
+ at SRC[i] corresponds to the character starting at (*RESULTP)[OFFSETS[i]],
+ and other offsets are set to (size_t)(-1).
+ *RESULTP and *LENGTH should initially be a scratch buffer and its size,
+ or *RESULTP can initially be NULL.
+ May erase the contents of the memory at *RESULTP.
+ Return value: 0 if successful, otherwise -1 and errno set.
+ If successful: The resulting string is stored in *RESULTP and its length
+ in *LENGTHP. *RESULTP is set to a freshly allocated memory block, or is
+ unchanged if no dynamic memory allocation was necessary. */
+extern int
+ mem_cd_iconveh (const char *src, size_t srclen,
+ iconv_t cd, iconv_t cd1, iconv_t cd2,
+ enum iconv_ilseq_handler handler,
+ size_t *offsets,
+ char **resultp, size_t *lengthp);
+
+/* Convert an entire string from one encoding to another, using iconv.
+ The original string is the NUL-terminated string starting at SRC.
+ CD is the conversion descriptor from FROMCODE to TOCODE, or (iconv_t)(-1) if
+ the system does not support a direct conversion from FROMCODE to TOCODE.
+ Both the "from" and the "to" encoding must use a single NUL byte at the end
+ of the string (i.e. not UCS-2, UCS-4, UTF-16, UTF-32).
+ CD1 is the conversion descriptor from FROM_CODESET to UTF-8 (or
+ (iconv_t)(-1) if FROM_CODESET is UTF-8).
+ CD2 is the conversion descriptor from UTF-8 to TO_CODESET (or (iconv_t)(-1)
+ if TO_CODESET is UTF-8).
+ Allocate a malloced memory block for the result.
+ Return value: the freshly allocated resulting NUL-terminated string if
+ successful, otherwise NULL and errno set. */
+extern char *
+ str_cd_iconveh (const char *src,
+ iconv_t cd, iconv_t cd1, iconv_t cd2,
+ enum iconv_ilseq_handler handler);
+
+#endif
+
+/* Convert an entire string from one encoding to another, using iconv.
+ The original string is at [SRC,...,SRC+SRCLEN-1].
+ If OFFSETS is not NULL, it should point to an array of SRCLEN integers; this
+ array is filled with offsets into the result, i.e. the character starting
+ at SRC[i] corresponds to the character starting at (*RESULTP)[OFFSETS[i]],
+ and other offsets are set to (size_t)(-1).
+ *RESULTP and *LENGTH should initially be a scratch buffer and its size,
+ or *RESULTP can initially be NULL.
+ May erase the contents of the memory at *RESULTP.
+ Return value: 0 if successful, otherwise -1 and errno set.
+ If successful: The resulting string is stored in *RESULTP and its length
+ in *LENGTHP. *RESULTP is set to a freshly allocated memory block, or is
+ unchanged if no dynamic memory allocation was necessary. */
+extern int
+ mem_iconveh (const char *src, size_t srclen,
+ const char *from_codeset, const char *to_codeset,
+ enum iconv_ilseq_handler handler,
+ size_t *offsets,
+ char **resultp, size_t *lengthp);
+
+/* Convert an entire string from one encoding to another, using iconv.
+ The original string is the NUL-terminated string starting at SRC.
+ Both the "from" and the "to" encoding must use a single NUL byte at the
+ end of the string (i.e. not UCS-2, UCS-4, UTF-16, UTF-32).
+ Allocate a malloced memory block for the result.
+ Return value: the freshly allocated resulting NUL-terminated string if
+ successful, otherwise NULL and errno set. */
+extern char *
+ str_iconveh (const char *src,
+ const char *from_codeset, const char *to_codeset,
+ enum iconv_ilseq_handler handler);
+
+
+#ifdef __cplusplus
+}
+#endif
+
+
+#endif /* _STRICONVEH_H */
diff --git a/lib/string.in.h b/lib/string.in.h
new file mode 100644
index 0000000..ca029d7
--- /dev/null
+++ b/lib/string.in.h
@@ -0,0 +1,605 @@
+/* A GNU-like <string.h>.
+
+ Copyright (C) 1995-1996, 2001-2008 Free Software Foundation, Inc.
+
+ This program is free software; you can redistribute it and/or modify
+ it under the terms of the GNU Lesser General Public License as published by
+ the Free Software Foundation; either version 2, or (at your option)
+ any later version.
+
+ This program is distributed in the hope that it will be useful,
+ but WITHOUT ANY WARRANTY; without even the implied warranty of
+ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ GNU Lesser General Public License for more details.
+
+ You should have received a copy of the GNU Lesser General Public License
+ along with this program; if not, write to the Free Software Foundation,
+ Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */
+
+#ifndef _GL_STRING_H
+
+#if __GNUC__ >= 3
address@hidden@
+#endif
+
+/* The include_next requires a split double-inclusion guard. */
address@hidden@ @NEXT_STRING_H@
+
+#ifndef _GL_STRING_H
+#define _GL_STRING_H
+
+
+#ifndef __attribute__
+/* This feature is available in gcc versions 2.5 and later. */
+# if __GNUC__ < 2 || (__GNUC__ == 2 && __GNUC_MINOR__ < 5)
+# define __attribute__(Spec) /* empty */
+# endif
+/* The attribute __pure__ was added in gcc 2.96. */
+# if __GNUC__ < 2 || (__GNUC__ == 2 && __GNUC_MINOR__ < 96)
+# define __pure__ /* empty */
+# endif
+#endif
+
+
+/* The definition of GL_LINK_WARNING is copied here. */
+
+
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+
+/* Return the first occurrence of NEEDLE in HAYSTACK. */
+#if @GNULIB_MEMMEM@
+# if @REPLACE_MEMMEM@
+# define memmem rpl_memmem
+# endif
+# if ! @HAVE_DECL_MEMMEM@ || @REPLACE_MEMMEM@
+extern void *memmem (void const *__haystack, size_t __haystack_len,
+ void const *__needle, size_t __needle_len)
+ __attribute__ ((__pure__));
+# endif
+#elif defined GNULIB_POSIXCHECK
+# undef memmem
+# define memmem(a,al,b,bl) \
+ (GL_LINK_WARNING ("memmem is unportable and often quadratic - " \
+ "use gnulib module memmem-simple for portability, " \
+ "and module memmem for speed" ), \
+ memmem (a, al, b, bl))
+#endif
+
+/* Copy N bytes of SRC to DEST, return pointer to bytes after the
+ last written byte. */
+#if @GNULIB_MEMPCPY@
+# if ! @HAVE_MEMPCPY@
+extern void *mempcpy (void *restrict __dest, void const *restrict __src,
+ size_t __n);
+# endif
+#elif defined GNULIB_POSIXCHECK
+# undef mempcpy
+# define mempcpy(a,b,n) \
+ (GL_LINK_WARNING ("mempcpy is unportable - " \
+ "use gnulib module mempcpy for portability"), \
+ mempcpy (a, b, n))
+#endif
+
+/* Search backwards through a block for a byte (specified as an int). */
+#if @GNULIB_MEMRCHR@
+# if ! @HAVE_DECL_MEMRCHR@
+extern void *memrchr (void const *, int, size_t)
+ __attribute__ ((__pure__));
+# endif
+#elif defined GNULIB_POSIXCHECK
+# undef memrchr
+# define memrchr(a,b,c) \
+ (GL_LINK_WARNING ("memrchr is unportable - " \
+ "use gnulib module memrchr for portability"), \
+ memrchr (a, b, c))
+#endif
+
+/* Find the first occurrence of C in S. More efficient than
+ memchr(S,C,N), at the expense of undefined behavior if C does not
+ occur within N bytes. */
+#if @GNULIB_RAWMEMCHR@
+# if ! @HAVE_RAWMEMCHR@
+extern void *rawmemchr (void const *__s, int __c_in)
+ __attribute__ ((__pure__));
+# endif
+#elif defined GNULIB_POSIXCHECK
+# undef rawmemchr
+# define rawmemchr(a,b) \
+ (GL_LINK_WARNING ("rawmemchr is unportable - " \
+ "use gnulib module rawmemchr for portability"), \
+ rawmemchr (a, b))
+#endif
+
+/* Copy SRC to DST, returning the address of the terminating '\0' in DST. */
+#if @GNULIB_STPCPY@
+# if ! @HAVE_STPCPY@
+extern char *stpcpy (char *restrict __dst, char const *restrict __src);
+# endif
+#elif defined GNULIB_POSIXCHECK
+# undef stpcpy
+# define stpcpy(a,b) \
+ (GL_LINK_WARNING ("stpcpy is unportable - " \
+ "use gnulib module stpcpy for portability"), \
+ stpcpy (a, b))
+#endif
+
+/* Copy no more than N bytes of SRC to DST, returning a pointer past the
+ last non-NUL byte written into DST. */
+#if @GNULIB_STPNCPY@
+# if ! @HAVE_STPNCPY@
+# define stpncpy gnu_stpncpy
+extern char *stpncpy (char *restrict __dst, char const *restrict __src,
+ size_t __n);
+# endif
+#elif defined GNULIB_POSIXCHECK
+# undef stpncpy
+# define stpncpy(a,b,n) \
+ (GL_LINK_WARNING ("stpncpy is unportable - " \
+ "use gnulib module stpncpy for portability"), \
+ stpncpy (a, b, n))
+#endif
+
+#if defined GNULIB_POSIXCHECK
+/* strchr() does not work with multibyte strings if the locale encoding is
+ GB18030 and the character to be searched is a digit. */
+# undef strchr
+# define strchr(s,c) \
+ (GL_LINK_WARNING ("strchr cannot work correctly on character strings " \
+ "in some multibyte locales - " \
+ "use mbschr if you care about internationalization"), \
+ strchr (s, c))
+#endif
+
+/* Find the first occurrence of C in S or the final NUL byte. */
+#if @GNULIB_STRCHRNUL@
+# if ! @HAVE_STRCHRNUL@
+extern char *strchrnul (char const *__s, int __c_in)
+ __attribute__ ((__pure__));
+# endif
+#elif defined GNULIB_POSIXCHECK
+# undef strchrnul
+# define strchrnul(a,b) \
+ (GL_LINK_WARNING ("strchrnul is unportable - " \
+ "use gnulib module strchrnul for portability"), \
+ strchrnul (a, b))
+#endif
+
+/* Duplicate S, returning an identical malloc'd string. */
+#if @GNULIB_STRDUP@
+# if @REPLACE_STRDUP@
+# undef strdup
+# define strdup rpl_strdup
+# endif
+# if !(@HAVE_DECL_STRDUP@ || defined strdup) || @REPLACE_STRDUP@
+extern char *strdup (char const *__s);
+# endif
+#elif defined GNULIB_POSIXCHECK
+# undef strdup
+# define strdup(a) \
+ (GL_LINK_WARNING ("strdup is unportable - " \
+ "use gnulib module strdup for portability"), \
+ strdup (a))
+#endif
+
+/* Return a newly allocated copy of at most N bytes of STRING. */
+#if @GNULIB_STRNDUP@
+# if ! @HAVE_STRNDUP@
+# undef strndup
+# define strndup rpl_strndup
+# endif
+# if ! @HAVE_STRNDUP@ || ! @HAVE_DECL_STRNDUP@
+extern char *strndup (char const *__string, size_t __n);
+# endif
+#elif defined GNULIB_POSIXCHECK
+# undef strndup
+# define strndup(a,n) \
+ (GL_LINK_WARNING ("strndup is unportable - " \
+ "use gnulib module strndup for portability"), \
+ strndup (a, n))
+#endif
+
+/* Find the length (number of bytes) of STRING, but scan at most
+ MAXLEN bytes. If no '\0' terminator is found in that many bytes,
+ return MAXLEN. */
+#if @GNULIB_STRNLEN@
+# if ! @HAVE_DECL_STRNLEN@
+extern size_t strnlen (char const *__string, size_t __maxlen)
+ __attribute__ ((__pure__));
+# endif
+#elif defined GNULIB_POSIXCHECK
+# undef strnlen
+# define strnlen(a,n) \
+ (GL_LINK_WARNING ("strnlen is unportable - " \
+ "use gnulib module strnlen for portability"), \
+ strnlen (a, n))
+#endif
+
+#if defined GNULIB_POSIXCHECK
+/* strcspn() assumes the second argument is a list of single-byte characters.
+ Even in this simple case, it does not work with multibyte strings if the
+ locale encoding is GB18030 and one of the characters to be searched is a
+ digit. */
+# undef strcspn
+# define strcspn(s,a) \
+ (GL_LINK_WARNING ("strcspn cannot work correctly on character strings " \
+ "in multibyte locales - " \
+ "use mbscspn if you care about internationalization"), \
+ strcspn (s, a))
+#endif
+
+/* Find the first occurrence in S of any character in ACCEPT. */
+#if @GNULIB_STRPBRK@
+# if ! @HAVE_STRPBRK@
+extern char *strpbrk (char const *__s, char const *__accept)
+ __attribute__ ((__pure__));
+# endif
+# if defined GNULIB_POSIXCHECK
+/* strpbrk() assumes the second argument is a list of single-byte characters.
+ Even in this simple case, it does not work with multibyte strings if the
+ locale encoding is GB18030 and one of the characters to be searched is a
+ digit. */
+# undef strpbrk
+# define strpbrk(s,a) \
+ (GL_LINK_WARNING ("strpbrk cannot work correctly on character strings " \
+ "in multibyte locales - " \
+ "use mbspbrk if you care about internationalization"), \
+ strpbrk (s, a))
+# endif
+#elif defined GNULIB_POSIXCHECK
+# undef strpbrk
+# define strpbrk(s,a) \
+ (GL_LINK_WARNING ("strpbrk is unportable - " \
+ "use gnulib module strpbrk for portability"), \
+ strpbrk (s, a))
+#endif
+
+#if defined GNULIB_POSIXCHECK
+/* strspn() assumes the second argument is a list of single-byte characters.
+ Even in this simple case, it cannot work with multibyte strings. */
+# undef strspn
+# define strspn(s,a) \
+ (GL_LINK_WARNING ("strspn cannot work correctly on character strings " \
+ "in multibyte locales - " \
+ "use mbsspn if you care about internationalization"), \
+ strspn (s, a))
+#endif
+
+#if defined GNULIB_POSIXCHECK
+/* strrchr() does not work with multibyte strings if the locale encoding is
+ GB18030 and the character to be searched is a digit. */
+# undef strrchr
+# define strrchr(s,c) \
+ (GL_LINK_WARNING ("strrchr cannot work correctly on character strings " \
+ "in some multibyte locales - " \
+ "use mbsrchr if you care about internationalization"), \
+ strrchr (s, c))
+#endif
+
+/* Search the next delimiter (char listed in DELIM) starting at *STRINGP.
+ If one is found, overwrite it with a NUL, and advance *STRINGP
+ to point to the next char after it. Otherwise, set *STRINGP to NULL.
+ If *STRINGP was already NULL, nothing happens.
+ Return the old value of *STRINGP.
+
+ This is a variant of strtok() that is multithread-safe and supports
+ empty fields.
+
+ Caveat: It modifies the original string.
+ Caveat: These functions cannot be used on constant strings.
+ Caveat: The identity of the delimiting character is lost.
+ Caveat: It doesn't work with multibyte strings unless all of the delimiter
+ characters are ASCII characters < 0x30.
+
+ See also strtok_r(). */
+#if @GNULIB_STRSEP@
+# if ! @HAVE_STRSEP@
+extern char *strsep (char **restrict __stringp, char const *restrict __delim);
+# endif
+# if defined GNULIB_POSIXCHECK
+# undef strsep
+# define strsep(s,d) \
+ (GL_LINK_WARNING ("strsep cannot work correctly on character strings " \
+ "in multibyte locales - " \
+ "use mbssep if you care about internationalization"), \
+ strsep (s, d))
+# endif
+#elif defined GNULIB_POSIXCHECK
+# undef strsep
+# define strsep(s,d) \
+ (GL_LINK_WARNING ("strsep is unportable - " \
+ "use gnulib module strsep for portability"), \
+ strsep (s, d))
+#endif
+
+#if @GNULIB_STRSTR@
+# if @REPLACE_STRSTR@
+# define strstr rpl_strstr
+char *strstr (const char *haystack, const char *needle)
+ __attribute__ ((__pure__));
+# endif
+#elif defined GNULIB_POSIXCHECK
+/* strstr() does not work with multibyte strings if the locale encoding is
+ different from UTF-8:
+ POSIX says that it operates on "strings", and "string" in POSIX is defined
+ as a sequence of bytes, not of characters. */
+# undef strstr
+# define strstr(a,b) \
+ (GL_LINK_WARNING ("strstr is quadratic on many systems, and cannot " \
+ "work correctly on character strings in most " \
+ "multibyte locales - " \
+ "use mbsstr if you care about internationalization, " \
+ "or use strstr if you care about speed"), \
+ strstr (a, b))
+#endif
+
+/* Find the first occurrence of NEEDLE in HAYSTACK, using case-insensitive
+ comparison. */
+#if @GNULIB_STRCASESTR@
+# if @REPLACE_STRCASESTR@
+# define strcasestr rpl_strcasestr
+# endif
+# if ! @HAVE_STRCASESTR@ || @REPLACE_STRCASESTR@
+extern char *strcasestr (const char *haystack, const char *needle)
+ __attribute__ ((__pure__));
+# endif
+#elif defined GNULIB_POSIXCHECK
+/* strcasestr() does not work with multibyte strings:
+ It is a glibc extension, and glibc implements it only for unibyte
+ locales. */
+# undef strcasestr
+# define strcasestr(a,b) \
+ (GL_LINK_WARNING ("strcasestr does work correctly on character strings " \
+ "in multibyte locales - " \
+ "use mbscasestr if you care about " \
+ "internationalization, or use c-strcasestr if you want "
\
+ "a locale independent function"), \
+ strcasestr (a, b))
+#endif
+
+/* Parse S into tokens separated by characters in DELIM.
+ If S is NULL, the saved pointer in SAVE_PTR is used as
+ the next starting point. For example:
+ char s[] = "-abc-=-def";
+ char *sp;
+ x = strtok_r(s, "-", &sp); // x = "abc", sp = "=-def"
+ x = strtok_r(NULL, "-=", &sp); // x = "def", sp = NULL
+ x = strtok_r(NULL, "=", &sp); // x = NULL
+ // s = "abc\0-def\0"
+
+ This is a variant of strtok() that is multithread-safe.
+
+ For the POSIX documentation for this function, see:
+ http://www.opengroup.org/susv3xsh/strtok.html
+
+ Caveat: It modifies the original string.
+ Caveat: These functions cannot be used on constant strings.
+ Caveat: The identity of the delimiting character is lost.
+ Caveat: It doesn't work with multibyte strings unless all of the delimiter
+ characters are ASCII characters < 0x30.
+
+ See also strsep(). */
+#if @GNULIB_STRTOK_R@
+# if ! @HAVE_DECL_STRTOK_R@
+extern char *strtok_r (char *restrict s, char const *restrict delim,
+ char **restrict save_ptr);
+# endif
+# if defined GNULIB_POSIXCHECK
+# undef strtok_r
+# define strtok_r(s,d,p) \
+ (GL_LINK_WARNING ("strtok_r cannot work correctly on character strings " \
+ "in multibyte locales - " \
+ "use mbstok_r if you care about internationalization"),
\
+ strtok_r (s, d, p))
+# endif
+#elif defined GNULIB_POSIXCHECK
+# undef strtok_r
+# define strtok_r(s,d,p) \
+ (GL_LINK_WARNING ("strtok_r is unportable - " \
+ "use gnulib module strtok_r for portability"), \
+ strtok_r (s, d, p))
+#endif
+
+
+/* The following functions are not specified by POSIX. They are gnulib
+ extensions. */
+
+#if @GNULIB_MBSLEN@
+/* Return the number of multibyte characters in the character string STRING.
+ This considers multibyte characters, unlike strlen, which counts bytes. */
+extern size_t mbslen (const char *string);
+#endif
+
+#if @GNULIB_MBSNLEN@
+/* Return the number of multibyte characters in the character string starting
+ at STRING and ending at STRING + LEN. */
+extern size_t mbsnlen (const char *string, size_t len);
+#endif
+
+#if @GNULIB_MBSCHR@
+/* Locate the first single-byte character C in the character string STRING,
+ and return a pointer to it. Return NULL if C is not found in STRING.
+ Unlike strchr(), this function works correctly in multibyte locales with
+ encodings such as GB18030. */
+# define mbschr rpl_mbschr /* avoid collision with HP-UX function */
+extern char * mbschr (const char *string, int c);
+#endif
+
+#if @GNULIB_MBSRCHR@
+/* Locate the last single-byte character C in the character string STRING,
+ and return a pointer to it. Return NULL if C is not found in STRING.
+ Unlike strrchr(), this function works correctly in multibyte locales with
+ encodings such as GB18030. */
+# define mbsrchr rpl_mbsrchr /* avoid collision with HP-UX function */
+extern char * mbsrchr (const char *string, int c);
+#endif
+
+#if @GNULIB_MBSSTR@
+/* Find the first occurrence of the character string NEEDLE in the character
+ string HAYSTACK. Return NULL if NEEDLE is not found in HAYSTACK.
+ Unlike strstr(), this function works correctly in multibyte locales with
+ encodings different from UTF-8. */
+extern char * mbsstr (const char *haystack, const char *needle);
+#endif
+
+#if @GNULIB_MBSCASECMP@
+/* Compare the character strings S1 and S2, ignoring case, returning less than,
+ equal to or greater than zero if S1 is lexicographically less than, equal to
+ or greater than S2.
+ Note: This function may, in multibyte locales, return 0 for strings of
+ different lengths!
+ Unlike strcasecmp(), this function works correctly in multibyte locales. */
+extern int mbscasecmp (const char *s1, const char *s2);
+#endif
+
+#if @GNULIB_MBSNCASECMP@
+/* Compare the initial segment of the character string S1 consisting of at most
+ N characters with the initial segment of the character string S2 consisting
+ of at most N characters, ignoring case, returning less than, equal to or
+ greater than zero if the initial segment of S1 is lexicographically less
+ than, equal to or greater than the initial segment of S2.
+ Note: This function may, in multibyte locales, return 0 for initial segments
+ of different lengths!
+ Unlike strncasecmp(), this function works correctly in multibyte locales.
+ But beware that N is not a byte count but a character count! */
+extern int mbsncasecmp (const char *s1, const char *s2, size_t n);
+#endif
+
+#if @GNULIB_MBSPCASECMP@
+/* Compare the initial segment of the character string STRING consisting of
+ at most mbslen (PREFIX) characters with the character string PREFIX,
+ ignoring case, returning less than, equal to or greater than zero if this
+ initial segment is lexicographically less than, equal to or greater than
+ PREFIX.
+ Note: This function may, in multibyte locales, return 0 if STRING is of
+ smaller length than PREFIX!
+ Unlike strncasecmp(), this function works correctly in multibyte
+ locales. */
+extern char * mbspcasecmp (const char *string, const char *prefix);
+#endif
+
+#if @GNULIB_MBSCASESTR@
+/* Find the first occurrence of the character string NEEDLE in the character
+ string HAYSTACK, using case-insensitive comparison.
+ Note: This function may, in multibyte locales, return success even if
+ strlen (haystack) < strlen (needle) !
+ Unlike strcasestr(), this function works correctly in multibyte locales. */
+extern char * mbscasestr (const char *haystack, const char *needle);
+#endif
+
+#if @GNULIB_MBSCSPN@
+/* Find the first occurrence in the character string STRING of any character
+ in the character string ACCEPT. Return the number of bytes from the
+ beginning of the string to this occurrence, or to the end of the string
+ if none exists.
+ Unlike strcspn(), this function works correctly in multibyte locales. */
+extern size_t mbscspn (const char *string, const char *accept);
+#endif
+
+#if @GNULIB_MBSPBRK@
+/* Find the first occurrence in the character string STRING of any character
+ in the character string ACCEPT. Return the pointer to it, or NULL if none
+ exists.
+ Unlike strpbrk(), this function works correctly in multibyte locales. */
+# define mbspbrk rpl_mbspbrk /* avoid collision with HP-UX function */
+extern char * mbspbrk (const char *string, const char *accept);
+#endif
+
+#if @GNULIB_MBSSPN@
+/* Find the first occurrence in the character string STRING of any character
+ not in the character string REJECT. Return the number of bytes from the
+ beginning of the string to this occurrence, or to the end of the string
+ if none exists.
+ Unlike strspn(), this function works correctly in multibyte locales. */
+extern size_t mbsspn (const char *string, const char *reject);
+#endif
+
+#if @GNULIB_MBSSEP@
+/* Search the next delimiter (multibyte character listed in the character
+ string DELIM) starting at the character string *STRINGP.
+ If one is found, overwrite it with a NUL, and advance *STRINGP to point
+ to the next multibyte character after it. Otherwise, set *STRINGP to NULL.
+ If *STRINGP was already NULL, nothing happens.
+ Return the old value of *STRINGP.
+
+ This is a variant of mbstok_r() that supports empty fields.
+
+ Caveat: It modifies the original string.
+ Caveat: These functions cannot be used on constant strings.
+ Caveat: The identity of the delimiting character is lost.
+
+ See also mbstok_r(). */
+extern char * mbssep (char **stringp, const char *delim);
+#endif
+
+#if @GNULIB_MBSTOK_R@
+/* Parse the character string STRING into tokens separated by characters in
+ the character string DELIM.
+ If STRING is NULL, the saved pointer in SAVE_PTR is used as
+ the next starting point. For example:
+ char s[] = "-abc-=-def";
+ char *sp;
+ x = mbstok_r(s, "-", &sp); // x = "abc", sp = "=-def"
+ x = mbstok_r(NULL, "-=", &sp); // x = "def", sp = NULL
+ x = mbstok_r(NULL, "=", &sp); // x = NULL
+ // s = "abc\0-def\0"
+
+ Caveat: It modifies the original string.
+ Caveat: These functions cannot be used on constant strings.
+ Caveat: The identity of the delimiting character is lost.
+
+ See also mbssep(). */
+extern char * mbstok_r (char *string, const char *delim, char **save_ptr);
+#endif
+
+/* Map any int, typically from errno, into an error message. */
+#if @GNULIB_STRERROR@
+# if @REPLACE_STRERROR@
+# undef strerror
+# define strerror rpl_strerror
+extern char *strerror (int);
+# endif
+#elif defined GNULIB_POSIXCHECK
+# undef strerror
+# define strerror(e) \
+ (GL_LINK_WARNING ("strerror is unportable - " \
+ "use gnulib module strerror to guarantee non-NULL
result"), \
+ strerror (e))
+#endif
+
+#if @GNULIB_STRSIGNAL@
+# if @REPLACE_STRSIGNAL@
+# define strsignal rpl_strsignal
+# endif
+# if ! @HAVE_DECL_STRSIGNAL@ || @REPLACE_STRSIGNAL@
+extern char *strsignal (int __sig);
+# endif
+#elif defined GNULIB_POSIXCHECK
+# undef strsignal
+# define strsignal(a) \
+ (GL_LINK_WARNING ("strsignal is unportable - " \
+ "use gnulib module strsignal for portability"), \
+ strsignal (a))
+#endif
+
+#if @GNULIB_STRVERSCMP@
+# if address@hidden@
+extern int strverscmp (const char *, const char *);
+# endif
+#elif defined GNULIB_POSIXCHECK
+# undef strverscmp
+# define strverscmp(a, b) \
+ (GL_LINK_WARNING ("strverscmp is unportable - " \
+ "use gnulib module strverscmp for portability"), \
+ strverscmp (a, b))
+#endif
+
+
+#ifdef __cplusplus
+}
+#endif
+
+#endif /* _GL_STRING_H */
+#endif /* _GL_STRING_H */
diff --git a/lib/unistr.h b/lib/unistr.h
new file mode 100644
index 0000000..83ff134
--- /dev/null
+++ b/lib/unistr.h
@@ -0,0 +1,681 @@
+/* Elementary Unicode string functions.
+ Copyright (C) 2001-2002, 2005-2009 Free Software Foundation, Inc.
+
+ This program is free software: you can redistribute it and/or modify it
+ under the terms of the GNU Lesser General Public License as published
+ by the Free Software Foundation; either version 3 of the License, or
+ (at your option) any later version.
+
+ This program is distributed in the hope that it will be useful,
+ but WITHOUT ANY WARRANTY; without even the implied warranty of
+ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+ Lesser General Public License for more details.
+
+ You should have received a copy of the GNU Lesser General Public License
+ along with this program. If not, see <http://www.gnu.org/licenses/>. */
+
+#ifndef _UNISTR_H
+#define _UNISTR_H
+
+#include "unitypes.h"
+
+/* Get bool. */
+#include <stdbool.h>
+
+/* Get size_t. */
+#include <stddef.h>
+
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+
+/* Conventions:
+
+ All functions prefixed with u8_ operate on UTF-8 encoded strings.
+ Their unit is an uint8_t (1 byte).
+
+ All functions prefixed with u16_ operate on UTF-16 encoded strings.
+ Their unit is an uint16_t (a 2-byte word).
+
+ All functions prefixed with u32_ operate on UCS-4 encoded strings.
+ Their unit is an uint32_t (a 4-byte word).
+
+ All argument pairs (s, n) denote a Unicode string s[0..n-1] with exactly
+ n units.
+
+ All arguments starting with "str" and the arguments of functions starting
+ with u8_str/u16_str/u32_str denote a NUL terminated string, i.e. a string
+ which terminates at the first NUL unit. This termination unit is
+ considered part of the string for all memory allocation purposes, but
+ is not considered part of the string for all other logical purposes.
+
+ Functions returning a string result take a (resultbuf, lengthp) argument
+ pair. If resultbuf is not NULL and the result fits into *lengthp units,
+ it is put in resultbuf, and resultbuf is returned. Otherwise, a freshly
+ allocated string is returned. In both cases, *lengthp is set to the
+ length (number of units) of the returned string. In case of error,
+ NULL is returned and errno is set. */
+
+
+/* Elementary string checks. */
+
+/* Check whether an UTF-8 string is well-formed.
+ Return NULL if valid, or a pointer to the first invalid unit otherwise. */
+extern const uint8_t *
+ u8_check (const uint8_t *s, size_t n);
+
+/* Check whether an UTF-16 string is well-formed.
+ Return NULL if valid, or a pointer to the first invalid unit otherwise. */
+extern const uint16_t *
+ u16_check (const uint16_t *s, size_t n);
+
+/* Check whether an UCS-4 string is well-formed.
+ Return NULL if valid, or a pointer to the first invalid unit otherwise. */
+extern const uint32_t *
+ u32_check (const uint32_t *s, size_t n);
+
+
+/* Elementary string conversions. */
+
+/* Convert an UTF-8 string to an UTF-16 string. */
+extern uint16_t *
+ u8_to_u16 (const uint8_t *s, size_t n, uint16_t *resultbuf,
+ size_t *lengthp);
+
+/* Convert an UTF-8 string to an UCS-4 string. */
+extern uint32_t *
+ u8_to_u32 (const uint8_t *s, size_t n, uint32_t *resultbuf,
+ size_t *lengthp);
+
+/* Convert an UTF-16 string to an UTF-8 string. */
+extern uint8_t *
+ u16_to_u8 (const uint16_t *s, size_t n, uint8_t *resultbuf,
+ size_t *lengthp);
+
+/* Convert an UTF-16 string to an UCS-4 string. */
+extern uint32_t *
+ u16_to_u32 (const uint16_t *s, size_t n, uint32_t *resultbuf,
+ size_t *lengthp);
+
+/* Convert an UCS-4 string to an UTF-8 string. */
+extern uint8_t *
+ u32_to_u8 (const uint32_t *s, size_t n, uint8_t *resultbuf,
+ size_t *lengthp);
+
+/* Convert an UCS-4 string to an UTF-16 string. */
+extern uint16_t *
+ u32_to_u16 (const uint32_t *s, size_t n, uint16_t *resultbuf,
+ size_t *lengthp);
+
+
+/* Elementary string functions. */
+
+/* Return the length (number of units) of the first character in S, which is
+ no longer than N. Return 0 if it is the NUL character. Return -1 upon
+ failure. */
+/* Similar to mblen(), except that s must not be NULL. */
+extern int
+ u8_mblen (const uint8_t *s, size_t n);
+extern int
+ u16_mblen (const uint16_t *s, size_t n);
+extern int
+ u32_mblen (const uint32_t *s, size_t n);
+
+/* Return the length (number of units) of the first character in S, putting
+ its 'ucs4_t' representation in *PUC. Upon failure, *PUC is set to 0xfffd,
+ and an appropriate number of units is returned.
+ The number of available units, N, must be > 0. */
+/* Similar to mbtowc(), except that puc and s must not be NULL, n must be > 0,
+ and the NUL character is not treated specially. */
+/* The variants with _safe suffix are safe, even if the library is compiled
+ without --enable-safety. */
+
+#ifdef GNULIB_UNISTR_U8_MBTOUC_UNSAFE
+# if !HAVE_INLINE
+extern int
+ u8_mbtouc_unsafe (ucs4_t *puc, const uint8_t *s, size_t n);
+# else
+extern int
+ u8_mbtouc_unsafe_aux (ucs4_t *puc, const uint8_t *s, size_t n);
+static inline int
+u8_mbtouc_unsafe (ucs4_t *puc, const uint8_t *s, size_t n)
+{
+ uint8_t c = *s;
+
+ if (c < 0x80)
+ {
+ *puc = c;
+ return 1;
+ }
+ else
+ return u8_mbtouc_unsafe_aux (puc, s, n);
+}
+# endif
+#endif
+
+#ifdef GNULIB_UNISTR_U16_MBTOUC_UNSAFE
+# if !HAVE_INLINE
+extern int
+ u16_mbtouc_unsafe (ucs4_t *puc, const uint16_t *s, size_t n);
+# else
+extern int
+ u16_mbtouc_unsafe_aux (ucs4_t *puc, const uint16_t *s, size_t n);
+static inline int
+u16_mbtouc_unsafe (ucs4_t *puc, const uint16_t *s, size_t n)
+{
+ uint16_t c = *s;
+
+ if (c < 0xd800 || c >= 0xe000)
+ {
+ *puc = c;
+ return 1;
+ }
+ else
+ return u16_mbtouc_unsafe_aux (puc, s, n);
+}
+# endif
+#endif
+
+#ifdef GNULIB_UNISTR_U32_MBTOUC_UNSAFE
+# if !HAVE_INLINE
+extern int
+ u32_mbtouc_unsafe (ucs4_t *puc, const uint32_t *s, size_t n);
+# else
+static inline int
+u32_mbtouc_unsafe (ucs4_t *puc, const uint32_t *s, size_t n _UNUSED_PARAMETER_)
+{
+ uint32_t c = *s;
+
+# if CONFIG_UNICODE_SAFETY
+ if (c < 0xd800 || (c >= 0xe000 && c < 0x110000))
+# endif
+ *puc = c;
+# if CONFIG_UNICODE_SAFETY
+ else
+ /* invalid multibyte character */
+ *puc = 0xfffd;
+# endif
+ return 1;
+}
+# endif
+#endif
+
+#ifdef GNULIB_UNISTR_U8_MBTOUC
+# if !HAVE_INLINE
+extern int
+ u8_mbtouc (ucs4_t *puc, const uint8_t *s, size_t n);
+# else
+extern int
+ u8_mbtouc_aux (ucs4_t *puc, const uint8_t *s, size_t n);
+static inline int
+u8_mbtouc (ucs4_t *puc, const uint8_t *s, size_t n)
+{
+ uint8_t c = *s;
+
+ if (c < 0x80)
+ {
+ *puc = c;
+ return 1;
+ }
+ else
+ return u8_mbtouc_aux (puc, s, n);
+}
+# endif
+#endif
+
+#ifdef GNULIB_UNISTR_U16_MBTOUC
+# if !HAVE_INLINE
+extern int
+ u16_mbtouc (ucs4_t *puc, const uint16_t *s, size_t n);
+# else
+extern int
+ u16_mbtouc_aux (ucs4_t *puc, const uint16_t *s, size_t n);
+static inline int
+u16_mbtouc (ucs4_t *puc, const uint16_t *s, size_t n)
+{
+ uint16_t c = *s;
+
+ if (c < 0xd800 || c >= 0xe000)
+ {
+ *puc = c;
+ return 1;
+ }
+ else
+ return u16_mbtouc_aux (puc, s, n);
+}
+# endif
+#endif
+
+#ifdef GNULIB_UNISTR_U32_MBTOUC
+# if !HAVE_INLINE
+extern int
+ u32_mbtouc (ucs4_t *puc, const uint32_t *s, size_t n);
+# else
+static inline int
+u32_mbtouc (ucs4_t *puc, const uint32_t *s, size_t n _UNUSED_PARAMETER_)
+{
+ uint32_t c = *s;
+
+ if (c < 0xd800 || (c >= 0xe000 && c < 0x110000))
+ *puc = c;
+ else
+ /* invalid multibyte character */
+ *puc = 0xfffd;
+ return 1;
+}
+# endif
+#endif
+
+/* Return the length (number of units) of the first character in S, putting
+ its 'ucs4_t' representation in *PUC. Upon failure, *PUC is set to 0xfffd,
+ and -1 is returned for an invalid sequence of units, -2 is returned for an
+ incomplete sequence of units.
+ The number of available units, N, must be > 0. */
+/* Similar to u*_mbtouc(), except that the return value gives more details
+ about the failure, similar to mbrtowc(). */
+
+#ifdef GNULIB_UNISTR_U8_MBTOUCR
+extern int
+ u8_mbtoucr (ucs4_t *puc, const uint8_t *s, size_t n);
+#endif
+
+#ifdef GNULIB_UNISTR_U16_MBTOUCR
+extern int
+ u16_mbtoucr (ucs4_t *puc, const uint16_t *s, size_t n);
+#endif
+
+#ifdef GNULIB_UNISTR_U32_MBTOUCR
+extern int
+ u32_mbtoucr (ucs4_t *puc, const uint32_t *s, size_t n);
+#endif
+
+/* Put the multibyte character represented by UC in S, returning its
+ length. Return -1 upon failure, -2 if the number of available units, N,
+ is too small. The latter case cannot occur if N >= 6/2/1, respectively. */
+/* Similar to wctomb(), except that s must not be NULL, and the argument n
+ must be specified. */
+
+#ifdef GNULIB_UNISTR_U8_UCTOMB
+/* Auxiliary function, also used by u8_chr, u8_strchr, u8_strrchr. */
+extern int
+ u8_uctomb_aux (uint8_t *s, ucs4_t uc, int n);
+# if !HAVE_INLINE
+extern int
+ u8_uctomb (uint8_t *s, ucs4_t uc, int n);
+# else
+static inline int
+u8_uctomb (uint8_t *s, ucs4_t uc, int n)
+{
+ if (uc < 0x80 && n > 0)
+ {
+ s[0] = uc;
+ return 1;
+ }
+ else
+ return u8_uctomb_aux (s, uc, n);
+}
+# endif
+#endif
+
+#ifdef GNULIB_UNISTR_U16_UCTOMB
+/* Auxiliary function, also used by u16_chr, u16_strchr, u16_strrchr. */
+extern int
+ u16_uctomb_aux (uint16_t *s, ucs4_t uc, int n);
+# if !HAVE_INLINE
+extern int
+ u16_uctomb (uint16_t *s, ucs4_t uc, int n);
+# else
+static inline int
+u16_uctomb (uint16_t *s, ucs4_t uc, int n)
+{
+ if (uc < 0xd800 && n > 0)
+ {
+ s[0] = uc;
+ return 1;
+ }
+ else
+ return u16_uctomb_aux (s, uc, n);
+}
+# endif
+#endif
+
+#ifdef GNULIB_UNISTR_U32_UCTOMB
+# if !HAVE_INLINE
+extern int
+ u32_uctomb (uint32_t *s, ucs4_t uc, int n);
+# else
+static inline int
+u32_uctomb (uint32_t *s, ucs4_t uc, int n)
+{
+ if (uc < 0xd800 || (uc >= 0xe000 && uc < 0x110000))
+ {
+ if (n > 0)
+ {
+ *s = uc;
+ return 1;
+ }
+ else
+ return -2;
+ }
+ else
+ return -1;
+}
+# endif
+#endif
+
+/* Copy N units from SRC to DEST. */
+/* Similar to memcpy(). */
+extern uint8_t *
+ u8_cpy (uint8_t *dest, const uint8_t *src, size_t n);
+extern uint16_t *
+ u16_cpy (uint16_t *dest, const uint16_t *src, size_t n);
+extern uint32_t *
+ u32_cpy (uint32_t *dest, const uint32_t *src, size_t n);
+
+/* Copy N units from SRC to DEST, guaranteeing correct behavior for
+ overlapping memory areas. */
+/* Similar to memmove(). */
+extern uint8_t *
+ u8_move (uint8_t *dest, const uint8_t *src, size_t n);
+extern uint16_t *
+ u16_move (uint16_t *dest, const uint16_t *src, size_t n);
+extern uint32_t *
+ u32_move (uint32_t *dest, const uint32_t *src, size_t n);
+
+/* Set the first N characters of S to UC. UC should be a character that
+ occupies only 1 unit. */
+/* Similar to memset(). */
+extern uint8_t *
+ u8_set (uint8_t *s, ucs4_t uc, size_t n);
+extern uint16_t *
+ u16_set (uint16_t *s, ucs4_t uc, size_t n);
+extern uint32_t *
+ u32_set (uint32_t *s, ucs4_t uc, size_t n);
+
+/* Compare S1 and S2, each of length N. */
+/* Similar to memcmp(). */
+extern int
+ u8_cmp (const uint8_t *s1, const uint8_t *s2, size_t n);
+extern int
+ u16_cmp (const uint16_t *s1, const uint16_t *s2, size_t n);
+extern int
+ u32_cmp (const uint32_t *s1, const uint32_t *s2, size_t n);
+
+/* Compare S1 and S2. */
+/* Similar to the gnulib function memcmp2(). */
+extern int
+ u8_cmp2 (const uint8_t *s1, size_t n1, const uint8_t *s2, size_t n2);
+extern int
+ u16_cmp2 (const uint16_t *s1, size_t n1, const uint16_t *s2, size_t n2);
+extern int
+ u32_cmp2 (const uint32_t *s1, size_t n1, const uint32_t *s2, size_t n2);
+
+/* Search the string at S for UC. */
+/* Similar to memchr(). */
+extern uint8_t *
+ u8_chr (const uint8_t *s, size_t n, ucs4_t uc);
+extern uint16_t *
+ u16_chr (const uint16_t *s, size_t n, ucs4_t uc);
+extern uint32_t *
+ u32_chr (const uint32_t *s, size_t n, ucs4_t uc);
+
+/* Count the number of Unicode characters in the N units from S. */
+/* Similar to mbsnlen(). */
+extern size_t
+ u8_mbsnlen (const uint8_t *s, size_t n);
+extern size_t
+ u16_mbsnlen (const uint16_t *s, size_t n);
+extern size_t
+ u32_mbsnlen (const uint32_t *s, size_t n);
+
+/* Elementary string functions with memory allocation. */
+
+/* Make a freshly allocated copy of S, of length N. */
+extern uint8_t *
+ u8_cpy_alloc (const uint8_t *s, size_t n);
+extern uint16_t *
+ u16_cpy_alloc (const uint16_t *s, size_t n);
+extern uint32_t *
+ u32_cpy_alloc (const uint32_t *s, size_t n);
+
+/* Elementary string functions on NUL terminated strings. */
+
+/* Return the length (number of units) of the first character in S.
+ Return 0 if it is the NUL character. Return -1 upon failure. */
+extern int
+ u8_strmblen (const uint8_t *s);
+extern int
+ u16_strmblen (const uint16_t *s);
+extern int
+ u32_strmblen (const uint32_t *s);
+
+/* Return the length (number of units) of the first character in S, putting
+ its 'ucs4_t' representation in *PUC. Return 0 if it is the NUL
+ character. Return -1 upon failure. */
+extern int
+ u8_strmbtouc (ucs4_t *puc, const uint8_t *s);
+extern int
+ u16_strmbtouc (ucs4_t *puc, const uint16_t *s);
+extern int
+ u32_strmbtouc (ucs4_t *puc, const uint32_t *s);
+
+/* Forward iteration step. Advances the pointer past the next character,
+ or returns NULL if the end of the string has been reached. Puts the
+ character's 'ucs4_t' representation in *PUC. */
+extern const uint8_t *
+ u8_next (ucs4_t *puc, const uint8_t *s);
+extern const uint16_t *
+ u16_next (ucs4_t *puc, const uint16_t *s);
+extern const uint32_t *
+ u32_next (ucs4_t *puc, const uint32_t *s);
+
+/* Backward iteration step. Advances the pointer to point to the previous
+ character, or returns NULL if the beginning of the string had been reached.
+ Puts the character's 'ucs4_t' representation in *PUC. */
+extern const uint8_t *
+ u8_prev (ucs4_t *puc, const uint8_t *s, const uint8_t *start);
+extern const uint16_t *
+ u16_prev (ucs4_t *puc, const uint16_t *s, const uint16_t *start);
+extern const uint32_t *
+ u32_prev (ucs4_t *puc, const uint32_t *s, const uint32_t *start);
+
+/* Return the number of units in S. */
+/* Similar to strlen(), wcslen(). */
+extern size_t
+ u8_strlen (const uint8_t *s);
+extern size_t
+ u16_strlen (const uint16_t *s);
+extern size_t
+ u32_strlen (const uint32_t *s);
+
+/* Return the number of units in S, but at most MAXLEN. */
+/* Similar to strnlen(), wcsnlen(). */
+extern size_t
+ u8_strnlen (const uint8_t *s, size_t maxlen);
+extern size_t
+ u16_strnlen (const uint16_t *s, size_t maxlen);
+extern size_t
+ u32_strnlen (const uint32_t *s, size_t maxlen);
+
+/* Copy SRC to DEST. */
+/* Similar to strcpy(), wcscpy(). */
+extern uint8_t *
+ u8_strcpy (uint8_t *dest, const uint8_t *src);
+extern uint16_t *
+ u16_strcpy (uint16_t *dest, const uint16_t *src);
+extern uint32_t *
+ u32_strcpy (uint32_t *dest, const uint32_t *src);
+
+/* Copy SRC to DEST, returning the address of the terminating NUL in DEST. */
+/* Similar to stpcpy(). */
+extern uint8_t *
+ u8_stpcpy (uint8_t *dest, const uint8_t *src);
+extern uint16_t *
+ u16_stpcpy (uint16_t *dest, const uint16_t *src);
+extern uint32_t *
+ u32_stpcpy (uint32_t *dest, const uint32_t *src);
+
+/* Copy no more than N units of SRC to DEST. */
+/* Similar to strncpy(), wcsncpy(). */
+extern uint8_t *
+ u8_strncpy (uint8_t *dest, const uint8_t *src, size_t n);
+extern uint16_t *
+ u16_strncpy (uint16_t *dest, const uint16_t *src, size_t n);
+extern uint32_t *
+ u32_strncpy (uint32_t *dest, const uint32_t *src, size_t n);
+
+/* Copy no more than N units of SRC to DEST, returning the address of
+ the last unit written into DEST. */
+/* Similar to stpncpy(). */
+extern uint8_t *
+ u8_stpncpy (uint8_t *dest, const uint8_t *src, size_t n);
+extern uint16_t *
+ u16_stpncpy (uint16_t *dest, const uint16_t *src, size_t n);
+extern uint32_t *
+ u32_stpncpy (uint32_t *dest, const uint32_t *src, size_t n);
+
+/* Append SRC onto DEST. */
+/* Similar to strcat(), wcscat(). */
+extern uint8_t *
+ u8_strcat (uint8_t *dest, const uint8_t *src);
+extern uint16_t *
+ u16_strcat (uint16_t *dest, const uint16_t *src);
+extern uint32_t *
+ u32_strcat (uint32_t *dest, const uint32_t *src);
+
+/* Append no more than N units of SRC onto DEST. */
+/* Similar to strncat(), wcsncat(). */
+extern uint8_t *
+ u8_strncat (uint8_t *dest, const uint8_t *src, size_t n);
+extern uint16_t *
+ u16_strncat (uint16_t *dest, const uint16_t *src, size_t n);
+extern uint32_t *
+ u32_strncat (uint32_t *dest, const uint32_t *src, size_t n);
+
+/* Compare S1 and S2. */
+/* Similar to strcmp(), wcscmp(). */
+extern int
+ u8_strcmp (const uint8_t *s1, const uint8_t *s2);
+extern int
+ u16_strcmp (const uint16_t *s1, const uint16_t *s2);
+extern int
+ u32_strcmp (const uint32_t *s1, const uint32_t *s2);
+
+/* Compare S1 and S2 using the collation rules of the current locale.
+ Return -1 if S1 < S2, 0 if S1 = S2, 1 if S1 > S2.
+ Upon failure, set errno and return any value. */
+/* Similar to strcoll(), wcscoll(). */
+extern int
+ u8_strcoll (const uint8_t *s1, const uint8_t *s2);
+extern int
+ u16_strcoll (const uint16_t *s1, const uint16_t *s2);
+extern int
+ u32_strcoll (const uint32_t *s1, const uint32_t *s2);
+
+/* Compare no more than N units of S1 and S2. */
+/* Similar to strncmp(), wcsncmp(). */
+extern int
+ u8_strncmp (const uint8_t *s1, const uint8_t *s2, size_t n);
+extern int
+ u16_strncmp (const uint16_t *s1, const uint16_t *s2, size_t n);
+extern int
+ u32_strncmp (const uint32_t *s1, const uint32_t *s2, size_t n);
+
+/* Duplicate S, returning an identical malloc'd string. */
+/* Similar to strdup(), wcsdup(). */
+extern uint8_t *
+ u8_strdup (const uint8_t *s);
+extern uint16_t *
+ u16_strdup (const uint16_t *s);
+extern uint32_t *
+ u32_strdup (const uint32_t *s);
+
+/* Find the first occurrence of UC in STR. */
+/* Similar to strchr(), wcschr(). */
+extern uint8_t *
+ u8_strchr (const uint8_t *str, ucs4_t uc);
+extern uint16_t *
+ u16_strchr (const uint16_t *str, ucs4_t uc);
+extern uint32_t *
+ u32_strchr (const uint32_t *str, ucs4_t uc);
+
+/* Find the last occurrence of UC in STR. */
+/* Similar to strrchr(), wcsrchr(). */
+extern uint8_t *
+ u8_strrchr (const uint8_t *str, ucs4_t uc);
+extern uint16_t *
+ u16_strrchr (const uint16_t *str, ucs4_t uc);
+extern uint32_t *
+ u32_strrchr (const uint32_t *str, ucs4_t uc);
+
+/* Return the length of the initial segment of STR which consists entirely
+ of Unicode characters not in REJECT. */
+/* Similar to strcspn(), wcscspn(). */
+extern size_t
+ u8_strcspn (const uint8_t *str, const uint8_t *reject);
+extern size_t
+ u16_strcspn (const uint16_t *str, const uint16_t *reject);
+extern size_t
+ u32_strcspn (const uint32_t *str, const uint32_t *reject);
+
+/* Return the length of the initial segment of STR which consists entirely
+ of Unicode characters in ACCEPT. */
+/* Similar to strspn(), wcsspn(). */
+extern size_t
+ u8_strspn (const uint8_t *str, const uint8_t *accept);
+extern size_t
+ u16_strspn (const uint16_t *str, const uint16_t *accept);
+extern size_t
+ u32_strspn (const uint32_t *str, const uint32_t *accept);
+
+/* Find the first occurrence in STR of any character in ACCEPT. */
+/* Similar to strpbrk(), wcspbrk(). */
+extern uint8_t *
+ u8_strpbrk (const uint8_t *str, const uint8_t *accept);
+extern uint16_t *
+ u16_strpbrk (const uint16_t *str, const uint16_t *accept);
+extern uint32_t *
+ u32_strpbrk (const uint32_t *str, const uint32_t *accept);
+
+/* Find the first occurrence of NEEDLE in HAYSTACK. */
+/* Similar to strstr(), wcsstr(). */
+extern uint8_t *
+ u8_strstr (const uint8_t *haystack, const uint8_t *needle);
+extern uint16_t *
+ u16_strstr (const uint16_t *haystack, const uint16_t *needle);
+extern uint32_t *
+ u32_strstr (const uint32_t *haystack, const uint32_t *needle);
+
+/* Test whether STR starts with PREFIX. */
+extern bool
+ u8_startswith (const uint8_t *str, const uint8_t *prefix);
+extern bool
+ u16_startswith (const uint16_t *str, const uint16_t *prefix);
+extern bool
+ u32_startswith (const uint32_t *str, const uint32_t *prefix);
+
+/* Test whether STR ends with SUFFIX. */
+extern bool
+ u8_endswith (const uint8_t *str, const uint8_t *suffix);
+extern bool
+ u16_endswith (const uint16_t *str, const uint16_t *suffix);
+extern bool
+ u32_endswith (const uint32_t *str, const uint32_t *suffix);
+
+/* Divide STR into tokens separated by characters in DELIM.
+ This interface is actually more similar to wcstok than to strtok. */
+/* Similar to strtok_r(), wcstok(). */
+extern uint8_t *
+ u8_strtok (uint8_t *str, const uint8_t *delim, uint8_t **ptr);
+extern uint16_t *
+ u16_strtok (uint16_t *str, const uint16_t *delim, uint16_t **ptr);
+extern uint32_t *
+ u32_strtok (uint32_t *str, const uint32_t *delim, uint32_t **ptr);
+
+
+#ifdef __cplusplus
+}
+#endif
+
+#endif /* _UNISTR_H */
diff --git a/lib/unistr/u8-mbtouc-aux.c b/lib/unistr/u8-mbtouc-aux.c
new file mode 100644
index 0000000..53d02bf
--- /dev/null
+++ b/lib/unistr/u8-mbtouc-aux.c
@@ -0,0 +1,158 @@
+/* Conversion UTF-8 to UCS-4.
+ Copyright (C) 2001-2002, 2006-2007, 2009 Free Software Foundation, Inc.
+ Written by Bruno Haible <address@hidden>, 2001.
+
+ This program is free software: you can redistribute it and/or modify it
+ under the terms of the GNU Lesser General Public License as published
+ by the Free Software Foundation; either version 3 of the License, or
+ (at your option) any later version.
+
+ This program is distributed in the hope that it will be useful,
+ but WITHOUT ANY WARRANTY; without even the implied warranty of
+ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+ Lesser General Public License for more details.
+
+ You should have received a copy of the GNU Lesser General Public License
+ along with this program. If not, see <http://www.gnu.org/licenses/>. */
+
+#include <config.h>
+
+/* Specification. */
+#include "unistr.h"
+
+#if defined IN_LIBUNISTRING || HAVE_INLINE
+
+int
+u8_mbtouc_aux (ucs4_t *puc, const uint8_t *s, size_t n)
+{
+ uint8_t c = *s;
+
+ if (c >= 0xc2)
+ {
+ if (c < 0xe0)
+ {
+ if (n >= 2)
+ {
+ if ((s[1] ^ 0x80) < 0x40)
+ {
+ *puc = ((unsigned int) (c & 0x1f) << 6)
+ | (unsigned int) (s[1] ^ 0x80);
+ return 2;
+ }
+ /* invalid multibyte character */
+ }
+ else
+ {
+ /* incomplete multibyte character */
+ *puc = 0xfffd;
+ return n;
+ }
+ }
+ else if (c < 0xf0)
+ {
+ if (n >= 3)
+ {
+ if ((s[1] ^ 0x80) < 0x40 && (s[2] ^ 0x80) < 0x40
+ && (c >= 0xe1 || s[1] >= 0xa0)
+ && (c != 0xed || s[1] < 0xa0))
+ {
+ *puc = ((unsigned int) (c & 0x0f) << 12)
+ | ((unsigned int) (s[1] ^ 0x80) << 6)
+ | (unsigned int) (s[2] ^ 0x80);
+ return 3;
+ }
+ /* invalid multibyte character */
+ }
+ else
+ {
+ /* incomplete multibyte character */
+ *puc = 0xfffd;
+ return n;
+ }
+ }
+ else if (c < 0xf8)
+ {
+ if (n >= 4)
+ {
+ if ((s[1] ^ 0x80) < 0x40 && (s[2] ^ 0x80) < 0x40
+ && (s[3] ^ 0x80) < 0x40
+ && (c >= 0xf1 || s[1] >= 0x90)
+#if 1
+ && (c < 0xf4 || (c == 0xf4 && s[1] < 0x90))
+#endif
+ )
+ {
+ *puc = ((unsigned int) (c & 0x07) << 18)
+ | ((unsigned int) (s[1] ^ 0x80) << 12)
+ | ((unsigned int) (s[2] ^ 0x80) << 6)
+ | (unsigned int) (s[3] ^ 0x80);
+ return 4;
+ }
+ /* invalid multibyte character */
+ }
+ else
+ {
+ /* incomplete multibyte character */
+ *puc = 0xfffd;
+ return n;
+ }
+ }
+#if 0
+ else if (c < 0xfc)
+ {
+ if (n >= 5)
+ {
+ if ((s[1] ^ 0x80) < 0x40 && (s[2] ^ 0x80) < 0x40
+ && (s[3] ^ 0x80) < 0x40 && (s[4] ^ 0x80) < 0x40
+ && (c >= 0xf9 || s[1] >= 0x88))
+ {
+ *puc = ((unsigned int) (c & 0x03) << 24)
+ | ((unsigned int) (s[1] ^ 0x80) << 18)
+ | ((unsigned int) (s[2] ^ 0x80) << 12)
+ | ((unsigned int) (s[3] ^ 0x80) << 6)
+ | (unsigned int) (s[4] ^ 0x80);
+ return 5;
+ }
+ /* invalid multibyte character */
+ }
+ else
+ {
+ /* incomplete multibyte character */
+ *puc = 0xfffd;
+ return n;
+ }
+ }
+ else if (c < 0xfe)
+ {
+ if (n >= 6)
+ {
+ if ((s[1] ^ 0x80) < 0x40 && (s[2] ^ 0x80) < 0x40
+ && (s[3] ^ 0x80) < 0x40 && (s[4] ^ 0x80) < 0x40
+ && (s[5] ^ 0x80) < 0x40
+ && (c >= 0xfd || s[1] >= 0x84))
+ {
+ *puc = ((unsigned int) (c & 0x01) << 30)
+ | ((unsigned int) (s[1] ^ 0x80) << 24)
+ | ((unsigned int) (s[2] ^ 0x80) << 18)
+ | ((unsigned int) (s[3] ^ 0x80) << 12)
+ | ((unsigned int) (s[4] ^ 0x80) << 6)
+ | (unsigned int) (s[5] ^ 0x80);
+ return 6;
+ }
+ /* invalid multibyte character */
+ }
+ else
+ {
+ /* incomplete multibyte character */
+ *puc = 0xfffd;
+ return n;
+ }
+ }
+#endif
+ }
+ /* invalid multibyte character */
+ *puc = 0xfffd;
+ return 1;
+}
+
+#endif
diff --git a/lib/unistr/u8-mbtouc-unsafe-aux.c
b/lib/unistr/u8-mbtouc-unsafe-aux.c
new file mode 100644
index 0000000..43e4a36
--- /dev/null
+++ b/lib/unistr/u8-mbtouc-unsafe-aux.c
@@ -0,0 +1,168 @@
+/* Conversion UTF-8 to UCS-4.
+ Copyright (C) 2001-2002, 2006-2007, 2009 Free Software Foundation, Inc.
+ Written by Bruno Haible <address@hidden>, 2001.
+
+ This program is free software: you can redistribute it and/or modify it
+ under the terms of the GNU Lesser General Public License as published
+ by the Free Software Foundation; either version 3 of the License, or
+ (at your option) any later version.
+
+ This program is distributed in the hope that it will be useful,
+ but WITHOUT ANY WARRANTY; without even the implied warranty of
+ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+ Lesser General Public License for more details.
+
+ You should have received a copy of the GNU Lesser General Public License
+ along with this program. If not, see <http://www.gnu.org/licenses/>. */
+
+#include <config.h>
+
+/* Specification. */
+#include "unistr.h"
+
+#if defined IN_LIBUNISTRING || HAVE_INLINE
+
+int
+u8_mbtouc_unsafe_aux (ucs4_t *puc, const uint8_t *s, size_t n)
+{
+ uint8_t c = *s;
+
+ if (c >= 0xc2)
+ {
+ if (c < 0xe0)
+ {
+ if (n >= 2)
+ {
+#if CONFIG_UNICODE_SAFETY
+ if ((s[1] ^ 0x80) < 0x40)
+#endif
+ {
+ *puc = ((unsigned int) (c & 0x1f) << 6)
+ | (unsigned int) (s[1] ^ 0x80);
+ return 2;
+ }
+ /* invalid multibyte character */
+ }
+ else
+ {
+ /* incomplete multibyte character */
+ *puc = 0xfffd;
+ return n;
+ }
+ }
+ else if (c < 0xf0)
+ {
+ if (n >= 3)
+ {
+#if CONFIG_UNICODE_SAFETY
+ if ((s[1] ^ 0x80) < 0x40 && (s[2] ^ 0x80) < 0x40
+ && (c >= 0xe1 || s[1] >= 0xa0)
+ && (c != 0xed || s[1] < 0xa0))
+#endif
+ {
+ *puc = ((unsigned int) (c & 0x0f) << 12)
+ | ((unsigned int) (s[1] ^ 0x80) << 6)
+ | (unsigned int) (s[2] ^ 0x80);
+ return 3;
+ }
+ /* invalid multibyte character */
+ }
+ else
+ {
+ /* incomplete multibyte character */
+ *puc = 0xfffd;
+ return n;
+ }
+ }
+ else if (c < 0xf8)
+ {
+ if (n >= 4)
+ {
+#if CONFIG_UNICODE_SAFETY
+ if ((s[1] ^ 0x80) < 0x40 && (s[2] ^ 0x80) < 0x40
+ && (s[3] ^ 0x80) < 0x40
+ && (c >= 0xf1 || s[1] >= 0x90)
+#if 1
+ && (c < 0xf4 || (c == 0xf4 && s[1] < 0x90))
+#endif
+ )
+#endif
+ {
+ *puc = ((unsigned int) (c & 0x07) << 18)
+ | ((unsigned int) (s[1] ^ 0x80) << 12)
+ | ((unsigned int) (s[2] ^ 0x80) << 6)
+ | (unsigned int) (s[3] ^ 0x80);
+ return 4;
+ }
+ /* invalid multibyte character */
+ }
+ else
+ {
+ /* incomplete multibyte character */
+ *puc = 0xfffd;
+ return n;
+ }
+ }
+#if 0
+ else if (c < 0xfc)
+ {
+ if (n >= 5)
+ {
+#if CONFIG_UNICODE_SAFETY
+ if ((s[1] ^ 0x80) < 0x40 && (s[2] ^ 0x80) < 0x40
+ && (s[3] ^ 0x80) < 0x40 && (s[4] ^ 0x80) < 0x40
+ && (c >= 0xf9 || s[1] >= 0x88))
+#endif
+ {
+ *puc = ((unsigned int) (c & 0x03) << 24)
+ | ((unsigned int) (s[1] ^ 0x80) << 18)
+ | ((unsigned int) (s[2] ^ 0x80) << 12)
+ | ((unsigned int) (s[3] ^ 0x80) << 6)
+ | (unsigned int) (s[4] ^ 0x80);
+ return 5;
+ }
+ /* invalid multibyte character */
+ }
+ else
+ {
+ /* incomplete multibyte character */
+ *puc = 0xfffd;
+ return n;
+ }
+ }
+ else if (c < 0xfe)
+ {
+ if (n >= 6)
+ {
+#if CONFIG_UNICODE_SAFETY
+ if ((s[1] ^ 0x80) < 0x40 && (s[2] ^ 0x80) < 0x40
+ && (s[3] ^ 0x80) < 0x40 && (s[4] ^ 0x80) < 0x40
+ && (s[5] ^ 0x80) < 0x40
+ && (c >= 0xfd || s[1] >= 0x84))
+#endif
+ {
+ *puc = ((unsigned int) (c & 0x01) << 30)
+ | ((unsigned int) (s[1] ^ 0x80) << 24)
+ | ((unsigned int) (s[2] ^ 0x80) << 18)
+ | ((unsigned int) (s[3] ^ 0x80) << 12)
+ | ((unsigned int) (s[4] ^ 0x80) << 6)
+ | (unsigned int) (s[5] ^ 0x80);
+ return 6;
+ }
+ /* invalid multibyte character */
+ }
+ else
+ {
+ /* incomplete multibyte character */
+ *puc = 0xfffd;
+ return n;
+ }
+ }
+#endif
+ }
+ /* invalid multibyte character */
+ *puc = 0xfffd;
+ return 1;
+}
+
+#endif
diff --git a/lib/unistr/u8-mbtouc-unsafe.c b/lib/unistr/u8-mbtouc-unsafe.c
new file mode 100644
index 0000000..4661569
--- /dev/null
+++ b/lib/unistr/u8-mbtouc-unsafe.c
@@ -0,0 +1,179 @@
+/* Look at first character in UTF-8 string.
+ Copyright (C) 1999-2002, 2006-2007, 2009 Free Software Foundation, Inc.
+ Written by Bruno Haible <address@hidden>, 2001.
+
+ This program is free software: you can redistribute it and/or modify it
+ under the terms of the GNU Lesser General Public License as published
+ by the Free Software Foundation; either version 3 of the License, or
+ (at your option) any later version.
+
+ This program is distributed in the hope that it will be useful,
+ but WITHOUT ANY WARRANTY; without even the implied warranty of
+ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+ Lesser General Public License for more details.
+
+ You should have received a copy of the GNU Lesser General Public License
+ along with this program. If not, see <http://www.gnu.org/licenses/>. */
+
+#include <config.h>
+
+#if defined IN_LIBUNISTRING
+/* Tell unistr.h to declare u8_mbtouc_unsafe as 'extern', not
+ 'static inline'. */
+# include "unistring-notinline.h"
+#endif
+
+/* Specification. */
+#include "unistr.h"
+
+#if !HAVE_INLINE
+
+int
+u8_mbtouc_unsafe (ucs4_t *puc, const uint8_t *s, size_t n)
+{
+ uint8_t c = *s;
+
+ if (c < 0x80)
+ {
+ *puc = c;
+ return 1;
+ }
+ else if (c >= 0xc2)
+ {
+ if (c < 0xe0)
+ {
+ if (n >= 2)
+ {
+#if CONFIG_UNICODE_SAFETY
+ if ((s[1] ^ 0x80) < 0x40)
+#endif
+ {
+ *puc = ((unsigned int) (c & 0x1f) << 6)
+ | (unsigned int) (s[1] ^ 0x80);
+ return 2;
+ }
+ /* invalid multibyte character */
+ }
+ else
+ {
+ /* incomplete multibyte character */
+ *puc = 0xfffd;
+ return n;
+ }
+ }
+ else if (c < 0xf0)
+ {
+ if (n >= 3)
+ {
+#if CONFIG_UNICODE_SAFETY
+ if ((s[1] ^ 0x80) < 0x40 && (s[2] ^ 0x80) < 0x40
+ && (c >= 0xe1 || s[1] >= 0xa0)
+ && (c != 0xed || s[1] < 0xa0))
+#endif
+ {
+ *puc = ((unsigned int) (c & 0x0f) << 12)
+ | ((unsigned int) (s[1] ^ 0x80) << 6)
+ | (unsigned int) (s[2] ^ 0x80);
+ return 3;
+ }
+ /* invalid multibyte character */
+ }
+ else
+ {
+ /* incomplete multibyte character */
+ *puc = 0xfffd;
+ return n;
+ }
+ }
+ else if (c < 0xf8)
+ {
+ if (n >= 4)
+ {
+#if CONFIG_UNICODE_SAFETY
+ if ((s[1] ^ 0x80) < 0x40 && (s[2] ^ 0x80) < 0x40
+ && (s[3] ^ 0x80) < 0x40
+ && (c >= 0xf1 || s[1] >= 0x90)
+#if 1
+ && (c < 0xf4 || (c == 0xf4 && s[1] < 0x90))
+#endif
+ )
+#endif
+ {
+ *puc = ((unsigned int) (c & 0x07) << 18)
+ | ((unsigned int) (s[1] ^ 0x80) << 12)
+ | ((unsigned int) (s[2] ^ 0x80) << 6)
+ | (unsigned int) (s[3] ^ 0x80);
+ return 4;
+ }
+ /* invalid multibyte character */
+ }
+ else
+ {
+ /* incomplete multibyte character */
+ *puc = 0xfffd;
+ return n;
+ }
+ }
+#if 0
+ else if (c < 0xfc)
+ {
+ if (n >= 5)
+ {
+#if CONFIG_UNICODE_SAFETY
+ if ((s[1] ^ 0x80) < 0x40 && (s[2] ^ 0x80) < 0x40
+ && (s[3] ^ 0x80) < 0x40 && (s[4] ^ 0x80) < 0x40
+ && (c >= 0xf9 || s[1] >= 0x88))
+#endif
+ {
+ *puc = ((unsigned int) (c & 0x03) << 24)
+ | ((unsigned int) (s[1] ^ 0x80) << 18)
+ | ((unsigned int) (s[2] ^ 0x80) << 12)
+ | ((unsigned int) (s[3] ^ 0x80) << 6)
+ | (unsigned int) (s[4] ^ 0x80);
+ return 5;
+ }
+ /* invalid multibyte character */
+ }
+ else
+ {
+ /* incomplete multibyte character */
+ *puc = 0xfffd;
+ return n;
+ }
+ }
+ else if (c < 0xfe)
+ {
+ if (n >= 6)
+ {
+#if CONFIG_UNICODE_SAFETY
+ if ((s[1] ^ 0x80) < 0x40 && (s[2] ^ 0x80) < 0x40
+ && (s[3] ^ 0x80) < 0x40 && (s[4] ^ 0x80) < 0x40
+ && (s[5] ^ 0x80) < 0x40
+ && (c >= 0xfd || s[1] >= 0x84))
+#endif
+ {
+ *puc = ((unsigned int) (c & 0x01) << 30)
+ | ((unsigned int) (s[1] ^ 0x80) << 24)
+ | ((unsigned int) (s[2] ^ 0x80) << 18)
+ | ((unsigned int) (s[3] ^ 0x80) << 12)
+ | ((unsigned int) (s[4] ^ 0x80) << 6)
+ | (unsigned int) (s[5] ^ 0x80);
+ return 6;
+ }
+ /* invalid multibyte character */
+ }
+ else
+ {
+ /* incomplete multibyte character */
+ *puc = 0xfffd;
+ return n;
+ }
+ }
+#endif
+ }
+ /* invalid multibyte character */
+ *puc = 0xfffd;
+ return 1;
+}
+
+#endif
diff --git a/lib/unistr/u8-mbtouc.c b/lib/unistr/u8-mbtouc.c
new file mode 100644
index 0000000..ff624f1
--- /dev/null
+++ b/lib/unistr/u8-mbtouc.c
@@ -0,0 +1,168 @@
+/* Look at first character in UTF-8 string.
+ Copyright (C) 1999-2002, 2006-2007, 2009 Free Software Foundation, Inc.
+ Written by Bruno Haible <address@hidden>, 2001.
+
+ This program is free software: you can redistribute it and/or modify it
+ under the terms of the GNU Lesser General Public License as published
+ by the Free Software Foundation; either version 3 of the License, or
+ (at your option) any later version.
+
+ This program is distributed in the hope that it will be useful,
+ but WITHOUT ANY WARRANTY; without even the implied warranty of
+ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+ Lesser General Public License for more details.
+
+ You should have received a copy of the GNU Lesser General Public License
+ along with this program. If not, see <http://www.gnu.org/licenses/>. */
+
+#include <config.h>
+
+#if defined IN_LIBUNISTRING
+/* Tell unistr.h to declare u8_mbtouc as 'extern', not 'static inline'. */
+# include "unistring-notinline.h"
+#endif
+
+/* Specification. */
+#include "unistr.h"
+
+#if !HAVE_INLINE
+
+int
+u8_mbtouc (ucs4_t *puc, const uint8_t *s, size_t n)
+{
+ uint8_t c = *s;
+
+ if (c < 0x80)
+ {
+ *puc = c;
+ return 1;
+ }
+ else if (c >= 0xc2)
+ {
+ if (c < 0xe0)
+ {
+ if (n >= 2)
+ {
+ if ((s[1] ^ 0x80) < 0x40)
+ {
+ *puc = ((unsigned int) (c & 0x1f) << 6)
+ | (unsigned int) (s[1] ^ 0x80);
+ return 2;
+ }
+ /* invalid multibyte character */
+ }
+ else
+ {
+ /* incomplete multibyte character */
+ *puc = 0xfffd;
+ return n;
+ }
+ }
+ else if (c < 0xf0)
+ {
+ if (n >= 3)
+ {
+ if ((s[1] ^ 0x80) < 0x40 && (s[2] ^ 0x80) < 0x40
+ && (c >= 0xe1 || s[1] >= 0xa0)
+ && (c != 0xed || s[1] < 0xa0))
+ {
+ *puc = ((unsigned int) (c & 0x0f) << 12)
+ | ((unsigned int) (s[1] ^ 0x80) << 6)
+ | (unsigned int) (s[2] ^ 0x80);
+ return 3;
+ }
+ /* invalid multibyte character */
+ }
+ else
+ {
+ /* incomplete multibyte character */
+ *puc = 0xfffd;
+ return n;
+ }
+ }
+ else if (c < 0xf8)
+ {
+ if (n >= 4)
+ {
+ if ((s[1] ^ 0x80) < 0x40 && (s[2] ^ 0x80) < 0x40
+ && (s[3] ^ 0x80) < 0x40
+ && (c >= 0xf1 || s[1] >= 0x90)
+#if 1
+ && (c < 0xf4 || (c == 0xf4 && s[1] < 0x90))
+#endif
+ )
+ {
+ *puc = ((unsigned int) (c & 0x07) << 18)
+ | ((unsigned int) (s[1] ^ 0x80) << 12)
+ | ((unsigned int) (s[2] ^ 0x80) << 6)
+ | (unsigned int) (s[3] ^ 0x80);
+ return 4;
+ }
+ /* invalid multibyte character */
+ }
+ else
+ {
+ /* incomplete multibyte character */
+ *puc = 0xfffd;
+ return n;
+ }
+ }
+#if 0
+ else if (c < 0xfc)
+ {
+ if (n >= 5)
+ {
+ if ((s[1] ^ 0x80) < 0x40 && (s[2] ^ 0x80) < 0x40
+ && (s[3] ^ 0x80) < 0x40 && (s[4] ^ 0x80) < 0x40
+ && (c >= 0xf9 || s[1] >= 0x88))
+ {
+ *puc = ((unsigned int) (c & 0x03) << 24)
+ | ((unsigned int) (s[1] ^ 0x80) << 18)
+ | ((unsigned int) (s[2] ^ 0x80) << 12)
+ | ((unsigned int) (s[3] ^ 0x80) << 6)
+ | (unsigned int) (s[4] ^ 0x80);
+ return 5;
+ }
+ /* invalid multibyte character */
+ }
+ else
+ {
+ /* incomplete multibyte character */
+ *puc = 0xfffd;
+ return n;
+ }
+ }
+ else if (c < 0xfe)
+ {
+ if (n >= 6)
+ {
+ if ((s[1] ^ 0x80) < 0x40 && (s[2] ^ 0x80) < 0x40
+ && (s[3] ^ 0x80) < 0x40 && (s[4] ^ 0x80) < 0x40
+ && (s[5] ^ 0x80) < 0x40
+ && (c >= 0xfd || s[1] >= 0x84))
+ {
+ *puc = ((unsigned int) (c & 0x01) << 30)
+ | ((unsigned int) (s[1] ^ 0x80) << 24)
+ | ((unsigned int) (s[2] ^ 0x80) << 18)
+ | ((unsigned int) (s[3] ^ 0x80) << 12)
+ | ((unsigned int) (s[4] ^ 0x80) << 6)
+ | (unsigned int) (s[5] ^ 0x80);
+ return 6;
+ }
+ /* invalid multibyte character */
+ }
+ else
+ {
+ /* incomplete multibyte character */
+ *puc = 0xfffd;
+ return n;
+ }
+ }
+#endif
+ }
+ /* invalid multibyte character */
+ *puc = 0xfffd;
+ return 1;
+}
+
+#endif
diff --git a/lib/unistr/u8-mbtoucr.c b/lib/unistr/u8-mbtoucr.c
new file mode 100644
index 0000000..dd83352
--- /dev/null
+++ b/lib/unistr/u8-mbtoucr.c
@@ -0,0 +1,285 @@
+/* Look at first character in UTF-8 string, returning an error code.
+ Copyright (C) 1999-2002, 2006-2007 Free Software Foundation, Inc.
+ Written by Bruno Haible <address@hidden>, 2001.
+
+ This program is free software: you can redistribute it and/or modify it
+ under the terms of the GNU Lesser General Public License as published
+ by the Free Software Foundation; either version 3 of the License, or
+ (at your option) any later version.
+
+ This program is distributed in the hope that it will be useful,
+ but WITHOUT ANY WARRANTY; without even the implied warranty of
+ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+ Lesser General Public License for more details.
+
+ You should have received a copy of the GNU Lesser General Public License
+ along with this program. If not, see <http://www.gnu.org/licenses/>. */
+
+#include <config.h>
+
+/* Specification. */
+#include "unistr.h"
+
+int
+u8_mbtoucr (ucs4_t *puc, const uint8_t *s, size_t n)
+{
+ uint8_t c = *s;
+
+ if (c < 0x80)
+ {
+ *puc = c;
+ return 1;
+ }
+ else if (c >= 0xc2)
+ {
+ if (c < 0xe0)
+ {
+ if (n >= 2)
+ {
+ if ((s[1] ^ 0x80) < 0x40)
+ {
+ *puc = ((unsigned int) (c & 0x1f) << 6)
+ | (unsigned int) (s[1] ^ 0x80);
+ return 2;
+ }
+ /* invalid multibyte character */
+ }
+ else
+ {
+ /* incomplete multibyte character */
+ *puc = 0xfffd;
+ return -2;
+ }
+ }
+ else if (c < 0xf0)
+ {
+ if (n >= 2)
+ {
+ if ((s[1] ^ 0x80) < 0x40
+ && (c >= 0xe1 || s[1] >= 0xa0)
+ && (c != 0xed || s[1] < 0xa0))
+ {
+ if (n >= 3)
+ {
+ if ((s[2] ^ 0x80) < 0x40)
+ {
+ *puc = ((unsigned int) (c & 0x0f) << 12)
+ | ((unsigned int) (s[1] ^ 0x80) << 6)
+ | (unsigned int) (s[2] ^ 0x80);
+ return 3;
+ }
+ /* invalid multibyte character */
+ }
+ else
+ {
+ /* incomplete multibyte character */
+ *puc = 0xfffd;
+ return -2;
+ }
+ }
+ /* invalid multibyte character */
+ }
+ else
+ {
+ /* incomplete multibyte character */
+ *puc = 0xfffd;
+ return -2;
+ }
+ }
+ else if (c < 0xf8)
+ {
+ if (n >= 2)
+ {
+ if ((s[1] ^ 0x80) < 0x40
+ && (c >= 0xf1 || s[1] >= 0x90)
+#if 1
+ && (c < 0xf4 || (c == 0xf4 && s[1] < 0x90))
+#endif
+ )
+ {
+ if (n >= 3)
+ {
+ if ((s[2] ^ 0x80) < 0x40)
+ {
+ if (n >= 4)
+ {
+ if ((s[3] ^ 0x80) < 0x40)
+ {
+ *puc = ((unsigned int) (c & 0x07) << 18)
+ | ((unsigned int) (s[1] ^ 0x80) << 12)
+ | ((unsigned int) (s[2] ^ 0x80) << 6)
+ | (unsigned int) (s[3] ^ 0x80);
+ return 4;
+ }
+ /* invalid multibyte character */
+ }
+ else
+ {
+ /* incomplete multibyte character */
+ *puc = 0xfffd;
+ return -2;
+ }
+ }
+ /* invalid multibyte character */
+ }
+ else
+ {
+ /* incomplete multibyte character */
+ *puc = 0xfffd;
+ return -2;
+ }
+ }
+ /* invalid multibyte character */
+ }
+ else
+ {
+ /* incomplete multibyte character */
+ *puc = 0xfffd;
+ return -2;
+ }
+ }
+#if 0
+ else if (c < 0xfc)
+ {
+ if (n >= 2)
+ {
+ if ((s[1] ^ 0x80) < 0x40
+ && (c >= 0xf9 || s[1] >= 0x88))
+ {
+ if (n >= 3)
+ {
+ if ((s[2] ^ 0x80) < 0x40)
+ {
+ if (n >= 4)
+ {
+ if ((s[3] ^ 0x80) < 0x40)
+ {
+ if (n >= 5)
+ {
+ if ((s[4] ^ 0x80) < 0x40)
+ {
+ *puc = ((unsigned int) (c & 0x03) <<
24)
+ | ((unsigned int) (s[1] ^
0x80) << 18)
+ | ((unsigned int) (s[2] ^
0x80) << 12)
+ | ((unsigned int) (s[3] ^
0x80) << 6)
+ | (unsigned int) (s[4] ^ 0x80);
+ return 5;
+ }
+ /* invalid multibyte character */
+ }
+ else
+ {
+ /* incomplete multibyte character */
+ *puc = 0xfffd;
+ return -2;
+ }
+ }
+ /* invalid multibyte character */
+ }
+ else
+ {
+ /* incomplete multibyte character */
+ *puc = 0xfffd;
+ return -2;
+ }
+ }
+ /* invalid multibyte character */
+ }
+ else
+ {
+ /* incomplete multibyte character */
+ *puc = 0xfffd;
+ return -2;
+ }
+ }
+ /* invalid multibyte character */
+ }
+ else
+ {
+ /* incomplete multibyte character */
+ *puc = 0xfffd;
+ return -2;
+ }
+ }
+ else if (c < 0xfe)
+ {
+ if (n >= 2)
+ {
+ if ((s[1] ^ 0x80) < 0x40
+ && (c >= 0xfd || s[1] >= 0x84))
+ {
+ if (n >= 3)
+ {
+ if ((s[2] ^ 0x80) < 0x40)
+ {
+ if (n >= 4)
+ {
+ if ((s[3] ^ 0x80) < 0x40)
+ {
+ if (n >= 5)
+ {
+ if ((s[4] ^ 0x80) < 0x40)
+ {
+ if (n >= 6)
+ {
+ if ((s[5] ^ 0x80) < 0x40)
+ {
+ *puc = ((unsigned int) (c &
0x01) << 30)
+ | ((unsigned int)
(s[1] ^ 0x80) << 24)
+ | ((unsigned int)
(s[2] ^ 0x80) << 18)
+ | ((unsigned int)
(s[3] ^ 0x80) << 12)
+ | ((unsigned int)
(s[4] ^ 0x80) << 6)
+ | (unsigned int) (s[5]
^ 0x80);
+ return 6;
+ }
+ /* invalid multibyte character */
+ }
+ else
+ {
+ /* incomplete multibyte character
*/
+ *puc = 0xfffd;
+ return -2;
+ }
+ }
+ /* invalid multibyte character */
+ }
+ else
+ {
+ /* incomplete multibyte character */
+ *puc = 0xfffd;
+ return -2;
+ }
+ }
+ /* invalid multibyte character */
+ }
+ else
+ {
+ /* incomplete multibyte character */
+ *puc = 0xfffd;
+ return -2;
+ }
+ }
+ /* invalid multibyte character */
+ }
+ else
+ {
+ /* incomplete multibyte character */
+ *puc = 0xfffd;
+ return -2;
+ }
+ }
+ /* invalid multibyte character */
+ }
+ else
+ {
+ /* incomplete multibyte character */
+ *puc = 0xfffd;
+ return -2;
+ }
+ }
+#endif
+ }
+ /* invalid multibyte character */
+ *puc = 0xfffd;
+ return -1;
+}
diff --git a/lib/unistr/u8-prev.c b/lib/unistr/u8-prev.c
new file mode 100644
index 0000000..245d22f
--- /dev/null
+++ b/lib/unistr/u8-prev.c
@@ -0,0 +1,93 @@
+/* Iterate over previous character in UTF-8 string.
+ Copyright (C) 2002, 2006-2007 Free Software Foundation, Inc.
+ Written by Bruno Haible <address@hidden>, 2002.
+
+ This program is free software: you can redistribute it and/or modify it
+ under the terms of the GNU Lesser General Public License as published
+ by the Free Software Foundation; either version 3 of the License, or
+ (at your option) any later version.
+
+ This program is distributed in the hope that it will be useful,
+ but WITHOUT ANY WARRANTY; without even the implied warranty of
+ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+ Lesser General Public License for more details.
+
+ You should have received a copy of the GNU Lesser General Public License
+ along with this program. If not, see <http://www.gnu.org/licenses/>. */
+
+#include <config.h>
+
+/* Specification. */
+#include "unistr.h"
+
+const uint8_t *
+u8_prev (ucs4_t *puc, const uint8_t *s, const uint8_t *start)
+{
+ /* Keep in sync with unistr.h and utf8-ucs4.c. */
+ if (s != start)
+ {
+ uint8_t c_1 = s[-1];
+
+ if (c_1 < 0x80)
+ {
+ *puc = c_1;
+ return s - 1;
+ }
+#if CONFIG_UNICODE_SAFETY
+ if ((c_1 ^ 0x80) < 0x40)
+#endif
+ if (s - 1 != start)
+ {
+ uint8_t c_2 = s[-2];
+
+ if (c_2 >= 0xc2 && c_2 < 0xe0)
+ {
+ *puc = ((unsigned int) (c_2 & 0x1f) << 6)
+ | (unsigned int) (c_1 ^ 0x80);
+ return s - 2;
+ }
+#if CONFIG_UNICODE_SAFETY
+ if ((c_2 ^ 0x80) < 0x40)
+#endif
+ if (s - 2 != start)
+ {
+ uint8_t c_3 = s[-3];
+
+ if (c_3 >= 0xe0 && c_3 < 0xf0
+#if CONFIG_UNICODE_SAFETY
+ && (c_3 >= 0xe1 || c_2 >= 0xa0)
+ && (c_3 != 0xed || c_2 < 0xa0)
+#endif
+ )
+ {
+ *puc = ((unsigned int) (c_3 & 0x0f) << 12)
+ | ((unsigned int) (c_2 ^ 0x80) << 6)
+ | (unsigned int) (c_1 ^ 0x80);
+ return s - 3;
+ }
+#if CONFIG_UNICODE_SAFETY
+ if ((c_3 ^ 0x80) < 0x40)
+#endif
+ if (s - 3 != start)
+ {
+ uint8_t c_4 = s[-4];
+
+ if (c_4 >= 0xf0 && c_4 < 0xf8
+#if CONFIG_UNICODE_SAFETY
+ && (c_4 >= 0xf1 || c_3 >= 0x90)
+ && (c_4 < 0xf4 || (c_4 == 0xf4 && c_3 < 0x90))
+#endif
+ )
+ {
+ *puc = ((unsigned int) (c_4 & 0x07) << 18)
+ | ((unsigned int) (c_3 ^ 0x80) << 12)
+ | ((unsigned int) (c_2 ^ 0x80) << 6)
+ | (unsigned int) (c_1 ^ 0x80);
+ return s - 4;
+ }
+ }
+ }
+ }
+ }
+ return NULL;
+}
diff --git a/lib/unistr/u8-uctomb-aux.c b/lib/unistr/u8-uctomb-aux.c
new file mode 100644
index 0000000..c42fa50
--- /dev/null
+++ b/lib/unistr/u8-uctomb-aux.c
@@ -0,0 +1,69 @@
+/* Conversion UCS-4 to UTF-8.
+ Copyright (C) 2002, 2006-2007 Free Software Foundation, Inc.
+ Written by Bruno Haible <address@hidden>, 2002.
+
+ This program is free software: you can redistribute it and/or modify it
+ under the terms of the GNU Lesser General Public License as published
+ by the Free Software Foundation; either version 3 of the License, or
+ (at your option) any later version.
+
+ This program is distributed in the hope that it will be useful,
+ but WITHOUT ANY WARRANTY; without even the implied warranty of
+ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+ Lesser General Public License for more details.
+
+ You should have received a copy of the GNU Lesser General Public License
+ along with this program. If not, see <http://www.gnu.org/licenses/>. */
+
+#include <config.h>
+
+/* Specification. */
+#include "unistr.h"
+
+int
+u8_uctomb_aux (uint8_t *s, ucs4_t uc, int n)
+{
+ int count;
+
+ if (uc < 0x80)
+ /* The case n >= 1 is already handled by the caller. */
+ return -2;
+ else if (uc < 0x800)
+ count = 2;
+ else if (uc < 0x10000)
+ {
+ if (uc < 0xd800 || uc >= 0xe000)
+ count = 3;
+ else
+ return -1;
+ }
+#if 0
+ else if (uc < 0x200000)
+ count = 4;
+ else if (uc < 0x4000000)
+ count = 5;
+ else if (uc <= 0x7fffffff)
+ count = 6;
+#else
+ else if (uc < 0x110000)
+ count = 4;
+#endif
+ else
+ return -1;
+
+ if (n < count)
+ return -2;
+
+ switch (count) /* note: code falls through cases! */
+ {
+#if 0
+ case 6: s[5] = 0x80 | (uc & 0x3f); uc = uc >> 6; uc |= 0x4000000;
+ case 5: s[4] = 0x80 | (uc & 0x3f); uc = uc >> 6; uc |= 0x200000;
+#endif
+ case 4: s[3] = 0x80 | (uc & 0x3f); uc = uc >> 6; uc |= 0x10000;
+ case 3: s[2] = 0x80 | (uc & 0x3f); uc = uc >> 6; uc |= 0x800;
+ case 2: s[1] = 0x80 | (uc & 0x3f); uc = uc >> 6; uc |= 0xc0;
+ /*case 1:*/ s[0] = uc;
+ }
+ return count;
+}
diff --git a/lib/unistr/u8-uctomb.c b/lib/unistr/u8-uctomb.c
new file mode 100644
index 0000000..3392166
--- /dev/null
+++ b/lib/unistr/u8-uctomb.c
@@ -0,0 +1,88 @@
+/* Store a character in UTF-8 string.
+ Copyright (C) 2002, 2005-2006, 2009 Free Software Foundation, Inc.
+ Written by Bruno Haible <address@hidden>, 2002.
+
+ This program is free software: you can redistribute it and/or modify it
+ under the terms of the GNU Lesser General Public License as published
+ by the Free Software Foundation; either version 3 of the License, or
+ (at your option) any later version.
+
+ This program is distributed in the hope that it will be useful,
+ but WITHOUT ANY WARRANTY; without even the implied warranty of
+ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+ Lesser General Public License for more details.
+
+ You should have received a copy of the GNU Lesser General Public License
+ along with this program. If not, see <http://www.gnu.org/licenses/>. */
+
+#include <config.h>
+
+#if defined IN_LIBUNISTRING
+/* Tell unistr.h to declare u8_uctomb as 'extern', not 'static inline'. */
+# include "unistring-notinline.h"
+#endif
+
+/* Specification. */
+#include "unistr.h"
+
+#if !HAVE_INLINE
+
+int
+u8_uctomb (uint8_t *s, ucs4_t uc, int n)
+{
+ if (uc < 0x80)
+ {
+ if (n > 0)
+ {
+ s[0] = uc;
+ return 1;
+ }
+ /* else return -2, below. */
+ }
+ else
+ {
+ int count;
+
+ if (uc < 0x800)
+ count = 2;
+ else if (uc < 0x10000)
+ {
+ if (uc < 0xd800 || uc >= 0xe000)
+ count = 3;
+ else
+ return -1;
+ }
+#if 0
+ else if (uc < 0x200000)
+ count = 4;
+ else if (uc < 0x4000000)
+ count = 5;
+ else if (uc <= 0x7fffffff)
+ count = 6;
+#else
+ else if (uc < 0x110000)
+ count = 4;
+#endif
+ else
+ return -1;
+
+ if (n >= count)
+ {
+ switch (count) /* note: code falls through cases! */
+ {
+#if 0
+ case 6: s[5] = 0x80 | (uc & 0x3f); uc = uc >> 6; uc |= 0x4000000;
+ case 5: s[4] = 0x80 | (uc & 0x3f); uc = uc >> 6; uc |= 0x200000;
+#endif
+ case 4: s[3] = 0x80 | (uc & 0x3f); uc = uc >> 6; uc |= 0x10000;
+ case 3: s[2] = 0x80 | (uc & 0x3f); uc = uc >> 6; uc |= 0x800;
+ case 2: s[1] = 0x80 | (uc & 0x3f); uc = uc >> 6; uc |= 0xc0;
+ /*case 1:*/ s[0] = uc;
+ }
+ return count;
+ }
+ }
+ return -2;
+}
+
+#endif
diff --git a/lib/unitypes.h b/lib/unitypes.h
new file mode 100644
index 0000000..fe8d877
--- /dev/null
+++ b/lib/unitypes.h
@@ -0,0 +1,26 @@
+/* Elementary types for the GNU UniString library.
+ Copyright (C) 2002, 2005-2006 Free Software Foundation, Inc.
+
+ This program is free software: you can redistribute it and/or modify it
+ under the terms of the GNU Lesser General Public License as published
+ by the Free Software Foundation; either version 3 of the License, or
+ (at your option) any later version.
+
+ This program is distributed in the hope that it will be useful,
+ but WITHOUT ANY WARRANTY; without even the implied warranty of
+ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+ Lesser General Public License for more details.
+
+ You should have received a copy of the GNU Lesser General Public License
+ along with this program. If not, see <http://www.gnu.org/licenses/>. */
+
+#ifndef _UNITYPES_H
+#define _UNITYPES_H
+
+/* Get uint8_t, uint16_t, uint32_t. */
+#include <stdint.h>
+
+/* Type representing a Unicode character. */
+typedef uint32_t ucs4_t;
+
+#endif /* _UNITYPES_H */
diff --git a/libguile.h b/libguile.h
index 40122df..6a6d232 100644
--- a/libguile.h
+++ b/libguile.h
@@ -1,7 +1,7 @@
#ifndef SCM_LIBGUILE_H
#define SCM_LIBGUILE_H
-/* Copyright (C) 1995,1996,1997,1998,2000,2001, 2002, 2003, 2004, 2006, 2008
Free Software Foundation, Inc.
+/* Copyright (C) 1995,1996,1997,1998,2000,2001, 2002, 2003, 2004, 2006, 2008,
2009 Free Software Foundation, Inc.
*
* This library is free software; you can redistribute it and/or
* modify it under the terms of the GNU Lesser General Public
@@ -32,6 +32,7 @@ extern "C" {
#include "libguile/arbiters.h"
#include "libguile/async.h"
#include "libguile/boolean.h"
+#include "libguile/bytevectors.h"
#include "libguile/chars.h"
#include "libguile/continuations.h"
#include "libguile/dynl.h"
@@ -75,6 +76,7 @@ extern "C" {
#include "libguile/procprop.h"
#include "libguile/properties.h"
#include "libguile/procs.h"
+#include "libguile/r6rs-ports.h"
#include "libguile/ramap.h"
#include "libguile/random.h"
#include "libguile/read.h"
diff --git a/libguile/Makefile.am b/libguile/Makefile.am
index 63f2ef2..fcf197a 100644
--- a/libguile/Makefile.am
+++ b/libguile/Makefile.am
@@ -106,7 +106,8 @@ guile_LDFLAGS = $(GUILE_CFLAGS)
libguile_la_CFLAGS = $(GUILE_CFLAGS) $(AM_CFLAGS)
libguile_la_SOURCES = alist.c arbiters.c async.c backtrace.c boolean.c \
- chars.c continuations.c convert.c debug.c deprecation.c \
+ bytevectors.c chars.c continuations.c \
+ convert.c debug.c deprecation.c \
deprecated.c discouraged.c dynwind.c eq.c error.c \
eval.c evalext.c extensions.c feature.c fluids.c fports.c \
futures.c gc.c gc-mark.c gc-segment.c gc-malloc.c gc-card.c
\
@@ -115,7 +116,8 @@ libguile_la_SOURCES = alist.c arbiters.c async.c
backtrace.c boolean.c \
guardians.c hash.c hashtab.c hooks.c init.c inline.c \
ioext.c keywords.c lang.c list.c load.c macros.c mallocs.c \
modules.c numbers.c objects.c objprop.c options.c pairs.c ports.c \
- print.c procprop.c procs.c properties.c random.c rdelim.c read.c \
+ print.c procprop.c procs.c properties.c \
+ r6rs-ports.c random.c rdelim.c read.c \
root.c rw.c scmsigs.c script.c simpos.c smob.c sort.c srcprop.c \
stackchk.c stacks.c stime.c strings.c srfi-4.c srfi-13.c srfi-14.c \
strorder.c strports.c struct.c symbols.c threads.c null-threads.c \
@@ -134,7 +136,8 @@ address@hidden@_la_LDFLAGS = \
-module -L$(builddir) -lguile \
-version-info @LIBGUILE_I18N_INTERFACE@
-DOT_X_FILES = alist.x arbiters.x async.x backtrace.x boolean.x chars.x \
+DOT_X_FILES = alist.x arbiters.x async.x backtrace.x boolean.x \
+ bytevectors.x chars.x \
continuations.x debug.x deprecation.x deprecated.x discouraged.x \
dynl.x dynwind.x eq.x error.x eval.x evalext.x \
extensions.x feature.x fluids.x fports.x futures.x gc.x gc-mark.x \
@@ -143,7 +146,8 @@ DOT_X_FILES = alist.x arbiters.x async.x backtrace.x
boolean.x chars.x \
hash.x hashtab.x hooks.x i18n.x init.x ioext.x keywords.x lang.x \
list.x load.x macros.x mallocs.x modules.x numbers.x objects.x \
objprop.x options.x pairs.x ports.x print.x procprop.x procs.x \
- properties.x random.x rdelim.x read.x root.x rw.x scmsigs.x
\
+ properties.x r6rs-ports.x random.x rdelim.x
\
+ read.x root.x rw.x scmsigs.x \
script.x simpos.x smob.x sort.x srcprop.x stackchk.x stacks.x \
stime.x strings.x srfi-4.x srfi-13.x srfi-14.x strorder.x \
strports.x struct.x symbols.x threads.x throw.x values.x \
@@ -155,7 +159,8 @@ DOT_X_FILES += frames.x instructions.x objcodes.x
programs.x vm.x
EXTRA_DOT_X_FILES = @EXTRA_DOT_X_FILES@
DOT_DOC_FILES = alist.doc arbiters.doc async.doc backtrace.doc \
- boolean.doc chars.doc continuations.doc debug.doc deprecation.doc \
+ boolean.doc bytevectors.doc chars.doc \
+ continuations.doc debug.doc deprecation.doc
\
deprecated.doc discouraged.doc dynl.doc dynwind.doc
\
eq.doc error.doc eval.doc evalext.doc \
extensions.doc feature.doc fluids.doc fports.doc futures.doc \
@@ -165,7 +170,8 @@ DOT_DOC_FILES = alist.doc arbiters.doc async.doc
backtrace.doc \
hooks.doc i18n.doc init.doc ioext.doc keywords.doc lang.doc
\
list.doc load.doc macros.doc mallocs.doc modules.doc numbers.doc \
objects.doc objprop.doc options.doc pairs.doc ports.doc print.doc \
- procprop.doc procs.doc properties.doc random.doc rdelim.doc
\
+ procprop.doc procs.doc properties.doc r6rs-ports.doc \
+ random.doc rdelim.doc \
read.doc root.doc rw.doc scmsigs.doc script.doc simpos.doc \
smob.doc sort.doc srcprop.doc stackchk.doc stacks.doc stime.doc \
strings.doc srfi-4.doc srfi-13.doc srfi-14.doc strorder.doc
\
@@ -204,7 +210,7 @@ install-exec-hook:
## working.
noinst_HEADERS = convert.i.c \
conv-integer.i.c conv-uinteger.i.c \
- eval.i.c \
+ eval.i.c ieee-754.h \
srfi-4.i.c \
quicksort.i.c \
win32-uname.h win32-dirent.h win32-socket.h \
@@ -223,7 +229,8 @@ pkginclude_HEADERS =
# These are headers visible as <libguile/mumble.h>.
modincludedir = $(includedir)/libguile
modinclude_HEADERS = __scm.h alist.h arbiters.h async.h backtrace.h \
- boolean.h chars.h continuations.h convert.h debug.h debug-malloc.h \
+ boolean.h bytevectors.h chars.h continuations.h convert.h \
+ debug.h debug-malloc.h \
deprecation.h deprecated.h discouraged.h dynl.h dynwind.h \
eq.h error.h eval.h evalext.h extensions.h \
feature.h filesys.h fluids.h fports.h futures.h gc.h \
@@ -232,7 +239,8 @@ modinclude_HEADERS = __scm.h alist.h arbiters.h async.h
backtrace.h \
hashtab.h hooks.h i18n.h init.h inline.h ioext.h iselect.h \
keywords.h lang.h list.h load.h macros.h mallocs.h modules.h \
net_db.h numbers.h objects.h objprop.h options.h pairs.h ports.h \
- posix.h regex-posix.h print.h procprop.h procs.h properties.h \
+ posix.h r6rs-ports.h regex-posix.h print.h \
+ procprop.h procs.h properties.h \
random.h ramap.h rdelim.h read.h root.h rw.h scmsigs.h validate.h \
script.h simpos.h smob.h snarf.h socket.h sort.h srcprop.h \
stackchk.h stacks.h stime.h strings.h srfi-4.h srfi-13.h srfi-14.h \
diff --git a/libguile/bytevectors.c b/libguile/bytevectors.c
new file mode 100644
index 0000000..4c3a353
--- /dev/null
+++ b/libguile/bytevectors.c
@@ -0,0 +1,1978 @@
+/* Copyright (C) 2009 Free Software Foundation, Inc.
+ *
+ * This library is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU Lesser General Public
+ * License as published by the Free Software Foundation; either
+ * version 2.1 of the License, or (at your option) any later version.
+ *
+ * This library is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+ * Lesser General Public License for more details.
+ *
+ * You should have received a copy of the GNU Lesser General Public
+ * License along with this library; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
+ */
+
+
+#ifdef HAVE_CONFIG_H
+# include <config.h>
+#endif
+
+#include <alloca.h>
+
+#include <gmp.h>
+
+#include "libguile/_scm.h"
+#include "libguile/bytevectors.h"
+#include "libguile/strings.h"
+#include "libguile/validate.h"
+#include "libguile/ieee-754.h"
+
+#include <byteswap.h>
+#include <striconveh.h>
+#include <uniconv.h>
+
+#ifdef HAVE_LIMITS_H
+# include <limits.h>
+#else
+/* Assuming 32-bit longs. */
+# define ULONG_MAX 4294967295UL
+#endif
+
+#include <string.h>
+
+
+
+/* Utilities. */
+
+/* Convenience macros. These are used by the various templates (macros) that
+ are parameterized by integer signedness. */
+#define INT8_T_signed scm_t_int8
+#define INT8_T_unsigned scm_t_uint8
+#define INT16_T_signed scm_t_int16
+#define INT16_T_unsigned scm_t_uint16
+#define INT32_T_signed scm_t_int32
+#define INT32_T_unsigned scm_t_uint32
+#define is_signed_int8(_x) (((_x) >= -128L) && ((_x) <= 127L))
+#define is_unsigned_int8(_x) ((_x) <= 255UL)
+#define is_signed_int16(_x) (((_x) >= -32768L) && ((_x) <= 32767L))
+#define is_unsigned_int16(_x) ((_x) <= 65535UL)
+#define is_signed_int32(_x) (((_x) >= -2147483648L) && ((_x) <=
2147483647L))
+#define is_unsigned_int32(_x) ((_x) <= 4294967295UL)
+#define SIGNEDNESS_signed 1
+#define SIGNEDNESS_unsigned 0
+
+#define INT_TYPE(_size, _sign) INT ## _size ## _T_ ## _sign
+#define INT_SWAP(_size) bswap_ ## _size
+#define INT_VALID_P(_size, _sign) is_ ## _sign ## _int ## _size
+#define SIGNEDNESS(_sign) SIGNEDNESS_ ## _sign
+
+
+#define INTEGER_ACCESSOR_PROLOGUE(_len, _sign) \
+ unsigned c_len, c_index; \
+ _sign char *c_bv; \
+ \
+ SCM_VALIDATE_BYTEVECTOR (1, bv); \
+ c_index = scm_to_uint (index); \
+ \
+ c_len = SCM_BYTEVECTOR_LENGTH (bv); \
+ c_bv = (_sign char *) SCM_BYTEVECTOR_CONTENTS (bv); \
+ \
+ if (SCM_UNLIKELY (c_index + ((_len) >> 3UL) - 1 >= c_len)) \
+ scm_out_of_range (FUNC_NAME, index);
+
+/* Template for fixed-size integer access (only 8, 16 or 32-bit). */
+#define INTEGER_REF(_len, _sign) \
+ SCM result; \
+ \
+ INTEGER_ACCESSOR_PROLOGUE (_len, _sign); \
+ SCM_VALIDATE_SYMBOL (3, endianness); \
+ \
+ { \
+ INT_TYPE (_len, _sign) c_result; \
+ \
+ memcpy (&c_result, &c_bv[c_index], (_len) / 8); \
+ if (!scm_is_eq (endianness, native_endianness)) \
+ c_result = INT_SWAP (_len) (c_result); \
+ \
+ result = SCM_I_MAKINUM (c_result); \
+ } \
+ \
+ return result;
+
+/* Template for fixed-size integer access using the native endianness. */
+#define INTEGER_NATIVE_REF(_len, _sign) \
+ SCM result; \
+ \
+ INTEGER_ACCESSOR_PROLOGUE (_len, _sign); \
+ \
+ { \
+ INT_TYPE (_len, _sign) c_result; \
+ \
+ memcpy (&c_result, &c_bv[c_index], (_len) / 8); \
+ result = SCM_I_MAKINUM (c_result); \
+ } \
+ \
+ return result;
+
+/* Template for fixed-size integer modification (only 8, 16 or 32-bit). */
+#define INTEGER_SET(_len, _sign) \
+ INTEGER_ACCESSOR_PROLOGUE (_len, _sign); \
+ SCM_VALIDATE_SYMBOL (3, endianness); \
+ \
+ { \
+ _sign long c_value; \
+ INT_TYPE (_len, _sign) c_value_short; \
+ \
+ if (SCM_UNLIKELY (!SCM_I_INUMP (value))) \
+ scm_wrong_type_arg (FUNC_NAME, 3, value); \
+ \
+ c_value = SCM_I_INUM (value); \
+ if (SCM_UNLIKELY (!INT_VALID_P (_len, _sign) (c_value))) \
+ scm_out_of_range (FUNC_NAME, value); \
+ \
+ c_value_short = (INT_TYPE (_len, _sign)) c_value; \
+ if (!scm_is_eq (endianness, native_endianness)) \
+ c_value_short = INT_SWAP (_len) (c_value_short); \
+ \
+ memcpy (&c_bv[c_index], &c_value_short, (_len) / 8); \
+ } \
+ \
+ return SCM_UNSPECIFIED;
+
+/* Template for fixed-size integer modification using the native
+ endianness. */
+#define INTEGER_NATIVE_SET(_len, _sign) \
+ INTEGER_ACCESSOR_PROLOGUE (_len, _sign); \
+ \
+ { \
+ _sign long c_value; \
+ INT_TYPE (_len, _sign) c_value_short; \
+ \
+ if (SCM_UNLIKELY (!SCM_I_INUMP (value))) \
+ scm_wrong_type_arg (FUNC_NAME, 3, value); \
+ \
+ c_value = SCM_I_INUM (value); \
+ if (SCM_UNLIKELY (!INT_VALID_P (_len, _sign) (c_value))) \
+ scm_out_of_range (FUNC_NAME, value); \
+ \
+ c_value_short = (INT_TYPE (_len, _sign)) c_value; \
+ \
+ memcpy (&c_bv[c_index], &c_value_short, (_len) / 8); \
+ } \
+ \
+ return SCM_UNSPECIFIED;
+
+
+
+/* Bytevector type. */
+
+SCM_GLOBAL_SMOB (scm_tc16_bytevector, "r6rs-bytevector", 0);
+
+#define SCM_BYTEVECTOR_SET_LENGTH(_bv, _len) \
+ SCM_SET_SMOB_DATA ((_bv), (scm_t_bits) (_len))
+#define SCM_BYTEVECTOR_SET_CONTENTS(_bv, _buf) \
+ SCM_SET_SMOB_DATA_2 ((_bv), (scm_t_bits) (_buf))
+
+/* The empty bytevector. */
+SCM scm_null_bytevector = SCM_UNSPECIFIED;
+
+
+static inline SCM
+make_bytevector_from_buffer (unsigned len, signed char *contents)
+{
+ /* Assuming LEN > SCM_BYTEVECTOR_INLINE_THRESHOLD. */
+ SCM_RETURN_NEWSMOB2 (scm_tc16_bytevector, len, contents);
+}
+
+static inline SCM
+make_bytevector (unsigned len)
+{
+ SCM bv;
+
+ if (SCM_UNLIKELY (len == 0))
+ bv = scm_null_bytevector;
+ else
+ {
+ signed char *contents = NULL;
+
+ if (!SCM_BYTEVECTOR_INLINEABLE_SIZE_P (len))
+ contents = (signed char *) scm_gc_malloc (len, SCM_GC_BYTEVECTOR);
+
+ bv = make_bytevector_from_buffer (len, contents);
+ }
+
+ return bv;
+}
+
+/* Return a new bytevector of size LEN octets. */
+SCM
+scm_c_make_bytevector (unsigned len)
+{
+ return (make_bytevector (len));
+}
+
+/* Return a bytevector of size LEN made up of CONTENTS. The area pointed to
+ by CONTENTS must have been allocated using `scm_gc_malloc ()'. */
+SCM
+scm_c_take_bytevector (signed char *contents, unsigned len)
+{
+ SCM bv;
+
+ if (SCM_UNLIKELY (SCM_BYTEVECTOR_INLINEABLE_SIZE_P (len)))
+ {
+ /* Copy CONTENTS into an "in-line" buffer, then free CONTENTS. */
+ signed char *c_bv;
+
+ bv = make_bytevector (len);
+ c_bv = SCM_BYTEVECTOR_CONTENTS (bv);
+ memcpy (c_bv, contents, len);
+ scm_gc_free (contents, len, SCM_GC_BYTEVECTOR);
+ }
+ else
+ bv = make_bytevector_from_buffer (len, contents);
+
+ return bv;
+}
+
+/* Shrink BV to C_NEW_LEN (which is assumed to be smaller than its current
+ size) and return BV. */
+SCM
+scm_i_shrink_bytevector (SCM bv, unsigned c_new_len)
+{
+ if (!SCM_BYTEVECTOR_INLINE_P (bv))
+ {
+ unsigned c_len;
+ signed char *c_bv, *c_new_bv;
+
+ c_len = SCM_BYTEVECTOR_LENGTH (bv);
+ c_bv = SCM_BYTEVECTOR_CONTENTS (bv);
+
+ SCM_BYTEVECTOR_SET_LENGTH (bv, c_new_len);
+
+ if (SCM_BYTEVECTOR_INLINEABLE_SIZE_P (c_new_len))
+ {
+ /* Copy to the in-line buffer and free the current buffer. */
+ c_new_bv = SCM_BYTEVECTOR_CONTENTS (bv);
+ memcpy (c_new_bv, c_bv, c_new_len);
+ scm_gc_free (c_bv, c_len, SCM_GC_BYTEVECTOR);
+ }
+ else
+ {
+ /* Resize the existing buffer. */
+ c_new_bv = scm_gc_realloc (c_bv, c_len, c_new_len,
+ SCM_GC_BYTEVECTOR);
+ SCM_BYTEVECTOR_SET_CONTENTS (bv, c_new_bv);
+ }
+ }
+
+ return bv;
+}
+
+SCM_SMOB_PRINT (scm_tc16_bytevector, print_bytevector,
+ bv, port, pstate)
+{
+ unsigned c_len, i;
+ unsigned char *c_bv;
+
+ c_len = SCM_BYTEVECTOR_LENGTH (bv);
+ c_bv = (unsigned char *) SCM_BYTEVECTOR_CONTENTS (bv);
+
+ scm_puts ("#vu8(", port);
+ for (i = 0; i < c_len; i++)
+ {
+ if (i > 0)
+ scm_putc (' ', port);
+
+ scm_uintprint (c_bv[i], 10, port);
+ }
+
+ scm_putc (')', port);
+
+ /* Make GCC think we use it. */
+ scm_remember_upto_here ((SCM) pstate);
+
+ return 1;
+}
+
+SCM_SMOB_FREE (scm_tc16_bytevector, free_bytevector, bv)
+{
+
+ if (!SCM_BYTEVECTOR_INLINE_P (bv))
+ {
+ unsigned c_len;
+ signed char *c_bv;
+
+ c_bv = SCM_BYTEVECTOR_CONTENTS (bv);
+ c_len = SCM_BYTEVECTOR_LENGTH (bv);
+
+ scm_gc_free (c_bv, c_len, SCM_GC_BYTEVECTOR);
+ }
+
+ return 0;
+}
+
+
+
+/* General operations. */
+
+SCM_SYMBOL (scm_sym_big, "big");
+SCM_SYMBOL (scm_sym_little, "little");
+
+SCM scm_endianness_big, scm_endianness_little;
+
+/* Host endianness (a symbol). */
+static SCM native_endianness = SCM_UNSPECIFIED;
+
+/* Byte-swapping. */
+#ifndef bswap_24
+# define bswap_24(_x) \
+ ((((_x) & 0xff0000) >> 16) | \
+ (((_x) & 0x00ff00)) | \
+ (((_x) & 0x0000ff) << 16))
+#endif
+
+
+SCM_DEFINE (scm_native_endianness, "native-endianness", 0, 0, 0,
+ (void),
+ "Return a symbol denoting the machine's native endianness.")
+#define FUNC_NAME s_scm_native_endianness
+{
+ return native_endianness;
+}
+#undef FUNC_NAME
+
+SCM_DEFINE (scm_bytevector_p, "bytevector?", 1, 0, 0,
+ (SCM obj),
+ "Return true if @var{obj} is a bytevector.")
+#define FUNC_NAME s_scm_bytevector_p
+{
+ return (scm_from_bool (SCM_SMOB_PREDICATE (scm_tc16_bytevector,
+ obj)));
+}
+#undef FUNC_NAME
+
+SCM_DEFINE (scm_make_bytevector, "make-bytevector", 1, 1, 0,
+ (SCM len, SCM fill),
+ "Return a newly allocated bytevector of @var{len} bytes, "
+ "optionally filled with @var{fill}.")
+#define FUNC_NAME s_scm_make_bytevector
+{
+ SCM bv;
+ unsigned c_len;
+ signed char c_fill = '\0';
+
+ SCM_VALIDATE_UINT_COPY (1, len, c_len);
+ if (fill != SCM_UNDEFINED)
+ {
+ int value;
+
+ value = scm_to_int (fill);
+ if (SCM_UNLIKELY ((value < -128) || (value > 255)))
+ scm_out_of_range (FUNC_NAME, fill);
+ c_fill = (signed char) value;
+ }
+
+ bv = make_bytevector (c_len);
+ if (fill != SCM_UNDEFINED)
+ {
+ unsigned i;
+ signed char *contents;
+
+ contents = SCM_BYTEVECTOR_CONTENTS (bv);
+ for (i = 0; i < c_len; i++)
+ contents[i] = c_fill;
+ }
+
+ return bv;
+}
+#undef FUNC_NAME
+
+SCM_DEFINE (scm_bytevector_length, "bytevector-length", 1, 0, 0,
+ (SCM bv),
+ "Return the length (in bytes) of @var{bv}.")
+#define FUNC_NAME s_scm_bytevector_length
+{
+ SCM_VALIDATE_BYTEVECTOR (1, bv);
+
+ return (scm_from_uint (SCM_BYTEVECTOR_LENGTH (bv)));
+}
+#undef FUNC_NAME
+
+SCM_DEFINE (scm_bytevector_eq_p, "bytevector=?", 2, 0, 0,
+ (SCM bv1, SCM bv2),
+ "Return is @var{bv1} equals to @var{bv2}---i.e., if they "
+ "have the same length and contents.")
+#define FUNC_NAME s_scm_bytevector_eq_p
+{
+ SCM result = SCM_BOOL_F;
+ unsigned c_len1, c_len2;
+
+ SCM_VALIDATE_BYTEVECTOR (1, bv1);
+ SCM_VALIDATE_BYTEVECTOR (2, bv2);
+
+ c_len1 = SCM_BYTEVECTOR_LENGTH (bv1);
+ c_len2 = SCM_BYTEVECTOR_LENGTH (bv2);
+
+ if (c_len1 == c_len2)
+ {
+ signed char *c_bv1, *c_bv2;
+
+ c_bv1 = SCM_BYTEVECTOR_CONTENTS (bv1);
+ c_bv2 = SCM_BYTEVECTOR_CONTENTS (bv2);
+
+ result = scm_from_bool (!memcmp (c_bv1, c_bv2, c_len1));
+ }
+
+ return result;
+}
+#undef FUNC_NAME
+
+SCM_DEFINE (scm_bytevector_fill_x, "bytevector-fill!", 2, 0, 0,
+ (SCM bv, SCM fill),
+ "Fill bytevector @var{bv} with @var{fill}, a byte.")
+#define FUNC_NAME s_scm_bytevector_fill_x
+{
+ unsigned c_len, i;
+ signed char *c_bv, c_fill;
+
+ SCM_VALIDATE_BYTEVECTOR (1, bv);
+ c_fill = scm_to_int8 (fill);
+
+ c_len = SCM_BYTEVECTOR_LENGTH (bv);
+ c_bv = SCM_BYTEVECTOR_CONTENTS (bv);
+
+ for (i = 0; i < c_len; i++)
+ c_bv[i] = c_fill;
+
+ return SCM_UNSPECIFIED;
+}
+#undef FUNC_NAME
+
+SCM_DEFINE (scm_bytevector_copy_x, "bytevector-copy!", 5, 0, 0,
+ (SCM source, SCM source_start, SCM target, SCM target_start,
+ SCM len),
+ "Copy @var{len} bytes from @var{source} into @var{target}, "
+ "starting reading from @var{source_start} (a positive index "
+ "within @var{source}) and start writing at "
+ "@var{target_start}.")
+#define FUNC_NAME s_scm_bytevector_copy_x
+{
+ unsigned c_len, c_source_len, c_target_len;
+ unsigned c_source_start, c_target_start;
+ signed char *c_source, *c_target;
+
+ SCM_VALIDATE_BYTEVECTOR (1, source);
+ SCM_VALIDATE_BYTEVECTOR (3, target);
+
+ c_len = scm_to_uint (len);
+ c_source_start = scm_to_uint (source_start);
+ c_target_start = scm_to_uint (target_start);
+
+ c_source = SCM_BYTEVECTOR_CONTENTS (source);
+ c_target = SCM_BYTEVECTOR_CONTENTS (target);
+ c_source_len = SCM_BYTEVECTOR_LENGTH (source);
+ c_target_len = SCM_BYTEVECTOR_LENGTH (target);
+
+ if (SCM_UNLIKELY (c_source_start + c_len > c_source_len))
+ scm_out_of_range (FUNC_NAME, source_start);
+ if (SCM_UNLIKELY (c_target_start + c_len > c_target_len))
+ scm_out_of_range (FUNC_NAME, target_start);
+
+ memcpy (c_target + c_target_start,
+ c_source + c_source_start,
+ c_len);
+
+ return SCM_UNSPECIFIED;
+}
+#undef FUNC_NAME
+
+SCM_DEFINE (scm_bytevector_copy, "bytevector-copy", 1, 0, 0,
+ (SCM bv),
+ "Return a newly allocated copy of @var{bv}.")
+#define FUNC_NAME s_scm_bytevector_copy
+{
+ SCM copy;
+ unsigned c_len;
+ signed char *c_bv, *c_copy;
+
+ SCM_VALIDATE_BYTEVECTOR (1, bv);
+
+ c_len = SCM_BYTEVECTOR_LENGTH (bv);
+ c_bv = SCM_BYTEVECTOR_CONTENTS (bv);
+
+ copy = make_bytevector (c_len);
+ c_copy = SCM_BYTEVECTOR_CONTENTS (copy);
+ memcpy (c_copy, c_bv, c_len);
+
+ return copy;
+}
+#undef FUNC_NAME
+
+
+/* Operations on bytes and octets. */
+
+SCM_DEFINE (scm_bytevector_u8_ref, "bytevector-u8-ref", 2, 0, 0,
+ (SCM bv, SCM index),
+ "Return the octet located at @var{index} in @var{bv}.")
+#define FUNC_NAME s_scm_bytevector_u8_ref
+{
+ INTEGER_NATIVE_REF (8, unsigned);
+}
+#undef FUNC_NAME
+
+SCM_DEFINE (scm_bytevector_s8_ref, "bytevector-s8-ref", 2, 0, 0,
+ (SCM bv, SCM index),
+ "Return the byte located at @var{index} in @var{bv}.")
+#define FUNC_NAME s_scm_bytevector_s8_ref
+{
+ INTEGER_NATIVE_REF (8, signed);
+}
+#undef FUNC_NAME
+
+SCM_DEFINE (scm_bytevector_u8_set_x, "bytevector-u8-set!", 3, 0, 0,
+ (SCM bv, SCM index, SCM value),
+ "Return the octet located at @var{index} in @var{bv}.")
+#define FUNC_NAME s_scm_bytevector_u8_set_x
+{
+ INTEGER_NATIVE_SET (8, unsigned);
+}
+#undef FUNC_NAME
+
+SCM_DEFINE (scm_bytevector_s8_set_x, "bytevector-s8-set!", 3, 0, 0,
+ (SCM bv, SCM index, SCM value),
+ "Return the octet located at @var{index} in @var{bv}.")
+#define FUNC_NAME s_scm_bytevector_u8_set_x
+{
+ INTEGER_NATIVE_SET (8, signed);
+}
+#undef FUNC_NAME
+
+#undef OCTET_ACCESSOR_PROLOGUE
+
+
+SCM_DEFINE (scm_bytevector_to_u8_list, "bytevector->u8-list", 1, 0, 0,
+ (SCM bv),
+ "Return a newly allocated list of octets containing the "
+ "contents of @var{bv}.")
+#define FUNC_NAME s_scm_bytevector_to_u8_list
+{
+ SCM lst, pair;
+ unsigned c_len, i;
+ unsigned char *c_bv;
+
+ SCM_VALIDATE_BYTEVECTOR (1, bv);
+
+ c_len = SCM_BYTEVECTOR_LENGTH (bv);
+ c_bv = (unsigned char *) SCM_BYTEVECTOR_CONTENTS (bv);
+
+ lst = scm_make_list (scm_from_uint (c_len), SCM_UNSPECIFIED);
+ for (i = 0, pair = lst;
+ i < c_len;
+ i++, pair = SCM_CDR (pair))
+ {
+ SCM_SETCAR (pair, SCM_I_MAKINUM (c_bv[i]));
+ }
+
+ return lst;
+}
+#undef FUNC_NAME
+
+SCM_DEFINE (scm_u8_list_to_bytevector, "u8-list->bytevector", 1, 0, 0,
+ (SCM lst),
+ "Turn @var{lst}, a list of octets, into a bytevector.")
+#define FUNC_NAME s_scm_u8_list_to_bytevector
+{
+ SCM bv, item;
+ long c_len, i;
+ unsigned char *c_bv;
+
+ SCM_VALIDATE_LIST_COPYLEN (1, lst, c_len);
+
+ bv = make_bytevector (c_len);
+ c_bv = (unsigned char *) SCM_BYTEVECTOR_CONTENTS (bv);
+
+ for (i = 0; i < c_len; lst = SCM_CDR (lst), i++)
+ {
+ item = SCM_CAR (lst);
+
+ if (SCM_LIKELY (SCM_I_INUMP (item)))
+ {
+ long c_item;
+
+ c_item = SCM_I_INUM (item);
+ if (SCM_LIKELY ((c_item >= 0) && (c_item < 256)))
+ c_bv[i] = (unsigned char) c_item;
+ else
+ goto type_error;
+ }
+ else
+ goto type_error;
+ }
+
+ return bv;
+
+ type_error:
+ scm_wrong_type_arg (FUNC_NAME, 1, item);
+
+ return SCM_BOOL_F;
+}
+#undef FUNC_NAME
+
+/* Compute the two's complement of VALUE (a positive integer) on SIZE octets
+ using (2^(SIZE * 8) - VALUE). */
+static inline void
+twos_complement (mpz_t value, size_t size)
+{
+ unsigned long bit_count;
+
+ /* We expect BIT_COUNT to fit in a unsigned long thanks to the range
+ checking on SIZE performed earlier. */
+ bit_count = (unsigned long) size << 3UL;
+
+ if (SCM_LIKELY (bit_count < sizeof (unsigned long)))
+ mpz_ui_sub (value, 1UL << bit_count, value);
+ else
+ {
+ mpz_t max;
+
+ mpz_init (max);
+ mpz_ui_pow_ui (max, 2, bit_count);
+ mpz_sub (value, max, value);
+ mpz_clear (max);
+ }
+}
+
+static inline SCM
+bytevector_large_ref (const char *c_bv, size_t c_size, int signed_p,
+ SCM endianness)
+{
+ SCM result;
+ mpz_t c_mpz;
+ int c_endianness, negative_p = 0;
+
+ if (signed_p)
+ {
+ if (scm_is_eq (endianness, scm_sym_big))
+ negative_p = c_bv[0] & 0x80;
+ else
+ negative_p = c_bv[c_size - 1] & 0x80;
+ }
+
+ c_endianness = scm_is_eq (endianness, scm_sym_big) ? 1 : -1;
+
+ mpz_init (c_mpz);
+ mpz_import (c_mpz, 1 /* 1 word */, 1 /* word order doesn't matter */,
+ c_size /* word is C_SIZE-byte long */,
+ c_endianness,
+ 0 /* nails */, c_bv);
+
+ if (signed_p && negative_p)
+ {
+ twos_complement (c_mpz, c_size);
+ mpz_neg (c_mpz, c_mpz);
+ }
+
+ result = scm_from_mpz (c_mpz);
+ mpz_clear (c_mpz); /* FIXME: Needed? */
+
+ return result;
+}
+
+static inline int
+bytevector_large_set (char *c_bv, size_t c_size, int signed_p,
+ SCM value, SCM endianness)
+{
+ mpz_t c_mpz;
+ int c_endianness, c_sign, err = 0;
+
+ c_endianness = scm_is_eq (endianness, scm_sym_big) ? 1 : -1;
+
+ mpz_init (c_mpz);
+ scm_to_mpz (value, c_mpz);
+
+ c_sign = mpz_sgn (c_mpz);
+ if (c_sign < 0)
+ {
+ if (SCM_LIKELY (signed_p))
+ {
+ mpz_neg (c_mpz, c_mpz);
+ twos_complement (c_mpz, c_size);
+ }
+ else
+ {
+ err = -1;
+ goto finish;
+ }
+ }
+
+ if (c_sign == 0)
+ /* Zero. */
+ memset (c_bv, 0, c_size);
+ else
+ {
+ size_t word_count, value_size;
+
+ value_size = (mpz_sizeinbase (c_mpz, 2) + (8 * c_size)) / (8 * c_size);
+ if (SCM_UNLIKELY (value_size > c_size))
+ {
+ err = -2;
+ goto finish;
+ }
+
+
+ mpz_export (c_bv, &word_count, 1 /* word order doesn't matter */,
+ c_size, c_endianness,
+ 0 /* nails */, c_mpz);
+ if (SCM_UNLIKELY (word_count != 1))
+ /* Shouldn't happen since we already checked with VALUE_SIZE. */
+ abort ();
+ }
+
+ finish:
+ mpz_clear (c_mpz);
+
+ return err;
+}
+
+#define GENERIC_INTEGER_ACCESSOR_PROLOGUE(_sign) \
+ unsigned long c_len, c_index, c_size;
\
+ char *c_bv; \
+ \
+ SCM_VALIDATE_BYTEVECTOR (1, bv); \
+ c_index = scm_to_ulong (index); \
+ c_size = scm_to_ulong (size);
\
+ \
+ c_len = SCM_BYTEVECTOR_LENGTH (bv); \
+ c_bv = (char *) SCM_BYTEVECTOR_CONTENTS (bv);
\
+ \
+ /* C_SIZE must have its 3 higher bits set to zero so that \
+ multiplying it by 8 yields a number that fits in an \
+ unsigned long. */
\
+ if (SCM_UNLIKELY ((c_size == 0) || (c_size >= (ULONG_MAX >> 3L)))) \
+ scm_out_of_range (FUNC_NAME, size);
\
+ if (SCM_UNLIKELY (c_index + c_size > c_len)) \
+ scm_out_of_range (FUNC_NAME, index);
+
+
+/* Template of an integer reference function. */
+#define GENERIC_INTEGER_REF(_sign) \
+ SCM result; \
+ \
+ if (c_size < 3) \
+ { \
+ int swap;
\
+ _sign int value; \
+ \
+ swap = !scm_is_eq (endianness, native_endianness); \
+ switch (c_size) \
+ { \
+ case 1: \
+ { \
+ _sign char c_value8; \
+ memcpy (&c_value8, c_bv, 1); \
+ value = c_value8; \
+ } \
+ break; \
+ case 2: \
+ { \
+ INT_TYPE (16, _sign) c_value16; \
+ memcpy (&c_value16, c_bv, 2); \
+ if (swap) \
+ value = (INT_TYPE (16, _sign)) bswap_16 (c_value16); \
+ else \
+ value = c_value16; \
+ } \
+ break; \
+ default: \
+ abort (); \
+ } \
+ \
+ result = SCM_I_MAKINUM ((_sign int) value); \
+ } \
+ else \
+ result = bytevector_large_ref ((char *) c_bv, \
+ c_size, SIGNEDNESS (_sign), \
+ endianness); \
+ \
+ return result;
+
+static inline SCM
+bytevector_signed_ref (const char *c_bv, size_t c_size, SCM endianness)
+{
+ GENERIC_INTEGER_REF (signed);
+}
+
+static inline SCM
+bytevector_unsigned_ref (const char *c_bv, size_t c_size, SCM endianness)
+{
+ GENERIC_INTEGER_REF (unsigned);
+}
+
+
+/* Template of an integer assignment function. */
+#define GENERIC_INTEGER_SET(_sign) \
+ if (c_size < 3) \
+ { \
+ _sign int c_value; \
+ \
+ if (SCM_UNLIKELY (!SCM_I_INUMP (value))) \
+ goto range_error; \
+ \
+ c_value = SCM_I_INUM (value); \
+ switch (c_size) \
+ { \
+ case 1: \
+ if (SCM_LIKELY (INT_VALID_P (8, _sign) (c_value))) \
+ { \
+ _sign char c_value8; \
+ c_value8 = (_sign char) c_value; \
+ memcpy (c_bv, &c_value8, 1); \
+ } \
+ else \
+ goto range_error; \
+ break; \
+ \
+ case 2: \
+ if (SCM_LIKELY (INT_VALID_P (16, _sign) (c_value))) \
+ { \
+ int swap; \
+ INT_TYPE (16, _sign) c_value16; \
+ \
+ swap = !scm_is_eq (endianness, native_endianness); \
+ \
+ if (swap) \
+ c_value16 = (INT_TYPE (16, _sign)) bswap_16 (c_value); \
+ else \
+ c_value16 = c_value; \
+ \
+ memcpy (c_bv, &c_value16, 2); \
+ } \
+ else \
+ goto range_error; \
+ break; \
+ \
+ default: \
+ abort (); \
+ } \
+ } \
+ else \
+ { \
+ int err; \
+ \
+ err = bytevector_large_set (c_bv, c_size,
\
+ SIGNEDNESS (_sign), \
+ value, endianness); \
+ if (err) \
+ goto range_error; \
+ } \
+ \
+ return; \
+ \
+ range_error: \
+ scm_out_of_range (FUNC_NAME, value); \
+ return;
+
+static inline void
+bytevector_signed_set (char *c_bv, size_t c_size,
+ SCM value, SCM endianness,
+ const char *func_name)
+#define FUNC_NAME func_name
+{
+ GENERIC_INTEGER_SET (signed);
+}
+#undef FUNC_NAME
+
+static inline void
+bytevector_unsigned_set (char *c_bv, size_t c_size,
+ SCM value, SCM endianness,
+ const char *func_name)
+#define FUNC_NAME func_name
+{
+ GENERIC_INTEGER_SET (unsigned);
+}
+#undef FUNC_NAME
+
+#undef GENERIC_INTEGER_SET
+#undef GENERIC_INTEGER_REF
+
+
+SCM_DEFINE (scm_bytevector_uint_ref, "bytevector-uint-ref", 4, 0, 0,
+ (SCM bv, SCM index, SCM endianness, SCM size),
+ "Return the @var{size}-octet long unsigned integer at index "
+ "@var{index} in @var{bv}.")
+#define FUNC_NAME s_scm_bytevector_uint_ref
+{
+ GENERIC_INTEGER_ACCESSOR_PROLOGUE (unsigned);
+
+ return (bytevector_unsigned_ref (&c_bv[c_index], c_size, endianness));
+}
+#undef FUNC_NAME
+
+SCM_DEFINE (scm_bytevector_sint_ref, "bytevector-sint-ref", 4, 0, 0,
+ (SCM bv, SCM index, SCM endianness, SCM size),
+ "Return the @var{size}-octet long unsigned integer at index "
+ "@var{index} in @var{bv}.")
+#define FUNC_NAME s_scm_bytevector_sint_ref
+{
+ GENERIC_INTEGER_ACCESSOR_PROLOGUE (signed);
+
+ return (bytevector_signed_ref (&c_bv[c_index], c_size, endianness));
+}
+#undef FUNC_NAME
+
+SCM_DEFINE (scm_bytevector_uint_set_x, "bytevector-uint-set!", 5, 0, 0,
+ (SCM bv, SCM index, SCM value, SCM endianness, SCM size),
+ "Set the @var{size}-octet long unsigned integer at @var{index} "
+ "to @var{value}.")
+#define FUNC_NAME s_scm_bytevector_uint_set_x
+{
+ GENERIC_INTEGER_ACCESSOR_PROLOGUE (unsigned);
+
+ bytevector_unsigned_set (&c_bv[c_index], c_size, value, endianness,
+ FUNC_NAME);
+
+ return SCM_UNSPECIFIED;
+}
+#undef FUNC_NAME
+
+SCM_DEFINE (scm_bytevector_sint_set_x, "bytevector-sint-set!", 5, 0, 0,
+ (SCM bv, SCM index, SCM value, SCM endianness, SCM size),
+ "Set the @var{size}-octet long signed integer at @var{index} "
+ "to @var{value}.")
+#define FUNC_NAME s_scm_bytevector_sint_set_x
+{
+ GENERIC_INTEGER_ACCESSOR_PROLOGUE (signed);
+
+ bytevector_signed_set (&c_bv[c_index], c_size, value, endianness,
+ FUNC_NAME);
+
+ return SCM_UNSPECIFIED;
+}
+#undef FUNC_NAME
+
+
+
+/* Operations on integers of arbitrary size. */
+
+#define INTEGERS_TO_LIST(_sign)
\
+ SCM lst, pair; \
+ size_t i, c_len, c_size; \
+ \
+ SCM_VALIDATE_BYTEVECTOR (1, bv); \
+ SCM_VALIDATE_SYMBOL (2, endianness); \
+ c_size = scm_to_uint (size); \
+ \
+ c_len = SCM_BYTEVECTOR_LENGTH (bv); \
+ if (SCM_UNLIKELY (c_len == 0)) \
+ lst = SCM_EOL; \
+ else if (SCM_UNLIKELY (c_len < c_size)) \
+ scm_out_of_range (FUNC_NAME, size);
\
+ else \
+ { \
+ const char *c_bv;
\
+ \
+ c_bv = (char *) SCM_BYTEVECTOR_CONTENTS (bv); \
+ \
+ lst = scm_make_list (scm_from_uint (c_len / c_size), \
+ SCM_UNSPECIFIED); \
+ for (i = 0, pair = lst; \
+ i <= c_len - c_size; \
+ i += c_size, c_bv += c_size, pair = SCM_CDR (pair)) \
+ { \
+ SCM_SETCAR (pair, \
+ bytevector_ ## _sign ## _ref (c_bv, c_size, \
+ endianness)); \
+ } \
+ } \
+ \
+ return lst;
+
+SCM_DEFINE (scm_bytevector_to_sint_list, "bytevector->sint-list",
+ 3, 0, 0,
+ (SCM bv, SCM endianness, SCM size),
+ "Return a list of signed integers of @var{size} octets "
+ "representing the contents of @var{bv}.")
+#define FUNC_NAME s_scm_bytevector_to_sint_list
+{
+ INTEGERS_TO_LIST (signed);
+}
+#undef FUNC_NAME
+
+SCM_DEFINE (scm_bytevector_to_uint_list, "bytevector->uint-list",
+ 3, 0, 0,
+ (SCM bv, SCM endianness, SCM size),
+ "Return a list of unsigned integers of @var{size} octets "
+ "representing the contents of @var{bv}.")
+#define FUNC_NAME s_scm_bytevector_to_uint_list
+{
+ INTEGERS_TO_LIST (unsigned);
+}
+#undef FUNC_NAME
+
+#undef INTEGER_TO_LIST
+
+
+#define INTEGER_LIST_TO_BYTEVECTOR(_sign) \
+ SCM bv; \
+ long c_len; \
+ size_t c_size; \
+ char *c_bv, *c_bv_ptr; \
+ \
+ SCM_VALIDATE_LIST_COPYLEN (1, lst, c_len); \
+ SCM_VALIDATE_SYMBOL (2, endianness); \
+ c_size = scm_to_uint (size); \
+ \
+ if (SCM_UNLIKELY ((c_size == 0) || (c_size >= (ULONG_MAX >> 3L)))) \
+ scm_out_of_range (FUNC_NAME, size);
\
+ \
+ bv = make_bytevector (c_len * c_size); \
+ c_bv = (char *) SCM_BYTEVECTOR_CONTENTS (bv);
\
+ \
+ for (c_bv_ptr = c_bv;
\
+ !scm_is_null (lst); \
+ lst = SCM_CDR (lst), c_bv_ptr += c_size)
\
+ { \
+ bytevector_ ## _sign ## _set (c_bv_ptr, c_size, \
+ SCM_CAR (lst), endianness, \
+ FUNC_NAME); \
+ } \
+ \
+ return bv;
+
+
+SCM_DEFINE (scm_uint_list_to_bytevector, "uint-list->bytevector",
+ 3, 0, 0,
+ (SCM lst, SCM endianness, SCM size),
+ "Return a bytevector containing the unsigned integers "
+ "listed in @var{lst} and encoded on @var{size} octets "
+ "according to @var{endianness}.")
+#define FUNC_NAME s_scm_uint_list_to_bytevector
+{
+ INTEGER_LIST_TO_BYTEVECTOR (unsigned);
+}
+#undef FUNC_NAME
+
+SCM_DEFINE (scm_sint_list_to_bytevector, "sint-list->bytevector",
+ 3, 0, 0,
+ (SCM lst, SCM endianness, SCM size),
+ "Return a bytevector containing the signed integers "
+ "listed in @var{lst} and encoded on @var{size} octets "
+ "according to @var{endianness}.")
+#define FUNC_NAME s_scm_sint_list_to_bytevector
+{
+ INTEGER_LIST_TO_BYTEVECTOR (signed);
+}
+#undef FUNC_NAME
+
+#undef INTEGER_LIST_TO_BYTEVECTOR
+
+
+
+/* Operations on 16-bit integers. */
+
+SCM_DEFINE (scm_bytevector_u16_ref, "bytevector-u16-ref",
+ 3, 0, 0,
+ (SCM bv, SCM index, SCM endianness),
+ "Return the unsigned 16-bit integer from @var{bv} at "
+ "@var{index}.")
+#define FUNC_NAME s_scm_bytevector_u16_ref
+{
+ INTEGER_REF (16, unsigned);
+}
+#undef FUNC_NAME
+
+SCM_DEFINE (scm_bytevector_s16_ref, "bytevector-s16-ref",
+ 3, 0, 0,
+ (SCM bv, SCM index, SCM endianness),
+ "Return the signed 16-bit integer from @var{bv} at "
+ "@var{index}.")
+#define FUNC_NAME s_scm_bytevector_s16_ref
+{
+ INTEGER_REF (16, signed);
+}
+#undef FUNC_NAME
+
+SCM_DEFINE (scm_bytevector_u16_native_ref, "bytevector-u16-native-ref",
+ 2, 0, 0,
+ (SCM bv, SCM index),
+ "Return the unsigned 16-bit integer from @var{bv} at "
+ "@var{index} using the native endianness.")
+#define FUNC_NAME s_scm_bytevector_u16_native_ref
+{
+ INTEGER_NATIVE_REF (16, unsigned);
+}
+#undef FUNC_NAME
+
+SCM_DEFINE (scm_bytevector_s16_native_ref, "bytevector-s16-native-ref",
+ 2, 0, 0,
+ (SCM bv, SCM index),
+ "Return the unsigned 16-bit integer from @var{bv} at "
+ "@var{index} using the native endianness.")
+#define FUNC_NAME s_scm_bytevector_s16_native_ref
+{
+ INTEGER_NATIVE_REF (16, signed);
+}
+#undef FUNC_NAME
+
+SCM_DEFINE (scm_bytevector_u16_set_x, "bytevector-u16-set!",
+ 4, 0, 0,
+ (SCM bv, SCM index, SCM value, SCM endianness),
+ "Store @var{value} in @var{bv} at @var{index} according to "
+ "@var{endianness}.")
+#define FUNC_NAME s_scm_bytevector_u16_set_x
+{
+ INTEGER_SET (16, unsigned);
+}
+#undef FUNC_NAME
+
+SCM_DEFINE (scm_bytevector_s16_set_x, "bytevector-s16-set!",
+ 4, 0, 0,
+ (SCM bv, SCM index, SCM value, SCM endianness),
+ "Store @var{value} in @var{bv} at @var{index} according to "
+ "@var{endianness}.")
+#define FUNC_NAME s_scm_bytevector_s16_set_x
+{
+ INTEGER_SET (16, signed);
+}
+#undef FUNC_NAME
+
+SCM_DEFINE (scm_bytevector_u16_native_set_x, "bytevector-u16-native-set!",
+ 3, 0, 0,
+ (SCM bv, SCM index, SCM value),
+ "Store the unsigned integer @var{value} at index @var{index} "
+ "of @var{bv} using the native endianness.")
+#define FUNC_NAME s_scm_bytevector_u16_native_set_x
+{
+ INTEGER_NATIVE_SET (16, unsigned);
+}
+#undef FUNC_NAME
+
+SCM_DEFINE (scm_bytevector_s16_native_set_x, "bytevector-s16-native-set!",
+ 3, 0, 0,
+ (SCM bv, SCM index, SCM value),
+ "Store the signed integer @var{value} at index @var{index} "
+ "of @var{bv} using the native endianness.")
+#define FUNC_NAME s_scm_bytevector_s16_native_set_x
+{
+ INTEGER_NATIVE_SET (16, signed);
+}
+#undef FUNC_NAME
+
+
+
+/* Operations on 32-bit integers. */
+
+/* Unfortunately, on 32-bit machines `SCM' is not large enough to hold
+ arbitrary 32-bit integers. Thus we fall back to using the
+ `large_{ref,set}' variants on 32-bit machines. */
+
+#define LARGE_INTEGER_REF(_len, _sign) \
+ INTEGER_ACCESSOR_PROLOGUE(_len, _sign); \
+ SCM_VALIDATE_SYMBOL (3, endianness); \
+ \
+ return (bytevector_large_ref ((char *) c_bv + c_index, _len / 8, \
+ SIGNEDNESS (_sign), endianness));
+
+#define LARGE_INTEGER_SET(_len, _sign) \
+ int err; \
+ INTEGER_ACCESSOR_PROLOGUE (_len, _sign); \
+ SCM_VALIDATE_SYMBOL (4, endianness); \
+ \
+ err = bytevector_large_set ((char *) c_bv + c_index, _len / 8, \
+ SIGNEDNESS (_sign), value, endianness); \
+ if (SCM_UNLIKELY (err)) \
+ scm_out_of_range (FUNC_NAME, value); \
+ \
+ return SCM_UNSPECIFIED;
+
+#define LARGE_INTEGER_NATIVE_REF(_len, _sign) \
+ INTEGER_ACCESSOR_PROLOGUE(_len, _sign); \
+ return (bytevector_large_ref ((char *) c_bv + c_index, _len / 8, \
+ SIGNEDNESS (_sign), native_endianness));
+
+#define LARGE_INTEGER_NATIVE_SET(_len, _sign) \
+ int err; \
+ INTEGER_ACCESSOR_PROLOGUE (_len, _sign); \
+ \
+ err = bytevector_large_set ((char *) c_bv + c_index, _len / 8, \
+ SIGNEDNESS (_sign), value, \
+ native_endianness); \
+ if (SCM_UNLIKELY (err)) \
+ scm_out_of_range (FUNC_NAME, value); \
+ \
+ return SCM_UNSPECIFIED;
+
+
+SCM_DEFINE (scm_bytevector_u32_ref, "bytevector-u32-ref",
+ 3, 0, 0,
+ (SCM bv, SCM index, SCM endianness),
+ "Return the unsigned 32-bit integer from @var{bv} at "
+ "@var{index}.")
+#define FUNC_NAME s_scm_bytevector_u32_ref
+{
+#if SIZEOF_VOID_P > 4
+ INTEGER_REF (32, unsigned);
+#else
+ LARGE_INTEGER_REF (32, unsigned);
+#endif
+}
+#undef FUNC_NAME
+
+SCM_DEFINE (scm_bytevector_s32_ref, "bytevector-s32-ref",
+ 3, 0, 0,
+ (SCM bv, SCM index, SCM endianness),
+ "Return the signed 32-bit integer from @var{bv} at "
+ "@var{index}.")
+#define FUNC_NAME s_scm_bytevector_s32_ref
+{
+#if SIZEOF_VOID_P > 4
+ INTEGER_REF (32, signed);
+#else
+ LARGE_INTEGER_REF (32, signed);
+#endif
+}
+#undef FUNC_NAME
+
+SCM_DEFINE (scm_bytevector_u32_native_ref, "bytevector-u32-native-ref",
+ 2, 0, 0,
+ (SCM bv, SCM index),
+ "Return the unsigned 32-bit integer from @var{bv} at "
+ "@var{index} using the native endianness.")
+#define FUNC_NAME s_scm_bytevector_u32_native_ref
+{
+#if SIZEOF_VOID_P > 4
+ INTEGER_NATIVE_REF (32, unsigned);
+#else
+ LARGE_INTEGER_NATIVE_REF (32, unsigned);
+#endif
+}
+#undef FUNC_NAME
+
+SCM_DEFINE (scm_bytevector_s32_native_ref, "bytevector-s32-native-ref",
+ 2, 0, 0,
+ (SCM bv, SCM index),
+ "Return the unsigned 32-bit integer from @var{bv} at "
+ "@var{index} using the native endianness.")
+#define FUNC_NAME s_scm_bytevector_s32_native_ref
+{
+#if SIZEOF_VOID_P > 4
+ INTEGER_NATIVE_REF (32, signed);
+#else
+ LARGE_INTEGER_NATIVE_REF (32, signed);
+#endif
+}
+#undef FUNC_NAME
+
+SCM_DEFINE (scm_bytevector_u32_set_x, "bytevector-u32-set!",
+ 4, 0, 0,
+ (SCM bv, SCM index, SCM value, SCM endianness),
+ "Store @var{value} in @var{bv} at @var{index} according to "
+ "@var{endianness}.")
+#define FUNC_NAME s_scm_bytevector_u32_set_x
+{
+#if SIZEOF_VOID_P > 4
+ INTEGER_SET (32, unsigned);
+#else
+ LARGE_INTEGER_SET (32, unsigned);
+#endif
+}
+#undef FUNC_NAME
+
+SCM_DEFINE (scm_bytevector_s32_set_x, "bytevector-s32-set!",
+ 4, 0, 0,
+ (SCM bv, SCM index, SCM value, SCM endianness),
+ "Store @var{value} in @var{bv} at @var{index} according to "
+ "@var{endianness}.")
+#define FUNC_NAME s_scm_bytevector_s32_set_x
+{
+#if SIZEOF_VOID_P > 4
+ INTEGER_SET (32, signed);
+#else
+ LARGE_INTEGER_SET (32, signed);
+#endif
+}
+#undef FUNC_NAME
+
+SCM_DEFINE (scm_bytevector_u32_native_set_x, "bytevector-u32-native-set!",
+ 3, 0, 0,
+ (SCM bv, SCM index, SCM value),
+ "Store the unsigned integer @var{value} at index @var{index} "
+ "of @var{bv} using the native endianness.")
+#define FUNC_NAME s_scm_bytevector_u32_native_set_x
+{
+#if SIZEOF_VOID_P > 4
+ INTEGER_NATIVE_SET (32, unsigned);
+#else
+ LARGE_INTEGER_NATIVE_SET (32, unsigned);
+#endif
+}
+#undef FUNC_NAME
+
+SCM_DEFINE (scm_bytevector_s32_native_set_x, "bytevector-s32-native-set!",
+ 3, 0, 0,
+ (SCM bv, SCM index, SCM value),
+ "Store the signed integer @var{value} at index @var{index} "
+ "of @var{bv} using the native endianness.")
+#define FUNC_NAME s_scm_bytevector_s32_native_set_x
+{
+#if SIZEOF_VOID_P > 4
+ INTEGER_NATIVE_SET (32, signed);
+#else
+ LARGE_INTEGER_NATIVE_SET (32, signed);
+#endif
+}
+#undef FUNC_NAME
+
+
+
+/* Operations on 64-bit integers. */
+
+/* For 64-bit integers, we use only the `large_{ref,set}' variant. */
+
+SCM_DEFINE (scm_bytevector_u64_ref, "bytevector-u64-ref",
+ 3, 0, 0,
+ (SCM bv, SCM index, SCM endianness),
+ "Return the unsigned 64-bit integer from @var{bv} at "
+ "@var{index}.")
+#define FUNC_NAME s_scm_bytevector_u64_ref
+{
+ LARGE_INTEGER_REF (64, unsigned);
+}
+#undef FUNC_NAME
+
+SCM_DEFINE (scm_bytevector_s64_ref, "bytevector-s64-ref",
+ 3, 0, 0,
+ (SCM bv, SCM index, SCM endianness),
+ "Return the signed 64-bit integer from @var{bv} at "
+ "@var{index}.")
+#define FUNC_NAME s_scm_bytevector_s64_ref
+{
+ LARGE_INTEGER_REF (64, signed);
+}
+#undef FUNC_NAME
+
+SCM_DEFINE (scm_bytevector_u64_native_ref, "bytevector-u64-native-ref",
+ 2, 0, 0,
+ (SCM bv, SCM index),
+ "Return the unsigned 64-bit integer from @var{bv} at "
+ "@var{index} using the native endianness.")
+#define FUNC_NAME s_scm_bytevector_u64_native_ref
+{
+ LARGE_INTEGER_NATIVE_REF (64, unsigned);
+}
+#undef FUNC_NAME
+
+SCM_DEFINE (scm_bytevector_s64_native_ref, "bytevector-s64-native-ref",
+ 2, 0, 0,
+ (SCM bv, SCM index),
+ "Return the unsigned 64-bit integer from @var{bv} at "
+ "@var{index} using the native endianness.")
+#define FUNC_NAME s_scm_bytevector_s64_native_ref
+{
+ LARGE_INTEGER_NATIVE_REF (64, signed);
+}
+#undef FUNC_NAME
+
+SCM_DEFINE (scm_bytevector_u64_set_x, "bytevector-u64-set!",
+ 4, 0, 0,
+ (SCM bv, SCM index, SCM value, SCM endianness),
+ "Store @var{value} in @var{bv} at @var{index} according to "
+ "@var{endianness}.")
+#define FUNC_NAME s_scm_bytevector_u64_set_x
+{
+ LARGE_INTEGER_SET (64, unsigned);
+}
+#undef FUNC_NAME
+
+SCM_DEFINE (scm_bytevector_s64_set_x, "bytevector-s64-set!",
+ 4, 0, 0,
+ (SCM bv, SCM index, SCM value, SCM endianness),
+ "Store @var{value} in @var{bv} at @var{index} according to "
+ "@var{endianness}.")
+#define FUNC_NAME s_scm_bytevector_s64_set_x
+{
+ LARGE_INTEGER_SET (64, signed);
+}
+#undef FUNC_NAME
+
+SCM_DEFINE (scm_bytevector_u64_native_set_x, "bytevector-u64-native-set!",
+ 3, 0, 0,
+ (SCM bv, SCM index, SCM value),
+ "Store the unsigned integer @var{value} at index @var{index} "
+ "of @var{bv} using the native endianness.")
+#define FUNC_NAME s_scm_bytevector_u64_native_set_x
+{
+ LARGE_INTEGER_NATIVE_SET (64, unsigned);
+}
+#undef FUNC_NAME
+
+SCM_DEFINE (scm_bytevector_s64_native_set_x, "bytevector-s64-native-set!",
+ 3, 0, 0,
+ (SCM bv, SCM index, SCM value),
+ "Store the signed integer @var{value} at index @var{index} "
+ "of @var{bv} using the native endianness.")
+#define FUNC_NAME s_scm_bytevector_s64_native_set_x
+{
+ LARGE_INTEGER_NATIVE_SET (64, signed);
+}
+#undef FUNC_NAME
+
+
+
+/* Operations on IEEE-754 numbers. */
+
+/* There are two possible word endians, visible in glibc's <ieee754.h>.
+ However, in R6RS, when the endianness is `little', little endian is
+ assumed for both the byte order and the word order. This is clear from
+ Section 2.1 of R6RS-lib (in response to
+ http://www.r6rs.org/formal-comments/comment-187.txt). */
+
+
+/* Convert to/from a floating-point number with different endianness. This
+ method is probably not the most efficient but it should be portable. */
+
+static inline void
+float_to_foreign_endianness (union scm_ieee754_float *target,
+ float source)
+{
+ union scm_ieee754_float src;
+
+ src.f = source;
+
+#ifdef WORDS_BIGENDIAN
+ /* Assuming little endian for both byte and word order. */
+ target->little_endian.negative = src.big_endian.negative;
+ target->little_endian.exponent = src.big_endian.exponent;
+ target->little_endian.mantissa = src.big_endian.mantissa;
+#else
+ target->big_endian.negative = src.little_endian.negative;
+ target->big_endian.exponent = src.little_endian.exponent;
+ target->big_endian.mantissa = src.little_endian.mantissa;
+#endif
+}
+
+static inline float
+float_from_foreign_endianness (const union scm_ieee754_float *source)
+{
+ union scm_ieee754_float result;
+
+#ifdef WORDS_BIGENDIAN
+ /* Assuming little endian for both byte and word order. */
+ result.big_endian.negative = source->little_endian.negative;
+ result.big_endian.exponent = source->little_endian.exponent;
+ result.big_endian.mantissa = source->little_endian.mantissa;
+#else
+ result.little_endian.negative = source->big_endian.negative;
+ result.little_endian.exponent = source->big_endian.exponent;
+ result.little_endian.mantissa = source->big_endian.mantissa;
+#endif
+
+ return (result.f);
+}
+
+static inline void
+double_to_foreign_endianness (union scm_ieee754_double *target,
+ double source)
+{
+ union scm_ieee754_double src;
+
+ src.d = source;
+
+#ifdef WORDS_BIGENDIAN
+ /* Assuming little endian for both byte and word order. */
+ target->little_little_endian.negative = src.big_endian.negative;
+ target->little_little_endian.exponent = src.big_endian.exponent;
+ target->little_little_endian.mantissa0 = src.big_endian.mantissa0;
+ target->little_little_endian.mantissa1 = src.big_endian.mantissa1;
+#else
+ target->big_endian.negative = src.little_little_endian.negative;
+ target->big_endian.exponent = src.little_little_endian.exponent;
+ target->big_endian.mantissa0 = src.little_little_endian.mantissa0;
+ target->big_endian.mantissa1 = src.little_little_endian.mantissa1;
+#endif
+}
+
+static inline double
+double_from_foreign_endianness (const union scm_ieee754_double *source)
+{
+ union scm_ieee754_double result;
+
+#ifdef WORDS_BIGENDIAN
+ /* Assuming little endian for both byte and word order. */
+ result.big_endian.negative = source->little_little_endian.negative;
+ result.big_endian.exponent = source->little_little_endian.exponent;
+ result.big_endian.mantissa0 = source->little_little_endian.mantissa0;
+ result.big_endian.mantissa1 = source->little_little_endian.mantissa1;
+#else
+ result.little_little_endian.negative = source->big_endian.negative;
+ result.little_little_endian.exponent = source->big_endian.exponent;
+ result.little_little_endian.mantissa0 = source->big_endian.mantissa0;
+ result.little_little_endian.mantissa1 = source->big_endian.mantissa1;
+#endif
+
+ return (result.d);
+}
+
+/* Template macros to abstract over doubles and floats.
+ XXX: Guile can only convert to/from doubles. */
+#define IEEE754_UNION(_c_type) union scm_ieee754_ ## _c_type
+#define IEEE754_TO_SCM(_c_type) scm_from_double
+#define IEEE754_FROM_SCM(_c_type) scm_to_double
+#define IEEE754_FROM_FOREIGN_ENDIANNESS(_c_type) \
+ _c_type ## _from_foreign_endianness
+#define IEEE754_TO_FOREIGN_ENDIANNESS(_c_type) \
+ _c_type ## _to_foreign_endianness
+
+
+/* Templace getters and setters. */
+
+#define IEEE754_ACCESSOR_PROLOGUE(_type) \
+ INTEGER_ACCESSOR_PROLOGUE (sizeof (_type) << 3UL, signed);
+
+#define IEEE754_REF(_type) \
+ _type c_result; \
+ \
+ IEEE754_ACCESSOR_PROLOGUE (_type); \
+ SCM_VALIDATE_SYMBOL (3, endianness); \
+ \
+ if (scm_is_eq (endianness, native_endianness)) \
+ memcpy (&c_result, &c_bv[c_index], sizeof (c_result)); \
+ else \
+ { \
+ IEEE754_UNION (_type) c_raw; \
+ \
+ memcpy (&c_raw, &c_bv[c_index], sizeof (c_raw)); \
+ c_result = \
+ IEEE754_FROM_FOREIGN_ENDIANNESS (_type) (&c_raw); \
+ } \
+ \
+ return (IEEE754_TO_SCM (_type) (c_result));
+
+#define IEEE754_NATIVE_REF(_type) \
+ _type c_result; \
+ \
+ IEEE754_ACCESSOR_PROLOGUE (_type); \
+ \
+ memcpy (&c_result, &c_bv[c_index], sizeof (c_result)); \
+ return (IEEE754_TO_SCM (_type) (c_result));
+
+#define IEEE754_SET(_type) \
+ _type c_value; \
+ \
+ IEEE754_ACCESSOR_PROLOGUE (_type); \
+ SCM_VALIDATE_REAL (3, value); \
+ SCM_VALIDATE_SYMBOL (4, endianness); \
+ c_value = IEEE754_FROM_SCM (_type) (value); \
+ \
+ if (scm_is_eq (endianness, native_endianness)) \
+ memcpy (&c_bv[c_index], &c_value, sizeof (c_value)); \
+ else \
+ { \
+ IEEE754_UNION (_type) c_raw; \
+ \
+ IEEE754_TO_FOREIGN_ENDIANNESS (_type) (&c_raw, c_value); \
+ memcpy (&c_bv[c_index], &c_raw, sizeof (c_raw)); \
+ } \
+ \
+ return SCM_UNSPECIFIED;
+
+#define IEEE754_NATIVE_SET(_type) \
+ _type c_value; \
+ \
+ IEEE754_ACCESSOR_PROLOGUE (_type); \
+ SCM_VALIDATE_REAL (3, value); \
+ c_value = IEEE754_FROM_SCM (_type) (value); \
+ \
+ memcpy (&c_bv[c_index], &c_value, sizeof (c_value)); \
+ return SCM_UNSPECIFIED;
+
+
+/* Single precision. */
+
+SCM_DEFINE (scm_bytevector_ieee_single_ref,
+ "bytevector-ieee-single-ref",
+ 3, 0, 0,
+ (SCM bv, SCM index, SCM endianness),
+ "Return the IEEE-754 single from @var{bv} at "
+ "@var{index}.")
+#define FUNC_NAME s_scm_bytevector_ieee_single_ref
+{
+ IEEE754_REF (float);
+}
+#undef FUNC_NAME
+
+SCM_DEFINE (scm_bytevector_ieee_single_native_ref,
+ "bytevector-ieee-single-native-ref",
+ 2, 0, 0,
+ (SCM bv, SCM index),
+ "Return the IEEE-754 single from @var{bv} at "
+ "@var{index} using the native endianness.")
+#define FUNC_NAME s_scm_bytevector_ieee_single_native_ref
+{
+ IEEE754_NATIVE_REF (float);
+}
+#undef FUNC_NAME
+
+SCM_DEFINE (scm_bytevector_ieee_single_set_x,
+ "bytevector-ieee-single-set!",
+ 4, 0, 0,
+ (SCM bv, SCM index, SCM value, SCM endianness),
+ "Store real @var{value} in @var{bv} at @var{index} according to "
+ "@var{endianness}.")
+#define FUNC_NAME s_scm_bytevector_ieee_single_set_x
+{
+ IEEE754_SET (float);
+}
+#undef FUNC_NAME
+
+SCM_DEFINE (scm_bytevector_ieee_single_native_set_x,
+ "bytevector-ieee-single-native-set!",
+ 3, 0, 0,
+ (SCM bv, SCM index, SCM value),
+ "Store the real @var{value} at index @var{index} "
+ "of @var{bv} using the native endianness.")
+#define FUNC_NAME s_scm_bytevector_ieee_single_native_set_x
+{
+ IEEE754_NATIVE_SET (float);
+}
+#undef FUNC_NAME
+
+
+/* Double precision. */
+
+SCM_DEFINE (scm_bytevector_ieee_double_ref,
+ "bytevector-ieee-double-ref",
+ 3, 0, 0,
+ (SCM bv, SCM index, SCM endianness),
+ "Return the IEEE-754 double from @var{bv} at "
+ "@var{index}.")
+#define FUNC_NAME s_scm_bytevector_ieee_double_ref
+{
+ IEEE754_REF (double);
+}
+#undef FUNC_NAME
+
+SCM_DEFINE (scm_bytevector_ieee_double_native_ref,
+ "bytevector-ieee-double-native-ref",
+ 2, 0, 0,
+ (SCM bv, SCM index),
+ "Return the IEEE-754 double from @var{bv} at "
+ "@var{index} using the native endianness.")
+#define FUNC_NAME s_scm_bytevector_ieee_double_native_ref
+{
+ IEEE754_NATIVE_REF (double);
+}
+#undef FUNC_NAME
+
+SCM_DEFINE (scm_bytevector_ieee_double_set_x,
+ "bytevector-ieee-double-set!",
+ 4, 0, 0,
+ (SCM bv, SCM index, SCM value, SCM endianness),
+ "Store real @var{value} in @var{bv} at @var{index} according to "
+ "@var{endianness}.")
+#define FUNC_NAME s_scm_bytevector_ieee_double_set_x
+{
+ IEEE754_SET (double);
+}
+#undef FUNC_NAME
+
+SCM_DEFINE (scm_bytevector_ieee_double_native_set_x,
+ "bytevector-ieee-double-native-set!",
+ 3, 0, 0,
+ (SCM bv, SCM index, SCM value),
+ "Store the real @var{value} at index @var{index} "
+ "of @var{bv} using the native endianness.")
+#define FUNC_NAME s_scm_bytevector_ieee_double_native_set_x
+{
+ IEEE754_NATIVE_SET (double);
+}
+#undef FUNC_NAME
+
+
+#undef IEEE754_UNION
+#undef IEEE754_TO_SCM
+#undef IEEE754_FROM_SCM
+#undef IEEE754_FROM_FOREIGN_ENDIANNESS
+#undef IEEE754_TO_FOREIGN_ENDIANNESS
+#undef IEEE754_REF
+#undef IEEE754_NATIVE_REF
+#undef IEEE754_SET
+#undef IEEE754_NATIVE_SET
+
+
+/* Operations on strings. */
+
+
+/* Produce a function that returns the length of a UTF-encoded string. */
+#define UTF_STRLEN_FUNCTION(_utf_width)
\
+static inline size_t \
+utf ## _utf_width ## _strlen (const uint ## _utf_width ## _t *str) \
+{ \
+ size_t len = 0; \
+ const uint ## _utf_width ## _t *ptr; \
+ for (ptr = str; \
+ *ptr != 0; \
+ ptr++) \
+ { \
+ len++; \
+ } \
+ \
+ return (len * ((_utf_width) / 8)); \
+}
+
+UTF_STRLEN_FUNCTION (8)
+
+
+/* Return the length (in bytes) of STR, a UTF-(UTF_WIDTH) encoded string. */
+#define UTF_STRLEN(_utf_width, _str) \
+ utf ## _utf_width ## _strlen (_str)
+
+/* Return the "portable" name of the UTF encoding of size UTF_WIDTH and
+ ENDIANNESS (Gnulib's `iconv_open' module guarantees the portability of the
+ encoding name). */
+static inline void
+utf_encoding_name (char *name, size_t utf_width, SCM endianness)
+{
+ strcpy (name, "UTF-");
+ strcat (name, ((utf_width == 8)
+ ? "8"
+ : ((utf_width == 16)
+ ? "16"
+ : ((utf_width == 32)
+ ? "32"
+ : "??"))));
+ strcat (name,
+ ((scm_is_eq (endianness, scm_sym_big))
+ ? "BE"
+ : ((scm_is_eq (endianness, scm_sym_little))
+ ? "LE"
+ : "unknown")));
+}
+
+/* Maximum length of a UTF encoding name. */
+#define MAX_UTF_ENCODING_NAME_LEN 16
+
+/* Produce the body of a `string->utf' function. */
+#define STRING_TO_UTF(_utf_width) \
+ SCM utf; \
+ int err; \
+ char *c_str; \
+ char c_utf_name[MAX_UTF_ENCODING_NAME_LEN]; \
+ char *c_utf = NULL, *c_locale; \
+ size_t c_strlen, c_raw_strlen, c_utf_len = 0;
\
+ \
+ SCM_VALIDATE_STRING (1, str);
\
+ if (endianness == SCM_UNDEFINED) \
+ endianness = scm_sym_big; \
+ else \
+ SCM_VALIDATE_SYMBOL (2, endianness); \
+ \
+ c_strlen = scm_c_string_length (str);
\
+ c_raw_strlen = c_strlen * ((_utf_width) / 8);
\
+ do \
+ { \
+ c_str = (char *) alloca (c_raw_strlen + 1); \
+ c_raw_strlen = scm_to_locale_stringbuf (str, c_str, c_strlen); \
+ } \
+ while (c_raw_strlen > c_strlen); \
+ c_str[c_raw_strlen] = '\0'; \
+ \
+ utf_encoding_name (c_utf_name, (_utf_width), endianness); \
+ \
+ c_locale = (char *) alloca (strlen (locale_charset ()) + 1); \
+ strcpy (c_locale, locale_charset ());
\
+ \
+ err = mem_iconveh (c_str, c_raw_strlen, \
+ c_locale, c_utf_name, \
+ iconveh_question_mark, NULL, \
+ &c_utf, &c_utf_len); \
+ if (SCM_UNLIKELY (err)) \
+ scm_syserror_msg (FUNC_NAME, "failed to convert string: ~A", \
+ scm_list_1 (str), err); \
+ else \
+ /* C_UTF is null-terminated. */ \
+ utf = scm_c_take_bytevector ((signed char *) c_utf,
\
+ c_utf_len); \
+ \
+ return (utf);
+
+
+
+SCM_DEFINE (scm_string_to_utf8, "string->utf8",
+ 1, 0, 0,
+ (SCM str),
+ "Return a newly allocated bytevector that contains the UTF-8 "
+ "encoding of @var{str}.")
+#define FUNC_NAME s_scm_string_to_utf8
+{
+ SCM utf;
+ char *c_str;
+ uint8_t *c_utf;
+ size_t c_strlen, c_raw_strlen;
+
+ SCM_VALIDATE_STRING (1, str);
+
+ c_strlen = scm_c_string_length (str);
+ c_raw_strlen = c_strlen;
+ do
+ {
+ c_str = (char *) alloca (c_raw_strlen + 1);
+ c_raw_strlen = scm_to_locale_stringbuf (str, c_str, c_strlen);
+ }
+ while (c_raw_strlen > c_strlen);
+ c_str[c_raw_strlen] = '\0';
+
+ c_utf = u8_strconv_from_locale (c_str);
+ if (SCM_UNLIKELY (c_utf == NULL))
+ scm_syserror (FUNC_NAME);
+ else
+ /* C_UTF is null-terminated. */
+ utf = scm_c_take_bytevector ((signed char *) c_utf,
+ UTF_STRLEN (8, c_utf));
+
+ return (utf);
+}
+#undef FUNC_NAME
+
+SCM_DEFINE (scm_string_to_utf16, "string->utf16",
+ 1, 1, 0,
+ (SCM str, SCM endianness),
+ "Return a newly allocated bytevector that contains the UTF-16 "
+ "encoding of @var{str}.")
+#define FUNC_NAME s_scm_string_to_utf16
+{
+ STRING_TO_UTF (16);
+}
+#undef FUNC_NAME
+
+SCM_DEFINE (scm_string_to_utf32, "string->utf32",
+ 1, 1, 0,
+ (SCM str, SCM endianness),
+ "Return a newly allocated bytevector that contains the UTF-32 "
+ "encoding of @var{str}.")
+#define FUNC_NAME s_scm_string_to_utf32
+{
+ STRING_TO_UTF (32);
+}
+#undef FUNC_NAME
+
+
+/* Produce the body of a function that converts a UTF-encoded bytevector to a
+ string. */
+#define UTF_TO_STRING(_utf_width) \
+ SCM str = SCM_BOOL_F;
\
+ int err; \
+ char *c_str = NULL, *c_locale; \
+ char c_utf_name[MAX_UTF_ENCODING_NAME_LEN]; \
+ const char *c_utf; \
+ size_t c_strlen = 0, c_utf_len; \
+ \
+ SCM_VALIDATE_BYTEVECTOR (1, utf); \
+ if (endianness == SCM_UNDEFINED) \
+ endianness = scm_sym_big; \
+ else \
+ SCM_VALIDATE_SYMBOL (2, endianness); \
+ \
+ c_utf_len = SCM_BYTEVECTOR_LENGTH (utf); \
+ c_utf = (char *) SCM_BYTEVECTOR_CONTENTS (utf); \
+ utf_encoding_name (c_utf_name, (_utf_width), endianness); \
+ \
+ c_locale = (char *) alloca (strlen (locale_charset ()) + 1); \
+ strcpy (c_locale, locale_charset ());
\
+ \
+ err = mem_iconveh (c_utf, c_utf_len, \
+ c_utf_name, c_locale, \
+ iconveh_question_mark, NULL, \
+ &c_str, &c_strlen); \
+ if (SCM_UNLIKELY (err)) \
+ scm_syserror_msg (FUNC_NAME, "failed to convert to string: ~A", \
+ scm_list_1 (utf), err); \
+ else \
+ /* C_STR is null-terminated. */ \
+ str = scm_take_locale_stringn (c_str, c_strlen); \
+ \
+ return (str);
+
+
+SCM_DEFINE (scm_utf8_to_string, "utf8->string",
+ 1, 0, 0,
+ (SCM utf),
+ "Return a newly allocate string that contains from the UTF-8-"
+ "encoded contents of bytevector @var{utf}.")
+#define FUNC_NAME s_scm_utf8_to_string
+{
+ SCM str;
+ int err;
+ char *c_str = NULL, *c_locale;
+ const char *c_utf;
+ size_t c_utf_len, c_strlen = 0;
+
+ SCM_VALIDATE_BYTEVECTOR (1, utf);
+
+ c_utf_len = SCM_BYTEVECTOR_LENGTH (utf);
+
+ c_locale = (char *) alloca (strlen (locale_charset ()) + 1);
+ strcpy (c_locale, locale_charset ());
+
+ c_utf = (char *) SCM_BYTEVECTOR_CONTENTS (utf);
+ err = mem_iconveh (c_utf, c_utf_len,
+ "UTF-8", c_locale,
+ iconveh_question_mark, NULL,
+ &c_str, &c_strlen);
+ if (SCM_UNLIKELY (err))
+ scm_syserror_msg (FUNC_NAME, "failed to convert to string: ~A",
+ scm_list_1 (utf), err);
+ else
+ /* C_STR is null-terminated. */
+ str = scm_take_locale_stringn (c_str, c_strlen);
+
+ return (str);
+}
+#undef FUNC_NAME
+
+SCM_DEFINE (scm_utf16_to_string, "utf16->string",
+ 1, 1, 0,
+ (SCM utf, SCM endianness),
+ "Return a newly allocate string that contains from the UTF-16-"
+ "encoded contents of bytevector @var{utf}.")
+#define FUNC_NAME s_scm_utf16_to_string
+{
+ UTF_TO_STRING (16);
+}
+#undef FUNC_NAME
+
+SCM_DEFINE (scm_utf32_to_string, "utf32->string",
+ 1, 1, 0,
+ (SCM utf, SCM endianness),
+ "Return a newly allocate string that contains from the UTF-32-"
+ "encoded contents of bytevector @var{utf}.")
+#define FUNC_NAME s_scm_utf32_to_string
+{
+ UTF_TO_STRING (32);
+}
+#undef FUNC_NAME
+
+
+
+/* Initialization. */
+
+void
+scm_init_bytevectors (void)
+{
+#include "libguile/bytevectors.x"
+
+#ifdef WORDS_BIGENDIAN
+ native_endianness = scm_sym_big;
+#else
+ native_endianness = scm_sym_little;
+#endif
+
+ scm_endianness_big = scm_sym_big;
+ scm_endianness_little = scm_sym_little;
+
+ scm_null_bytevector =
+ scm_gc_protect_object (make_bytevector_from_buffer (0, NULL));
+}
diff --git a/libguile/bytevectors.h b/libguile/bytevectors.h
new file mode 100644
index 0000000..98c38ac
--- /dev/null
+++ b/libguile/bytevectors.h
@@ -0,0 +1,133 @@
+#ifndef SCM_BYTEVECTORS_H
+#define SCM_BYTEVECTORS_H
+
+/* Copyright (C) 2009 Free Software Foundation, Inc.
+ *
+ * This library is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU Lesser General Public
+ * License as published by the Free Software Foundation; either
+ * version 2.1 of the License, or (at your option) any later version.
+ *
+ * This library is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+ * Lesser General Public License for more details.
+ *
+ * You should have received a copy of the GNU Lesser General Public
+ * License along with this library; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
+ */
+
+
+
+#include "libguile/__scm.h"
+
+
+/* R6RS bytevectors. */
+
+#define SCM_BYTEVECTOR_LENGTH(_bv) \
+ ((unsigned) SCM_SMOB_DATA (_bv))
+#define SCM_BYTEVECTOR_CONTENTS(_bv) \
+ (SCM_BYTEVECTOR_INLINE_P (_bv) \
+ ? (signed char *) SCM_SMOB_OBJECT_2_LOC (_bv) \
+ : (signed char *) SCM_SMOB_DATA_2 (_bv))
+
+
+SCM_API SCM scm_endianness_big;
+SCM_API SCM scm_endianness_little;
+
+SCM_API SCM scm_make_bytevector (SCM, SCM);
+SCM_API SCM scm_c_make_bytevector (unsigned);
+SCM_API SCM scm_native_endianness (void);
+SCM_API SCM scm_bytevector_p (SCM);
+SCM_API SCM scm_bytevector_length (SCM);
+SCM_API SCM scm_bytevector_eq_p (SCM, SCM);
+SCM_API SCM scm_bytevector_fill_x (SCM, SCM);
+SCM_API SCM scm_bytevector_copy_x (SCM, SCM, SCM, SCM, SCM);
+SCM_API SCM scm_bytevector_copy (SCM);
+
+SCM_API SCM scm_bytevector_to_u8_list (SCM);
+SCM_API SCM scm_u8_list_to_bytevector (SCM);
+SCM_API SCM scm_uint_list_to_bytevector (SCM, SCM, SCM);
+SCM_API SCM scm_bytevector_to_uint_list (SCM, SCM, SCM);
+SCM_API SCM scm_sint_list_to_bytevector (SCM, SCM, SCM);
+SCM_API SCM scm_bytevector_to_sint_list (SCM, SCM, SCM);
+
+SCM_API SCM scm_bytevector_u16_native_ref (SCM, SCM);
+SCM_API SCM scm_bytevector_s16_native_ref (SCM, SCM);
+SCM_API SCM scm_bytevector_u32_native_ref (SCM, SCM);
+SCM_API SCM scm_bytevector_s32_native_ref (SCM, SCM);
+SCM_API SCM scm_bytevector_u64_native_ref (SCM, SCM);
+SCM_API SCM scm_bytevector_s64_native_ref (SCM, SCM);
+SCM_API SCM scm_bytevector_u8_ref (SCM, SCM);
+SCM_API SCM scm_bytevector_s8_ref (SCM, SCM);
+SCM_API SCM scm_bytevector_uint_ref (SCM, SCM, SCM, SCM);
+SCM_API SCM scm_bytevector_sint_ref (SCM, SCM, SCM, SCM);
+SCM_API SCM scm_bytevector_u16_ref (SCM, SCM, SCM);
+SCM_API SCM scm_bytevector_s16_ref (SCM, SCM, SCM);
+SCM_API SCM scm_bytevector_u32_ref (SCM, SCM, SCM);
+SCM_API SCM scm_bytevector_s32_ref (SCM, SCM, SCM);
+SCM_API SCM scm_bytevector_u64_ref (SCM, SCM, SCM);
+SCM_API SCM scm_bytevector_s64_ref (SCM, SCM, SCM);
+SCM_API SCM scm_bytevector_u16_native_set_x (SCM, SCM, SCM);
+SCM_API SCM scm_bytevector_s16_native_set_x (SCM, SCM, SCM);
+SCM_API SCM scm_bytevector_u32_native_set_x (SCM, SCM, SCM);
+SCM_API SCM scm_bytevector_s32_native_set_x (SCM, SCM, SCM);
+SCM_API SCM scm_bytevector_u64_native_set_x (SCM, SCM, SCM);
+SCM_API SCM scm_bytevector_s64_native_set_x (SCM, SCM, SCM);
+SCM_API SCM scm_bytevector_u8_set_x (SCM, SCM, SCM);
+SCM_API SCM scm_bytevector_s8_set_x (SCM, SCM, SCM);
+SCM_API SCM scm_bytevector_uint_set_x (SCM, SCM, SCM, SCM, SCM);
+SCM_API SCM scm_bytevector_sint_set_x (SCM, SCM, SCM, SCM, SCM);
+SCM_API SCM scm_bytevector_u16_set_x (SCM, SCM, SCM, SCM);
+SCM_API SCM scm_bytevector_s16_set_x (SCM, SCM, SCM, SCM);
+SCM_API SCM scm_bytevector_u32_set_x (SCM, SCM, SCM, SCM);
+SCM_API SCM scm_bytevector_s32_set_x (SCM, SCM, SCM, SCM);
+SCM_API SCM scm_bytevector_u64_set_x (SCM, SCM, SCM, SCM);
+SCM_API SCM scm_bytevector_s64_set_x (SCM, SCM, SCM, SCM);
+SCM_API SCM scm_bytevector_ieee_single_ref (SCM, SCM, SCM);
+SCM_API SCM scm_bytevector_ieee_single_native_ref (SCM, SCM);
+SCM_API SCM scm_bytevector_ieee_single_set_x (SCM, SCM, SCM, SCM);
+SCM_API SCM scm_bytevector_ieee_single_native_set_x (SCM, SCM, SCM);
+SCM_API SCM scm_bytevector_ieee_double_ref (SCM, SCM, SCM);
+SCM_API SCM scm_bytevector_ieee_double_native_ref (SCM, SCM);
+SCM_API SCM scm_bytevector_ieee_double_set_x (SCM, SCM, SCM, SCM);
+SCM_API SCM scm_bytevector_ieee_double_native_set_x (SCM, SCM, SCM);
+SCM_API SCM scm_string_to_utf8 (SCM);
+SCM_API SCM scm_string_to_utf16 (SCM, SCM);
+SCM_API SCM scm_string_to_utf32 (SCM, SCM);
+SCM_API SCM scm_utf8_to_string (SCM);
+SCM_API SCM scm_utf16_to_string (SCM, SCM);
+SCM_API SCM scm_utf32_to_string (SCM, SCM);
+
+
+
+/* Internal API. */
+
+/* The threshold (in octets) under which bytevectors are stored "in-line",
+ i.e., without allocating memory beside the SMOB itself (a double cell).
+ This optimization is necessary since small bytevectors are expected to be
+ common. */
+#define SCM_BYTEVECTOR_INLINE_THRESHOLD (2 * sizeof (SCM))
+#define SCM_BYTEVECTOR_INLINEABLE_SIZE_P(_size) \
+ ((_size) <= SCM_BYTEVECTOR_INLINE_THRESHOLD)
+#define SCM_BYTEVECTOR_INLINE_P(_bv) \
+ (SCM_BYTEVECTOR_INLINEABLE_SIZE_P (SCM_BYTEVECTOR_LENGTH (_bv)))
+
+/* Hint that is passed to `scm_gc_malloc ()' and friends. */
+#define SCM_GC_BYTEVECTOR "bytevector"
+
+SCM_API void scm_init_bytevectors (void);
+
+SCM_INTERNAL scm_t_bits scm_tc16_bytevector;
+SCM_INTERNAL SCM scm_c_take_bytevector (signed char *, unsigned);
+
+#define scm_c_shrink_bytevector(_bv, _len) \
+ (SCM_BYTEVECTOR_INLINE_P (_bv) \
+ ? (_bv) \
+ : scm_i_shrink_bytevector ((_bv), (_len)))
+
+SCM_INTERNAL SCM scm_i_shrink_bytevector (SCM, unsigned);
+SCM_INTERNAL SCM scm_null_bytevector;
+
+#endif /* SCM_BYTEVECTORS_H */
diff --git a/libguile/ieee-754.h b/libguile/ieee-754.h
new file mode 100644
index 0000000..e345efa
--- /dev/null
+++ b/libguile/ieee-754.h
@@ -0,0 +1,90 @@
+/* Copyright (C) 1992, 1995, 1996, 1999 Free Software Foundation, Inc.
+ This file is part of the GNU C Library.
+
+ The GNU C Library is free software; you can redistribute it and/or
+ modify it under the terms of the GNU Lesser General Public
+ License as published by the Free Software Foundation; either
+ version 2.1 of the License, or (at your option) any later version.
+
+ The GNU C Library is distributed in the hope that it will be useful,
+ but WITHOUT ANY WARRANTY; without even the implied warranty of
+ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+ Lesser General Public License for more details.
+
+ You should have received a copy of the GNU Lesser General Public
+ License along with the GNU C Library; if not, write to the Free
+ Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
+ 02111-1307 USA. */
+
+#ifndef SCM_IEEE_754_H
+#define SCM_IEEE_754_H 1
+
+/* Based on glibc's <ieee754.h> and modified by Ludovic Courtès to include
+ all possible IEEE-754 double-precision representations. */
+
+
+/* IEEE 754 simple-precision format (32-bit). */
+
+union scm_ieee754_float
+ {
+ float f;
+
+ struct
+ {
+ unsigned int negative:1;
+ unsigned int exponent:8;
+ unsigned int mantissa:23;
+ } big_endian;
+
+ struct
+ {
+ unsigned int mantissa:23;
+ unsigned int exponent:8;
+ unsigned int negative:1;
+ } little_endian;
+ };
+
+
+
+/* IEEE 754 double-precision format (64-bit). */
+
+union scm_ieee754_double
+ {
+ double d;
+
+ struct
+ {
+ /* Big endian. */
+
+ unsigned int negative:1;
+ unsigned int exponent:11;
+ /* Together these comprise the mantissa. */
+ unsigned int mantissa0:20;
+ unsigned int mantissa1:32;
+ } big_endian;
+
+ struct
+ {
+ /* Both byte order and word order are little endian. */
+
+ /* Together these comprise the mantissa. */
+ unsigned int mantissa1:32;
+ unsigned int mantissa0:20;
+ unsigned int exponent:11;
+ unsigned int negative:1;
+ } little_little_endian;
+
+ struct
+ {
+ /* Byte order is little endian but word order is big endian. Not
+ sure this is very wide spread. */
+ unsigned int mantissa0:20;
+ unsigned int exponent:11;
+ unsigned int negative:1;
+ unsigned int mantissa1:32;
+ } little_big_endian;
+
+ };
+
+
+#endif /* SCM_IEEE_754_H */
diff --git a/libguile/r6rs-ports.c b/libguile/r6rs-ports.c
new file mode 100644
index 0000000..a07636f
--- /dev/null
+++ b/libguile/r6rs-ports.c
@@ -0,0 +1,1118 @@
+/* Copyright (C) 2009 Free Software Foundation, Inc.
+ *
+ * This library is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU Lesser General Public
+ * License as published by the Free Software Foundation; either
+ * version 2.1 of the License, or (at your option) any later version.
+ *
+ * This library is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+ * Lesser General Public License for more details.
+ *
+ * You should have received a copy of the GNU Lesser General Public
+ * License along with this library; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
+ */
+
+#ifdef HAVE_CONFIG_H
+# include <config.h>
+#endif
+
+#ifdef HAVE_UNISTD_H
+# include <unistd.h>
+#endif
+
+#include <string.h>
+#include <stdio.h>
+#include <assert.h>
+
+#include "libguile/_scm.h"
+#include "libguile/bytevectors.h"
+#include "libguile/chars.h"
+#include "libguile/eval.h"
+#include "libguile/r6rs-ports.h"
+#include "libguile/strings.h"
+#include "libguile/validate.h"
+#include "libguile/values.h"
+#include "libguile/vectors.h"
+
+
+
+/* Unimplemented features. */
+
+
+/* Transoders are currently not implemented since Guile 1.8 is not
+ Unicode-capable. Thus, most of the code here assumes the use of the
+ binary transcoder. */
+static inline void
+transcoders_not_implemented (void)
+{
+ fprintf (stderr, "%s: warning: transcoders not implemented\n",
+ PACKAGE_NAME);
+}
+
+
+/* End-of-file object. */
+
+SCM_DEFINE (scm_eof_object, "eof-object", 0, 0, 0,
+ (void),
+ "Return the end-of-file object.")
+#define FUNC_NAME s_scm_eof_object
+{
+ return (SCM_EOF_VAL);
+}
+#undef FUNC_NAME
+
+
+/* Input ports. */
+
+#ifndef MIN
+# define MIN(a,b) ((a) < (b) ? (a) : (b))
+#endif
+
+/* Bytevector input ports or "bip" for short. */
+static scm_t_bits bytevector_input_port_type = 0;
+
+static inline SCM
+make_bip (SCM bv)
+{
+ SCM port;
+ char *c_bv;
+ unsigned c_len;
+ scm_t_port *c_port;
+ const unsigned long mode_bits = SCM_OPN | SCM_RDNG;
+
+ port = scm_new_port_table_entry (bytevector_input_port_type);
+
+ /* Prevent BV from being GC'd. */
+ SCM_SETSTREAM (port, SCM_UNPACK (bv));
+
+ /* Have the port directly access the bytevector. */
+ c_bv = (char *) SCM_BYTEVECTOR_CONTENTS (bv);
+ c_len = SCM_BYTEVECTOR_LENGTH (bv);
+
+ c_port = SCM_PTAB_ENTRY (port);
+ c_port->read_pos = c_port->read_buf = (unsigned char *) c_bv;
+ c_port->read_end = (unsigned char *) c_bv + c_len;
+ c_port->read_buf_size = c_len;
+
+ /* Mark PORT as open, readable and unbuffered (hmm, how elegant...). */
+ SCM_SET_CELL_TYPE (port, bytevector_input_port_type | mode_bits);
+
+ return port;
+}
+
+static SCM
+bip_mark (SCM port)
+{
+ /* Mark the underlying bytevector. */
+ return (SCM_PACK (SCM_STREAM (port)));
+}
+
+static int
+bip_fill_input (SCM port)
+{
+ int result;
+ scm_t_port *c_port = SCM_PTAB_ENTRY (port);
+
+ if (c_port->read_pos >= c_port->read_end)
+ result = EOF;
+ else
+ result = (int) *c_port->read_pos;
+
+ return result;
+}
+
+static off_t
+bip_seek (SCM port, off_t offset, int whence)
+#define FUNC_NAME "bip_seek"
+{
+ off_t c_result = 0;
+ scm_t_port *c_port = SCM_PTAB_ENTRY (port);
+
+ switch (whence)
+ {
+ case SEEK_CUR:
+ offset += c_port->read_pos - c_port->read_buf;
+ /* Fall through. */
+
+ case SEEK_SET:
+ if (c_port->read_buf + offset < c_port->read_end)
+ {
+ c_port->read_pos = c_port->read_buf + offset;
+ c_result = offset;
+ }
+ else
+ scm_out_of_range (FUNC_NAME, scm_from_int (offset));
+ break;
+
+ case SEEK_END:
+ if (c_port->read_end - offset >= c_port->read_buf)
+ {
+ c_port->read_pos = c_port->read_end - offset;
+ c_result = c_port->read_pos - c_port->read_buf;
+ }
+ else
+ scm_out_of_range (FUNC_NAME, scm_from_int (offset));
+ break;
+
+ default:
+ scm_wrong_type_arg_msg (FUNC_NAME, 0, port,
+ "invalid `seek' parameter");
+ }
+
+ return c_result;
+}
+#undef FUNC_NAME
+
+
+/* Instantiate the bytevector input port type. */
+static inline void
+initialize_bytevector_input_ports (void)
+{
+ bytevector_input_port_type =
+ scm_make_port_type ("r6rs-bytevector-input-port", bip_fill_input,
+ NULL);
+
+ scm_set_port_mark (bytevector_input_port_type, bip_mark);
+ scm_set_port_seek (bytevector_input_port_type, bip_seek);
+}
+
+
+SCM_DEFINE (scm_open_bytevector_input_port,
+ "open-bytevector-input-port", 1, 1, 0,
+ (SCM bv, SCM transcoder),
+ "Return an input port whose contents are drawn from "
+ "bytevector @var{bv}.")
+#define FUNC_NAME s_scm_open_bytevector_input_port
+{
+ SCM_VALIDATE_BYTEVECTOR (1, bv);
+ if (!SCM_UNBNDP (transcoder) && !scm_is_false (transcoder))
+ transcoders_not_implemented ();
+
+ return (make_bip (bv));
+}
+#undef FUNC_NAME
+
+
+/* Custom binary ports. The following routines are shared by input and
+ output custom binary ports. */
+
+#define SCM_CBP_GET_POSITION_PROC(_port) \
+ SCM_SIMPLE_VECTOR_REF (SCM_PACK (SCM_STREAM (_port)), 1)
+#define SCM_CBP_SET_POSITION_PROC(_port) \
+ SCM_SIMPLE_VECTOR_REF (SCM_PACK (SCM_STREAM (_port)), 2)
+#define SCM_CBP_CLOSE_PROC(_port) \
+ SCM_SIMPLE_VECTOR_REF (SCM_PACK (SCM_STREAM (_port)), 3)
+
+static SCM
+cbp_mark (SCM port)
+{
+ /* Mark the underlying method and object vector. */
+ return (SCM_PACK (SCM_STREAM (port)));
+}
+
+static off_t
+cbp_seek (SCM port, off_t offset, int whence)
+#define FUNC_NAME "cbp_seek"
+{
+ SCM result;
+ off_t c_result = 0;
+
+ switch (whence)
+ {
+ case SEEK_CUR:
+ {
+ SCM get_position_proc;
+
+ get_position_proc = SCM_CBP_GET_POSITION_PROC (port);
+ if (SCM_LIKELY (scm_is_true (get_position_proc)))
+ result = scm_call_0 (get_position_proc);
+ else
+ scm_wrong_type_arg_msg (FUNC_NAME, 0, port,
+ "R6RS custom binary port does not "
+ "support `port-position'");
+
+ offset += scm_to_int (result);
+ /* Fall through. */
+ }
+
+ case SEEK_SET:
+ {
+ SCM set_position_proc;
+
+ set_position_proc = SCM_CBP_SET_POSITION_PROC (port);
+ if (SCM_LIKELY (scm_is_true (set_position_proc)))
+ result = scm_call_1 (set_position_proc, scm_from_int (offset));
+ else
+ scm_wrong_type_arg_msg (FUNC_NAME, 0, port,
+ "R6RS custom binary port does not "
+ "support `set-port-position!'");
+
+ /* Assuming setting the position succeeded. */
+ c_result = offset;
+ break;
+ }
+
+ default:
+ /* `SEEK_END' cannot be supported. */
+ scm_wrong_type_arg_msg (FUNC_NAME, 0, port,
+ "R6RS custom binary ports do not "
+ "support `SEEK_END'");
+ }
+
+ return c_result;
+}
+#undef FUNC_NAME
+
+static int
+cbp_close (SCM port)
+{
+ SCM close_proc;
+
+ close_proc = SCM_CBP_CLOSE_PROC (port);
+ if (scm_is_true (close_proc))
+ /* Invoke the `close' thunk. */
+ scm_call_0 (close_proc);
+
+ return 1;
+}
+
+
+/* Custom binary input port ("cbip" for short). */
+
+static scm_t_bits custom_binary_input_port_type = 0;
+
+/* Size of the buffer embedded in custom binary input ports. */
+#define CBIP_BUFFER_SIZE 4096
+
+/* Return the bytevector associated with PORT. */
+#define SCM_CBIP_BYTEVECTOR(_port) \
+ SCM_SIMPLE_VECTOR_REF (SCM_PACK (SCM_STREAM (_port)), 4)
+
+/* Return the various procedures of PORT. */
+#define SCM_CBIP_READ_PROC(_port) \
+ SCM_SIMPLE_VECTOR_REF (SCM_PACK (SCM_STREAM (_port)), 0)
+
+
+static inline SCM
+make_cbip (SCM read_proc, SCM get_position_proc,
+ SCM set_position_proc, SCM close_proc)
+{
+ SCM port, bv, method_vector;
+ char *c_bv;
+ unsigned c_len;
+ scm_t_port *c_port;
+ const unsigned long mode_bits = SCM_OPN | SCM_RDNG;
+
+ /* Use a bytevector as the underlying buffer. */
+ c_len = CBIP_BUFFER_SIZE;
+ bv = scm_c_make_bytevector (c_len);
+ c_bv = (char *) SCM_BYTEVECTOR_CONTENTS (bv);
+
+ /* Store the various methods and bytevector in a vector. */
+ method_vector = scm_c_make_vector (5, SCM_BOOL_F);
+ SCM_SIMPLE_VECTOR_SET (method_vector, 4, bv);
+ SCM_SIMPLE_VECTOR_SET (method_vector, 0, read_proc);
+ SCM_SIMPLE_VECTOR_SET (method_vector, 1, get_position_proc);
+ SCM_SIMPLE_VECTOR_SET (method_vector, 2, set_position_proc);
+ SCM_SIMPLE_VECTOR_SET (method_vector, 3, close_proc);
+
+ port = scm_new_port_table_entry (custom_binary_input_port_type);
+
+ /* Attach it the method vector. */
+ SCM_SETSTREAM (port, SCM_UNPACK (method_vector));
+
+ /* Have the port directly access the buffer (bytevector). */
+ c_port = SCM_PTAB_ENTRY (port);
+ c_port->read_pos = c_port->read_buf = (unsigned char *) c_bv;
+ c_port->read_end = (unsigned char *) c_bv;
+ c_port->read_buf_size = c_len;
+
+ /* Mark PORT as open, readable and unbuffered (hmm, how elegant...). */
+ SCM_SET_CELL_TYPE (port, custom_binary_input_port_type | mode_bits);
+
+ return port;
+}
+
+static int
+cbip_fill_input (SCM port)
+#define FUNC_NAME "cbip_fill_input"
+{
+ int result;
+ scm_t_port *c_port = SCM_PTAB_ENTRY (port);
+
+ again:
+ if (c_port->read_pos >= c_port->read_end)
+ {
+ /* Invoke the user's `read!' procedure. */
+ unsigned c_octets;
+ SCM bv, read_proc, octets;
+
+ /* Use the bytevector associated with PORT as the buffer passed to the
+ `read!' procedure, thereby avoiding additional allocations. */
+ bv = SCM_CBIP_BYTEVECTOR (port);
+ read_proc = SCM_CBIP_READ_PROC (port);
+
+ /* The assumption here is that C_PORT's internal buffer wasn't changed
+ behind our back. */
+ assert (c_port->read_buf ==
+ (unsigned char *) SCM_BYTEVECTOR_CONTENTS (bv));
+ assert ((unsigned) c_port->read_buf_size
+ == SCM_BYTEVECTOR_LENGTH (bv));
+
+ octets = scm_call_3 (read_proc, bv, SCM_INUM0,
+ SCM_I_MAKINUM (CBIP_BUFFER_SIZE));
+ c_octets = scm_to_uint (octets);
+
+ c_port->read_pos = (unsigned char *) SCM_BYTEVECTOR_CONTENTS (bv);
+ c_port->read_end = (unsigned char *) c_port->read_pos + c_octets;
+
+ if (c_octets > 0)
+ goto again;
+ else
+ result = EOF;
+ }
+ else
+ result = (int) *c_port->read_pos;
+
+ return result;
+}
+#undef FUNC_NAME
+
+
+SCM_DEFINE (scm_make_custom_binary_input_port,
+ "make-custom-binary-input-port", 5, 0, 0,
+ (SCM id, SCM read_proc, SCM get_position_proc,
+ SCM set_position_proc, SCM close_proc),
+ "Return a new custom binary input port whose input is drained "
+ "by invoking @var{read_proc} and passing it a bytevector, an "
+ "index where octets should be written, and an octet count.")
+#define FUNC_NAME s_scm_make_custom_binary_input_port
+{
+ SCM_VALIDATE_STRING (1, id);
+ SCM_VALIDATE_PROC (2, read_proc);
+
+ if (!scm_is_false (get_position_proc))
+ SCM_VALIDATE_PROC (3, get_position_proc);
+
+ if (!scm_is_false (set_position_proc))
+ SCM_VALIDATE_PROC (4, set_position_proc);
+
+ if (!scm_is_false (close_proc))
+ SCM_VALIDATE_PROC (5, close_proc);
+
+ return (make_cbip (read_proc, get_position_proc, set_position_proc,
+ close_proc));
+}
+#undef FUNC_NAME
+
+
+/* Instantiate the custom binary input port type. */
+static inline void
+initialize_custom_binary_input_ports (void)
+{
+ custom_binary_input_port_type =
+ scm_make_port_type ("r6rs-custom-binary-input-port",
+ cbip_fill_input, NULL);
+
+ scm_set_port_mark (custom_binary_input_port_type, cbp_mark);
+ scm_set_port_seek (custom_binary_input_port_type, cbp_seek);
+ scm_set_port_close (custom_binary_input_port_type, cbp_close);
+}
+
+
+
+/* Binary input. */
+
+/* We currently don't support specific binary input ports. */
+#define SCM_VALIDATE_BINARY_INPUT_PORT SCM_VALIDATE_OPINPORT
+
+SCM_DEFINE (scm_get_u8, "get-u8", 1, 0, 0,
+ (SCM port),
+ "Read an octet from @var{port}, a binary input port, "
+ "blocking as necessary.")
+#define FUNC_NAME s_scm_get_u8
+{
+ SCM result;
+ int c_result;
+
+ SCM_VALIDATE_BINARY_INPUT_PORT (1, port);
+
+ c_result = scm_getc (port);
+ if (c_result == EOF)
+ result = SCM_EOF_VAL;
+ else
+ result = SCM_I_MAKINUM ((unsigned char) c_result);
+
+ return result;
+}
+#undef FUNC_NAME
+
+SCM_DEFINE (scm_lookahead_u8, "lookahead-u8", 1, 0, 0,
+ (SCM port),
+ "Like @code{get-u8} but does not update @var{port} to "
+ "point past the octet.")
+#define FUNC_NAME s_scm_lookahead_u8
+{
+ SCM result;
+
+ SCM_VALIDATE_BINARY_INPUT_PORT (1, port);
+
+ result = scm_peek_char (port);
+ if (SCM_CHARP (result))
+ result = SCM_I_MAKINUM ((signed char) SCM_CHAR (result));
+ else
+ result = SCM_EOF_VAL;
+
+ return result;
+}
+#undef FUNC_NAME
+
+SCM_DEFINE (scm_get_bytevector_n, "get-bytevector-n", 2, 0, 0,
+ (SCM port, SCM count),
+ "Read @var{count} octets from @var{port}, blocking as "
+ "necessary and return a bytevector containing the octets "
+ "read. If fewer bytes are available, a bytevector smaller "
+ "than @var{count} is returned.")
+#define FUNC_NAME s_scm_get_bytevector_n
+{
+ SCM result;
+ char *c_bv;
+ unsigned c_count;
+ size_t c_read;
+
+ SCM_VALIDATE_BINARY_INPUT_PORT (1, port);
+ c_count = scm_to_uint (count);
+
+ result = scm_c_make_bytevector (c_count);
+ c_bv = (char *) SCM_BYTEVECTOR_CONTENTS (result);
+
+ if (SCM_LIKELY (c_count > 0))
+ /* XXX: `scm_c_read ()' does not update the port position. */
+ c_read = scm_c_read (port, c_bv, c_count);
+ else
+ /* Don't invoke `scm_c_read ()' since it may block. */
+ c_read = 0;
+
+ if ((c_read == 0) && (c_count > 0))
+ {
+ if (SCM_EOF_OBJECT_P (scm_peek_char (port)))
+ result = SCM_EOF_VAL;
+ else
+ result = scm_null_bytevector;
+ }
+ else
+ {
+ if (c_read < c_count)
+ result = scm_c_shrink_bytevector (result, c_read);
+ }
+
+ return result;
+}
+#undef FUNC_NAME
+
+SCM_DEFINE (scm_get_bytevector_n_x, "get-bytevector-n!", 4, 0, 0,
+ (SCM port, SCM bv, SCM start, SCM count),
+ "Read @var{count} bytes from @var{port} and store them "
+ "in @var{bv} starting at index @var{start}. Return either "
+ "the number of bytes actually read or the end-of-file "
+ "object.")
+#define FUNC_NAME s_scm_get_bytevector_n_x
+{
+ SCM result;
+ char *c_bv;
+ unsigned c_start, c_count, c_len;
+ size_t c_read;
+
+ SCM_VALIDATE_BINARY_INPUT_PORT (1, port);
+ SCM_VALIDATE_BYTEVECTOR (2, bv);
+ c_start = scm_to_uint (start);
+ c_count = scm_to_uint (count);
+
+ c_bv = (char *) SCM_BYTEVECTOR_CONTENTS (bv);
+ c_len = SCM_BYTEVECTOR_LENGTH (bv);
+
+ if (SCM_UNLIKELY (c_start + c_count > c_len))
+ scm_out_of_range (FUNC_NAME, count);
+
+ if (SCM_LIKELY (c_count > 0))
+ c_read = scm_c_read (port, c_bv + c_start, c_count);
+ else
+ /* Don't invoke `scm_c_read ()' since it may block. */
+ c_read = 0;
+
+ if ((c_read == 0) && (c_count > 0))
+ {
+ if (SCM_EOF_OBJECT_P (scm_peek_char (port)))
+ result = SCM_EOF_VAL;
+ else
+ result = SCM_I_MAKINUM (0);
+ }
+ else
+ result = scm_from_size_t (c_read);
+
+ return result;
+}
+#undef FUNC_NAME
+
+
+SCM_DEFINE (scm_get_bytevector_some, "get-bytevector-some", 1, 0, 0,
+ (SCM port),
+ "Read from @var{port}, blocking as necessary, until data "
+ "are available or and end-of-file is reached. Return either "
+ "a new bytevector containing the data read or the "
+ "end-of-file object.")
+#define FUNC_NAME s_scm_get_bytevector_some
+{
+ /* Read at least one byte, unless the end-of-file is already reached, and
+ read while characters are available (buffered). */
+
+ SCM result;
+ char *c_bv;
+ unsigned c_len;
+ size_t c_total;
+
+ SCM_VALIDATE_BINARY_INPUT_PORT (1, port);
+
+ c_len = 4096;
+ c_bv = (char *) scm_gc_malloc (c_len, SCM_GC_BYTEVECTOR);
+ c_total = 0;
+
+ do
+ {
+ int c_chr;
+
+ if (c_total + 1 > c_len)
+ {
+ /* Grow the bytevector. */
+ c_bv = (char *) scm_gc_realloc (c_bv, c_len, c_len * 2,
+ SCM_GC_BYTEVECTOR);
+ c_len *= 2;
+ }
+
+ /* We can't use `scm_c_read ()' since it blocks. */
+ c_chr = scm_getc (port);
+ if (c_chr != EOF)
+ {
+ c_bv[c_total] = (char) c_chr;
+ c_total++;
+ }
+ }
+ while ((scm_is_true (scm_char_ready_p (port)))
+ && (!SCM_EOF_OBJECT_P (scm_peek_char (port))));
+
+ if (c_total == 0)
+ {
+ result = SCM_EOF_VAL;
+ scm_gc_free (c_bv, c_len, SCM_GC_BYTEVECTOR);
+ }
+ else
+ {
+ if (c_len > c_total)
+ {
+ /* Shrink the bytevector. */
+ c_bv = (char *) scm_gc_realloc (c_bv, c_len, c_total,
+ SCM_GC_BYTEVECTOR);
+ c_len = (unsigned) c_total;
+ }
+
+ result = scm_c_take_bytevector ((signed char *) c_bv, c_len);
+ }
+
+ return result;
+}
+#undef FUNC_NAME
+
+SCM_DEFINE (scm_get_bytevector_all, "get-bytevector-all", 1, 0, 0,
+ (SCM port),
+ "Read from @var{port}, blocking as necessary, until "
+ "the end-of-file is reached. Return either "
+ "a new bytevector containing the data read or the "
+ "end-of-file object (if no data were available).")
+#define FUNC_NAME s_scm_get_bytevector_all
+{
+ SCM result;
+ char *c_bv;
+ unsigned c_len, c_count;
+ size_t c_read, c_total;
+
+ SCM_VALIDATE_BINARY_INPUT_PORT (1, port);
+
+ c_len = c_count = 4096;
+ c_bv = (char *) scm_gc_malloc (c_len, SCM_GC_BYTEVECTOR);
+ c_total = c_read = 0;
+
+ do
+ {
+ if (c_total + c_read > c_len)
+ {
+ /* Grow the bytevector. */
+ c_bv = (char *) scm_gc_realloc (c_bv, c_len, c_len * 2,
+ SCM_GC_BYTEVECTOR);
+ c_count = c_len;
+ c_len *= 2;
+ }
+
+ /* `scm_c_read ()' blocks until C_COUNT bytes are available or EOF is
+ reached. */
+ c_read = scm_c_read (port, c_bv + c_total, c_count);
+ c_total += c_read, c_count -= c_read;
+ }
+ while (!SCM_EOF_OBJECT_P (scm_peek_char (port)));
+
+ if (c_total == 0)
+ {
+ result = SCM_EOF_VAL;
+ scm_gc_free (c_bv, c_len, SCM_GC_BYTEVECTOR);
+ }
+ else
+ {
+ if (c_len > c_total)
+ {
+ /* Shrink the bytevector. */
+ c_bv = (char *) scm_gc_realloc (c_bv, c_len, c_total,
+ SCM_GC_BYTEVECTOR);
+ c_len = (unsigned) c_total;
+ }
+
+ result = scm_c_take_bytevector ((signed char *) c_bv, c_len);
+ }
+
+ return result;
+}
+#undef FUNC_NAME
+
+
+
+/* Binary output. */
+
+/* We currently don't support specific binary input ports. */
+#define SCM_VALIDATE_BINARY_OUTPUT_PORT SCM_VALIDATE_OPOUTPORT
+
+
+SCM_DEFINE (scm_put_u8, "put-u8", 2, 0, 0,
+ (SCM port, SCM octet),
+ "Write @var{octet} to binary port @var{port}.")
+#define FUNC_NAME s_scm_put_u8
+{
+ scm_t_uint8 c_octet;
+
+ SCM_VALIDATE_BINARY_OUTPUT_PORT (1, port);
+ c_octet = scm_to_uint8 (octet);
+
+ scm_putc ((char) c_octet, port);
+
+ return SCM_UNSPECIFIED;
+}
+#undef FUNC_NAME
+
+SCM_DEFINE (scm_put_bytevector, "put-bytevector", 2, 2, 0,
+ (SCM port, SCM bv, SCM start, SCM count),
+ "Write the contents of @var{bv} to @var{port}, optionally "
+ "starting at index @var{start} and limiting to @var{count} "
+ "octets.")
+#define FUNC_NAME s_scm_put_bytevector
+{
+ char *c_bv;
+ unsigned c_start, c_count, c_len;
+
+ SCM_VALIDATE_BINARY_OUTPUT_PORT (1, port);
+ SCM_VALIDATE_BYTEVECTOR (2, bv);
+
+ c_len = SCM_BYTEVECTOR_LENGTH (bv);
+ c_bv = (char *) SCM_BYTEVECTOR_CONTENTS (bv);
+
+ if (start != SCM_UNDEFINED)
+ {
+ c_start = scm_to_uint (start);
+
+ if (count != SCM_UNDEFINED)
+ {
+ c_count = scm_to_uint (count);
+ if (SCM_UNLIKELY (c_start + c_count > c_len))
+ scm_out_of_range (FUNC_NAME, count);
+ }
+ else
+ {
+ if (SCM_UNLIKELY (c_start >= c_len))
+ scm_out_of_range (FUNC_NAME, start);
+ else
+ c_count = c_len - c_start;
+ }
+ }
+ else
+ c_start = 0, c_count = c_len;
+
+ scm_c_write (port, c_bv + c_start, c_count);
+
+ return SCM_UNSPECIFIED;
+}
+#undef FUNC_NAME
+
+
+
+/* Bytevector output port ("bop" for short). */
+
+/* Implementation of "bops".
+
+ Each bop has an internal buffer, of type `scm_t_bop_buffer', attached to
+ it. The procedure returned along with the output port is actually an
+ applicable SMOB. The SMOB holds a reference to the port. When applied,
+ the SMOB swallows the port's internal buffer, turning it into a
+ bytevector, and resets it.
+
+ XXX: Access to a bop's internal buffer is not thread-safe. */
+
+static scm_t_bits bytevector_output_port_type = 0;
+
+SCM_SMOB (bytevector_output_port_procedure,
+ "r6rs-bytevector-output-port-procedure",
+ 0);
+
+#define SCM_GC_BOP "r6rs-bytevector-output-port"
+#define SCM_BOP_BUFFER_INITIAL_SIZE 4096
+
+/* Representation of a bop's internal buffer. */
+typedef struct
+{
+ size_t total_len;
+ size_t len;
+ size_t pos;
+ char *buffer;
+} scm_t_bop_buffer;
+
+
+/* Accessing a bop's buffer. */
+#define SCM_BOP_BUFFER(_port) \
+ ((scm_t_bop_buffer *) SCM_STREAM (_port))
+#define SCM_SET_BOP_BUFFER(_port, _buf) \
+ (SCM_SETSTREAM ((_port), (scm_t_bits) (_buf)))
+
+
+static inline void
+bop_buffer_init (scm_t_bop_buffer *buf)
+{
+ buf->total_len = buf->len = buf->pos = 0;
+ buf->buffer = NULL;
+}
+
+static inline void
+bop_buffer_grow (scm_t_bop_buffer *buf, size_t min_size)
+{
+ char *new_buf;
+ size_t new_size;
+
+ for (new_size = buf->total_len
+ ? buf->total_len : SCM_BOP_BUFFER_INITIAL_SIZE;
+ new_size < min_size;
+ new_size *= 2);
+
+ if (buf->buffer)
+ new_buf = scm_gc_realloc ((void *) buf->buffer, buf->total_len,
+ new_size, SCM_GC_BOP);
+ else
+ new_buf = scm_gc_malloc (new_size, SCM_GC_BOP);
+
+ buf->buffer = new_buf;
+ buf->total_len = new_size;
+}
+
+static inline SCM
+make_bop (void)
+{
+ SCM port, bop_proc;
+ scm_t_port *c_port;
+ scm_t_bop_buffer *buf;
+ const unsigned long mode_bits = SCM_OPN | SCM_WRTNG;
+
+ port = scm_new_port_table_entry (bytevector_output_port_type);
+
+ buf = (scm_t_bop_buffer *) scm_gc_malloc (sizeof (* buf), SCM_GC_BOP);
+ bop_buffer_init (buf);
+
+ c_port = SCM_PTAB_ENTRY (port);
+ c_port->write_buf = c_port->write_pos = c_port->write_end = NULL;
+ c_port->write_buf_size = 0;
+
+ SCM_SET_BOP_BUFFER (port, buf);
+
+ /* Mark PORT as open and writable. */
+ SCM_SET_CELL_TYPE (port, bytevector_output_port_type | mode_bits);
+
+ /* Make the bop procedure. */
+ SCM_NEWSMOB (bop_proc, bytevector_output_port_procedure,
+ SCM_PACK (port));
+
+ return (scm_values (scm_list_2 (port, bop_proc)));
+}
+
+static size_t
+bop_free (SCM port)
+{
+ /* The port itself is necessarily freed _after_ the bop proc, since the bop
+ proc holds a reference to it. Thus we can safely free the internal
+ buffer when the bop becomes unreferenced. */
+ scm_t_bop_buffer *buf;
+
+ buf = SCM_BOP_BUFFER (port);
+ if (buf->buffer)
+ scm_gc_free (buf->buffer, buf->total_len, SCM_GC_BOP);
+
+ scm_gc_free (buf, sizeof (* buf), SCM_GC_BOP);
+
+ return 0;
+}
+
+/* Write SIZE octets from DATA to PORT. */
+static void
+bop_write (SCM port, const void *data, size_t size)
+{
+ scm_t_bop_buffer *buf;
+
+ buf = SCM_BOP_BUFFER (port);
+
+ if (buf->pos + size > buf->total_len)
+ bop_buffer_grow (buf, buf->pos + size);
+
+ memcpy (buf->buffer + buf->pos, data, size);
+ buf->pos += size;
+ buf->len = (buf->len > buf->pos) ? buf->len : buf->pos;
+}
+
+static off_t
+bop_seek (SCM port, off_t offset, int whence)
+#define FUNC_NAME "bop_seek"
+{
+ scm_t_bop_buffer *buf;
+
+ buf = SCM_BOP_BUFFER (port);
+ switch (whence)
+ {
+ case SEEK_CUR:
+ offset += (off_t) buf->pos;
+ /* Fall through. */
+
+ case SEEK_SET:
+ if (offset < 0 || (unsigned) offset > buf->len)
+ scm_out_of_range (FUNC_NAME, scm_from_int (offset));
+ else
+ buf->pos = offset;
+ break;
+
+ case SEEK_END:
+ if (offset < 0 || (unsigned) offset >= buf->len)
+ scm_out_of_range (FUNC_NAME, scm_from_int (offset));
+ else
+ buf->pos = buf->len - (offset + 1);
+ break;
+
+ default:
+ scm_wrong_type_arg_msg (FUNC_NAME, 0, port,
+ "invalid `seek' parameter");
+ }
+
+ return buf->pos;
+}
+#undef FUNC_NAME
+
+/* Fetch data from a bop. */
+SCM_SMOB_APPLY (bytevector_output_port_procedure,
+ bop_proc_apply, 0, 0, 0, (SCM bop_proc))
+{
+ SCM port, bv;
+ scm_t_bop_buffer *buf, result_buf;
+
+ port = SCM_PACK (SCM_SMOB_DATA (bop_proc));
+ buf = SCM_BOP_BUFFER (port);
+
+ result_buf = *buf;
+ bop_buffer_init (buf);
+
+ if (result_buf.len == 0)
+ bv = scm_c_take_bytevector (NULL, 0);
+ else
+ {
+ if (result_buf.total_len > result_buf.len)
+ /* Shrink the buffer. */
+ result_buf.buffer = scm_gc_realloc ((void *) result_buf.buffer,
+ result_buf.total_len,
+ result_buf.len,
+ SCM_GC_BOP);
+
+ bv = scm_c_take_bytevector ((signed char *) result_buf.buffer,
+ result_buf.len);
+ }
+
+ return bv;
+}
+
+SCM_SMOB_MARK (bytevector_output_port_procedure, bop_proc_mark,
+ bop_proc)
+{
+ /* Mark the port associated with BOP_PROC. */
+ return (SCM_PACK (SCM_SMOB_DATA (bop_proc)));
+}
+
+
+SCM_DEFINE (scm_open_bytevector_output_port,
+ "open-bytevector-output-port", 0, 1, 0,
+ (SCM transcoder),
+ "Return two values: an output port and a procedure. The latter "
+ "should be called with zero arguments to obtain a bytevector "
+ "containing the data accumulated by the port.")
+#define FUNC_NAME s_scm_open_bytevector_output_port
+{
+ if (!SCM_UNBNDP (transcoder) && !scm_is_false (transcoder))
+ transcoders_not_implemented ();
+
+ return (make_bop ());
+}
+#undef FUNC_NAME
+
+static inline void
+initialize_bytevector_output_ports (void)
+{
+ bytevector_output_port_type =
+ scm_make_port_type ("r6rs-bytevector-output-port",
+ NULL, bop_write);
+
+ scm_set_port_seek (bytevector_output_port_type, bop_seek);
+ scm_set_port_free (bytevector_output_port_type, bop_free);
+}
+
+
+/* Custom binary output port ("cbop" for short). */
+
+static scm_t_bits custom_binary_output_port_type;
+
+/* Return the various procedures of PORT. */
+#define SCM_CBOP_WRITE_PROC(_port) \
+ SCM_SIMPLE_VECTOR_REF (SCM_PACK (SCM_STREAM (_port)), 0)
+
+
+static inline SCM
+make_cbop (SCM write_proc, SCM get_position_proc,
+ SCM set_position_proc, SCM close_proc)
+{
+ SCM port, method_vector;
+ scm_t_port *c_port;
+ const unsigned long mode_bits = SCM_OPN | SCM_WRTNG;
+
+ /* Store the various methods and bytevector in a vector. */
+ method_vector = scm_c_make_vector (4, SCM_BOOL_F);
+ SCM_SIMPLE_VECTOR_SET (method_vector, 0, write_proc);
+ SCM_SIMPLE_VECTOR_SET (method_vector, 1, get_position_proc);
+ SCM_SIMPLE_VECTOR_SET (method_vector, 2, set_position_proc);
+ SCM_SIMPLE_VECTOR_SET (method_vector, 3, close_proc);
+
+ port = scm_new_port_table_entry (custom_binary_output_port_type);
+
+ /* Attach it the method vector. */
+ SCM_SETSTREAM (port, SCM_UNPACK (method_vector));
+
+ /* Have the port directly access the buffer (bytevector). */
+ c_port = SCM_PTAB_ENTRY (port);
+ c_port->write_buf = c_port->write_pos = c_port->write_end = NULL;
+ c_port->write_buf_size = c_port->read_buf_size = 0;
+
+ /* Mark PORT as open, writable and unbuffered. */
+ SCM_SET_CELL_TYPE (port, custom_binary_output_port_type | mode_bits);
+
+ return port;
+}
+
+/* Write SIZE octets from DATA to PORT. */
+static void
+cbop_write (SCM port, const void *data, size_t size)
+#define FUNC_NAME "cbop_write"
+{
+ long int c_result;
+ size_t c_written;
+ SCM bv, write_proc, result;
+
+ /* XXX: Allocating a new bytevector at each `write' call is inefficient,
+ but necessary since (1) we don't control the lifetime of the buffer
+ pointed to by DATA, and (2) the `write!' procedure could capture the
+ bytevector it is passed. */
+ bv = scm_c_make_bytevector (size);
+ memcpy (SCM_BYTEVECTOR_CONTENTS (bv), data, size);
+
+ write_proc = SCM_CBOP_WRITE_PROC (port);
+
+ /* Since the `write' procedure of Guile's ports has type `void', it must
+ try hard to write exactly SIZE bytes, regardless of how many bytes the
+ sink can handle. */
+ for (c_written = 0;
+ c_written < size;
+ c_written += c_result)
+ {
+ result = scm_call_3 (write_proc, bv,
+ scm_from_size_t (c_written),
+ scm_from_size_t (size - c_written));
+
+ c_result = scm_to_long (result);
+ if (SCM_UNLIKELY (c_result < 0
+ || (size_t) c_result > (size - c_written)))
+ scm_wrong_type_arg_msg (FUNC_NAME, 0, result,
+ "R6RS custom binary output port `write!' "
+ "returned a incorrect integer");
+ }
+}
+#undef FUNC_NAME
+
+
+SCM_DEFINE (scm_make_custom_binary_output_port,
+ "make-custom-binary-output-port", 5, 0, 0,
+ (SCM id, SCM write_proc, SCM get_position_proc,
+ SCM set_position_proc, SCM close_proc),
+ "Return a new custom binary output port whose output is drained "
+ "by invoking @var{write_proc} and passing it a bytevector, an "
+ "index where octets should be written, and an octet count.")
+#define FUNC_NAME s_scm_make_custom_binary_output_port
+{
+ SCM_VALIDATE_STRING (1, id);
+ SCM_VALIDATE_PROC (2, write_proc);
+
+ if (!scm_is_false (get_position_proc))
+ SCM_VALIDATE_PROC (3, get_position_proc);
+
+ if (!scm_is_false (set_position_proc))
+ SCM_VALIDATE_PROC (4, set_position_proc);
+
+ if (!scm_is_false (close_proc))
+ SCM_VALIDATE_PROC (5, close_proc);
+
+ return (make_cbop (write_proc, get_position_proc, set_position_proc,
+ close_proc));
+}
+#undef FUNC_NAME
+
+
+/* Instantiate the custom binary output port type. */
+static inline void
+initialize_custom_binary_output_ports (void)
+{
+ custom_binary_output_port_type =
+ scm_make_port_type ("r6rs-custom-binary-output-port",
+ NULL, cbop_write);
+
+ scm_set_port_mark (custom_binary_output_port_type, cbp_mark);
+ scm_set_port_seek (custom_binary_output_port_type, cbp_seek);
+ scm_set_port_close (custom_binary_output_port_type, cbp_close);
+}
+
+
+/* Initialization. */
+
+void
+scm_init_r6rs_ports (void)
+{
+#include "r6rs-ports.x"
+
+ initialize_bytevector_input_ports ();
+ initialize_custom_binary_input_ports ();
+ initialize_bytevector_output_ports ();
+ initialize_custom_binary_output_ports ();
+}
diff --git a/libguile/r6rs-ports.h b/libguile/r6rs-ports.h
new file mode 100644
index 0000000..e29d962
--- /dev/null
+++ b/libguile/r6rs-ports.h
@@ -0,0 +1,43 @@
+#ifndef SCM_R6RS_PORTS_H
+#define SCM_R6RS_PORTS_H
+
+/* Copyright (C) 2009 Free Software Foundation, Inc.
+ *
+ * This library is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU Lesser General Public
+ * License as published by the Free Software Foundation; either
+ * version 2.1 of the License, or (at your option) any later version.
+ *
+ * This library is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+ * Lesser General Public License for more details.
+ *
+ * You should have received a copy of the GNU Lesser General Public
+ * License along with this library; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
+ */
+
+
+
+#include "libguile/__scm.h"
+
+/* R6RS I/O Ports. */
+
+SCM_API SCM scm_eof_object (void);
+SCM_API SCM scm_open_bytevector_input_port (SCM, SCM);
+SCM_API SCM scm_make_custom_binary_input_port (SCM, SCM, SCM, SCM, SCM);
+SCM_API SCM scm_get_u8 (SCM);
+SCM_API SCM scm_lookahead_u8 (SCM);
+SCM_API SCM scm_get_bytevector_n (SCM, SCM);
+SCM_API SCM scm_get_bytevector_n_x (SCM, SCM, SCM, SCM);
+SCM_API SCM scm_get_bytevector_some (SCM);
+SCM_API SCM scm_get_bytevector_all (SCM);
+SCM_API SCM scm_put_u8 (SCM, SCM);
+SCM_API SCM scm_put_bytevector (SCM, SCM, SCM, SCM);
+SCM_API SCM scm_open_bytevector_output_port (SCM);
+SCM_API SCM scm_make_custom_binary_output_port (SCM, SCM, SCM, SCM, SCM);
+
+SCM_API void scm_init_r6rs_ports (void);
+
+#endif /* SCM_R6RS_PORTS_H */
diff --git a/libguile/validate.h b/libguile/validate.h
index e05b7dd..c362c02 100644
--- a/libguile/validate.h
+++ b/libguile/validate.h
@@ -3,7 +3,7 @@
#ifndef SCM_VALIDATE_H
#define SCM_VALIDATE_H
-/* Copyright (C) 1999,2000,2001, 2002, 2004, 2006, 2007 Free Software
Foundation, Inc.
+/* Copyright (C) 1999,2000,2001, 2002, 2004, 2006, 2007, 2009 Free Software
Foundation, Inc.
*
* This library is free software; you can redistribute it and/or
* modify it under the terms of the GNU Lesser General Public
@@ -150,6 +150,9 @@
cvar = scm_to_bool (flag); \
} while (0)
+#define SCM_VALIDATE_BYTEVECTOR(_pos, _obj) \
+ SCM_VALIDATE_SMOB ((_pos), (_obj), bytevector)
+
#define SCM_VALIDATE_CHAR(pos, scm) SCM_MAKE_VALIDATE_MSG (pos, scm, CHARP,
"character")
#define SCM_VALIDATE_CHAR_COPY(pos, scm, cvar) \
diff --git a/m4/byteswap.m4 b/m4/byteswap.m4
new file mode 100644
index 0000000..ad13f22
--- /dev/null
+++ b/m4/byteswap.m4
@@ -0,0 +1,18 @@
+# byteswap.m4 serial 3
+dnl Copyright (C) 2005, 2007, 2009 Free Software Foundation, Inc.
+dnl This file is free software; the Free Software Foundation
+dnl gives unlimited permission to copy and/or distribute it,
+dnl with or without modifications, as long as this notice is preserved.
+
+dnl Written by Oskar Liljeblad.
+
+AC_DEFUN([gl_BYTESWAP],
+[
+ dnl Prerequisites of lib/byteswap.in.h.
+ AC_CHECK_HEADERS([byteswap.h], [
+ BYTESWAP_H=''
+ ], [
+ BYTESWAP_H='byteswap.h'
+ ])
+ AC_SUBST([BYTESWAP_H])
+])
diff --git a/m4/gnulib-cache.m4 b/m4/gnulib-cache.m4
index c7cfb83..0fbe119 100644
--- a/m4/gnulib-cache.m4
+++ b/m4/gnulib-cache.m4
@@ -15,13 +15,14 @@
# Specification in the form of a command-line invocation:
-# gnulib-tool --import --dir=. --lib=libgnu --source-base=lib --m4-base=m4
--doc-base=doc --tests-base=tests --aux-dir=build-aux --lgpl --libtool
--macro-prefix=gl --no-vc-files alloca-opt autobuild count-one-bits environ
extensions flock fpieee full-read full-write lib-symbol-visibility putenv
stdlib strcase strftime
+# gnulib-tool --import --dir=. --lib=libgnu --source-base=lib --m4-base=m4
--doc-base=doc --tests-base=tests --aux-dir=build-aux --lgpl --libtool
--macro-prefix=gl --no-vc-files alloca-opt autobuild byteswap count-one-bits
environ extensions flock fpieee full-read full-write iconv_open-utf
lib-symbol-visibility libunistring putenv stdlib strcase strftime striconveh
string
# Specification in the form of a few gnulib-tool.m4 macro invocations:
gl_LOCAL_DIR([])
gl_MODULES([
alloca-opt
autobuild
+ byteswap
count-one-bits
environ
extensions
@@ -29,11 +30,15 @@ gl_MODULES([
fpieee
full-read
full-write
+ iconv_open-utf
lib-symbol-visibility
+ libunistring
putenv
stdlib
strcase
strftime
+ striconveh
+ string
])
gl_AVOID([])
gl_SOURCE_BASE([lib])
diff --git a/m4/gnulib-comp.m4 b/m4/gnulib-comp.m4
index b6d10a8..8f77510 100644
--- a/m4/gnulib-comp.m4
+++ b/m4/gnulib-comp.m4
@@ -25,6 +25,7 @@ AC_DEFUN([gl_EARLY],
m4_pattern_allow([^gl_LIBOBJS$])dnl a variable
m4_pattern_allow([^gl_LTLIBOBJS$])dnl a variable
AC_REQUIRE([AC_PROG_RANLIB])
+ AC_REQUIRE([AM_PROG_CC_C_O])
AB_INIT
AC_REQUIRE([gl_USE_SYSTEM_EXTENSIONS])
AC_REQUIRE([gl_FP_IEEE])
@@ -44,13 +45,19 @@ AC_DEFUN([gl_INIT],
gl_COMMON
gl_source_base='lib'
gl_FUNC_ALLOCA
+ gl_BYTESWAP
gl_COUNT_ONE_BITS
gl_ENVIRON
gl_UNISTD_MODULE_INDICATOR([environ])
gl_FUNC_FLOCK
gl_HEADER_SYS_FILE_MODULE_INDICATOR([flock])
+ AM_ICONV
+ gl_ICONV_H
+ gl_FUNC_ICONV_OPEN
+ gl_FUNC_ICONV_OPEN_UTF
gl_INLINE
gl_VISIBILITY
+ gl_LIBUNISTRING
gl_LOCALCHARSET
LOCALCHARSET_TESTS_ENVIRONMENT="CHARSETALIASDIR=\"\$(top_builddir)/$gl_source_base\""
AC_SUBST([LOCALCHARSET_TESTS_ENVIRONMENT])
@@ -73,12 +80,21 @@ AC_DEFUN([gl_INIT],
gl_STDLIB_H
gl_STRCASE
gl_FUNC_GNU_STRFTIME
+ if test $gl_cond_libtool = false; then
+ gl_ltlibdeps="$gl_ltlibdeps $LTLIBICONV"
+ gl_libdeps="$gl_libdeps $LIBICONV"
+ fi
+ gl_HEADER_STRING_H
gl_HEADER_STRINGS_H
gl_HEADER_SYS_FILE_H
AC_PROG_MKDIR_P
gl_HEADER_TIME_H
gl_TIME_R
gl_UNISTD_H
+ gl_MODULE_INDICATOR([unistr/u8-mbtouc])
+ gl_MODULE_INDICATOR([unistr/u8-mbtouc-unsafe])
+ gl_MODULE_INDICATOR([unistr/u8-mbtoucr])
+ gl_MODULE_INDICATOR([unistr/u8-uctomb])
gl_WCHAR_H
gl_FUNC_WRITE
gl_UNISTD_MODULE_INDICATOR([write])
@@ -210,8 +226,16 @@ AC_DEFUN([gltests_LIBSOURCES], [
# This macro records the list of files which have been installed by
# gnulib-tool and may be removed by future gnulib-tool invocations.
AC_DEFUN([gl_FILE_LIST], [
+ build-aux/config.rpath
build-aux/link-warning.h
lib/alloca.in.h
+ lib/byteswap.in.h
+ lib/c-ctype.c
+ lib/c-ctype.h
+ lib/c-strcase.h
+ lib/c-strcasecmp.c
+ lib/c-strcaseeq.h
+ lib/c-strncasecmp.c
lib/config.charset
lib/count-one-bits.h
lib/flock.c
@@ -219,6 +243,15 @@ AC_DEFUN([gl_FILE_LIST], [
lib/full-read.h
lib/full-write.c
lib/full-write.h
+ lib/iconv.c
+ lib/iconv.in.h
+ lib/iconv_close.c
+ lib/iconv_open-aix.gperf
+ lib/iconv_open-hpux.gperf
+ lib/iconv_open-irix.gperf
+ lib/iconv_open-osf.gperf
+ lib/iconv_open.c
+ lib/iconveh.h
lib/localcharset.c
lib/localcharset.h
lib/malloc.c
@@ -239,18 +272,32 @@ AC_DEFUN([gl_FILE_LIST], [
lib/streq.h
lib/strftime.c
lib/strftime.h
+ lib/striconveh.c
+ lib/striconveh.h
+ lib/string.in.h
lib/strings.in.h
lib/strncasecmp.c
lib/sys_file.in.h
lib/time.in.h
lib/time_r.c
lib/unistd.in.h
+ lib/unistr.h
+ lib/unistr/u8-mbtouc-aux.c
+ lib/unistr/u8-mbtouc-unsafe-aux.c
+ lib/unistr/u8-mbtouc-unsafe.c
+ lib/unistr/u8-mbtouc.c
+ lib/unistr/u8-mbtoucr.c
+ lib/unistr/u8-prev.c
+ lib/unistr/u8-uctomb-aux.c
+ lib/unistr/u8-uctomb.c
+ lib/unitypes.h
lib/verify.h
lib/wchar.in.h
lib/write.c
m4/00gnulib.m4
m4/alloca.m4
m4/autobuild.m4
+ m4/byteswap.m4
m4/codeset.m4
m4/count-one-bits.m4
m4/environ.m4
@@ -259,8 +306,15 @@ AC_DEFUN([gl_FILE_LIST], [
m4/fpieee.m4
m4/glibc21.m4
m4/gnulib-common.m4
+ m4/iconv.m4
+ m4/iconv_h.m4
+ m4/iconv_open.m4
m4/include_next.m4
m4/inline.m4
+ m4/lib-ld.m4
+ m4/lib-link.m4
+ m4/lib-prefix.m4
+ m4/libunistring.m4
m4/localcharset.m4
m4/locale-fr.m4
m4/locale-ja.m4
@@ -281,6 +335,7 @@ AC_DEFUN([gl_FILE_LIST], [
m4/stdlib_h.m4
m4/strcase.m4
m4/strftime.m4
+ m4/string_h.m4
m4/strings_h.m4
m4/sys_file_h.m4
m4/time_h.m4
diff --git a/m4/iconv.m4 b/m4/iconv.m4
new file mode 100644
index 0000000..3cc6268
--- /dev/null
+++ b/m4/iconv.m4
@@ -0,0 +1,180 @@
+# iconv.m4 serial AM7 (gettext-0.18)
+dnl Copyright (C) 2000-2002, 2007-2009 Free Software Foundation, Inc.
+dnl This file is free software; the Free Software Foundation
+dnl gives unlimited permission to copy and/or distribute it,
+dnl with or without modifications, as long as this notice is preserved.
+
+dnl From Bruno Haible.
+
+AC_DEFUN([AM_ICONV_LINKFLAGS_BODY],
+[
+ dnl Prerequisites of AC_LIB_LINKFLAGS_BODY.
+ AC_REQUIRE([AC_LIB_PREPARE_PREFIX])
+ AC_REQUIRE([AC_LIB_RPATH])
+
+ dnl Search for libiconv and define LIBICONV, LTLIBICONV and INCICONV
+ dnl accordingly.
+ AC_LIB_LINKFLAGS_BODY([iconv])
+])
+
+AC_DEFUN([AM_ICONV_LINK],
+[
+ dnl Some systems have iconv in libc, some have it in libiconv (OSF/1 and
+ dnl those with the standalone portable GNU libiconv installed).
+ AC_REQUIRE([AC_CANONICAL_HOST]) dnl for cross-compiles
+
+ dnl Search for libiconv and define LIBICONV, LTLIBICONV and INCICONV
+ dnl accordingly.
+ AC_REQUIRE([AM_ICONV_LINKFLAGS_BODY])
+
+ dnl Add $INCICONV to CPPFLAGS before performing the following checks,
+ dnl because if the user has installed libiconv and not disabled its use
+ dnl via --without-libiconv-prefix, he wants to use it. The first
+ dnl AC_TRY_LINK will then fail, the second AC_TRY_LINK will succeed.
+ am_save_CPPFLAGS="$CPPFLAGS"
+ AC_LIB_APPENDTOVAR([CPPFLAGS], [$INCICONV])
+
+ AC_CACHE_CHECK([for iconv], [am_cv_func_iconv], [
+ am_cv_func_iconv="no, consider installing GNU libiconv"
+ am_cv_lib_iconv=no
+ AC_TRY_LINK([#include <stdlib.h>
+#include <iconv.h>],
+ [iconv_t cd = iconv_open("","");
+ iconv(cd,NULL,NULL,NULL,NULL);
+ iconv_close(cd);],
+ [am_cv_func_iconv=yes])
+ if test "$am_cv_func_iconv" != yes; then
+ am_save_LIBS="$LIBS"
+ LIBS="$LIBS $LIBICONV"
+ AC_TRY_LINK([#include <stdlib.h>
+#include <iconv.h>],
+ [iconv_t cd = iconv_open("","");
+ iconv(cd,NULL,NULL,NULL,NULL);
+ iconv_close(cd);],
+ [am_cv_lib_iconv=yes]
+ [am_cv_func_iconv=yes])
+ LIBS="$am_save_LIBS"
+ fi
+ ])
+ if test "$am_cv_func_iconv" = yes; then
+ AC_CACHE_CHECK([for working iconv], [am_cv_func_iconv_works], [
+ dnl This tests against bugs in AIX 5.1 and HP-UX 11.11.
+ am_save_LIBS="$LIBS"
+ if test $am_cv_lib_iconv = yes; then
+ LIBS="$LIBS $LIBICONV"
+ fi
+ AC_TRY_RUN([
+#include <iconv.h>
+#include <string.h>
+int main ()
+{
+ /* Test against AIX 5.1 bug: Failures are not distinguishable from successful
+ returns. */
+ {
+ iconv_t cd_utf8_to_88591 = iconv_open ("ISO8859-1", "UTF-8");
+ if (cd_utf8_to_88591 != (iconv_t)(-1))
+ {
+ static const char input[] = "\342\202\254"; /* EURO SIGN */
+ char buf[10];
+ const char *inptr = input;
+ size_t inbytesleft = strlen (input);
+ char *outptr = buf;
+ size_t outbytesleft = sizeof (buf);
+ size_t res = iconv (cd_utf8_to_88591,
+ (char **) &inptr, &inbytesleft,
+ &outptr, &outbytesleft);
+ if (res == 0)
+ return 1;
+ }
+ }
+#if 0 /* This bug could be worked around by the caller. */
+ /* Test against HP-UX 11.11 bug: Positive return value instead of 0. */
+ {
+ iconv_t cd_88591_to_utf8 = iconv_open ("utf8", "iso88591");
+ if (cd_88591_to_utf8 != (iconv_t)(-1))
+ {
+ static const char input[] = "\304rger mit b\366sen B\374bchen ohne
Augenma\337";
+ char buf[50];
+ const char *inptr = input;
+ size_t inbytesleft = strlen (input);
+ char *outptr = buf;
+ size_t outbytesleft = sizeof (buf);
+ size_t res = iconv (cd_88591_to_utf8,
+ (char **) &inptr, &inbytesleft,
+ &outptr, &outbytesleft);
+ if ((int)res > 0)
+ return 1;
+ }
+ }
+#endif
+ /* Test against HP-UX 11.11 bug: No converter from EUC-JP to UTF-8 is
+ provided. */
+ if (/* Try standardized names. */
+ iconv_open ("UTF-8", "EUC-JP") == (iconv_t)(-1)
+ /* Try IRIX, OSF/1 names. */
+ && iconv_open ("UTF-8", "eucJP") == (iconv_t)(-1)
+ /* Try AIX names. */
+ && iconv_open ("UTF-8", "IBM-eucJP") == (iconv_t)(-1)
+ /* Try HP-UX names. */
+ && iconv_open ("utf8", "eucJP") == (iconv_t)(-1))
+ return 1;
+ return 0;
+}], [am_cv_func_iconv_works=yes], [am_cv_func_iconv_works=no],
+ [case "$host_os" in
+ aix* | hpux*) am_cv_func_iconv_works="guessing no" ;;
+ *) am_cv_func_iconv_works="guessing yes" ;;
+ esac])
+ LIBS="$am_save_LIBS"
+ ])
+ case "$am_cv_func_iconv_works" in
+ *no) am_func_iconv=no am_cv_lib_iconv=no ;;
+ *) am_func_iconv=yes ;;
+ esac
+ else
+ am_func_iconv=no am_cv_lib_iconv=no
+ fi
+ if test "$am_func_iconv" = yes; then
+ AC_DEFINE([HAVE_ICONV], [1],
+ [Define if you have the iconv() function and it works.])
+ fi
+ if test "$am_cv_lib_iconv" = yes; then
+ AC_MSG_CHECKING([how to link with libiconv])
+ AC_MSG_RESULT([$LIBICONV])
+ else
+ dnl If $LIBICONV didn't lead to a usable library, we don't need $INCICONV
+ dnl either.
+ CPPFLAGS="$am_save_CPPFLAGS"
+ LIBICONV=
+ LTLIBICONV=
+ fi
+ AC_SUBST([LIBICONV])
+ AC_SUBST([LTLIBICONV])
+])
+
+AC_DEFUN([AM_ICONV],
+[
+ AM_ICONV_LINK
+ if test "$am_cv_func_iconv" = yes; then
+ AC_MSG_CHECKING([for iconv declaration])
+ AC_CACHE_VAL([am_cv_proto_iconv], [
+ AC_TRY_COMPILE([
+#include <stdlib.h>
+#include <iconv.h>
+extern
+#ifdef __cplusplus
+"C"
+#endif
+#if defined(__STDC__) || defined(__cplusplus)
+size_t iconv (iconv_t cd, char * *inbuf, size_t *inbytesleft, char * *outbuf,
size_t *outbytesleft);
+#else
+size_t iconv();
+#endif
+], [], [am_cv_proto_iconv_arg1=""], [am_cv_proto_iconv_arg1="const"])
+ am_cv_proto_iconv="extern size_t iconv (iconv_t cd,
$am_cv_proto_iconv_arg1 char * *inbuf, size_t *inbytesleft, char * *outbuf,
size_t *outbytesleft);"])
+ am_cv_proto_iconv=`echo "[$]am_cv_proto_iconv" | tr -s ' ' | sed -e 's/(
/(/'`
+ AC_MSG_RESULT([${ac_t:-
+ }$am_cv_proto_iconv])
+ AC_DEFINE_UNQUOTED([ICONV_CONST], [$am_cv_proto_iconv_arg1],
+ [Define as const if the declaration of iconv() needs const.])
+ fi
+])
diff --git a/m4/iconv_h.m4 b/m4/iconv_h.m4
new file mode 100644
index 0000000..bc05b05
--- /dev/null
+++ b/m4/iconv_h.m4
@@ -0,0 +1,34 @@
+# iconv_h.m4 serial 4
+dnl Copyright (C) 2007-2008 Free Software Foundation, Inc.
+dnl This file is free software; the Free Software Foundation
+dnl gives unlimited permission to copy and/or distribute it,
+dnl with or without modifications, as long as this notice is preserved.
+
+AC_DEFUN([gl_ICONV_H],
+[
+ AC_REQUIRE([gl_ICONV_H_DEFAULTS])
+ gl_CHECK_NEXT_HEADERS([iconv.h])
+])
+
+dnl Unconditionally enables the replacement of <iconv.h>.
+AC_DEFUN([gl_REPLACE_ICONV_H],
+[
+ AC_REQUIRE([gl_ICONV_H_DEFAULTS])
+ ICONV_H='iconv.h'
+])
+
+AC_DEFUN([gl_ICONV_MODULE_INDICATOR],
+[
+ dnl Use AC_REQUIRE here, so that the default settings are expanded once only.
+ AC_REQUIRE([gl_ICONV_H_DEFAULTS])
+
GNULIB_[]m4_translit([$1],[abcdefghijklmnopqrstuvwxyz./-],[ABCDEFGHIJKLMNOPQRSTUVWXYZ___])=1
+])
+
+AC_DEFUN([gl_ICONV_H_DEFAULTS],
+[
+ dnl Assume proper GNU behavior unless another module says otherwise.
+ REPLACE_ICONV=0; AC_SUBST([REPLACE_ICONV])
+ REPLACE_ICONV_OPEN=0; AC_SUBST([REPLACE_ICONV_OPEN])
+ REPLACE_ICONV_UTF=0; AC_SUBST([REPLACE_ICONV_UTF])
+ ICONV_H=''; AC_SUBST([ICONV_H])
+])
diff --git a/m4/iconv_open.m4 b/m4/iconv_open.m4
new file mode 100644
index 0000000..c7b948e
--- /dev/null
+++ b/m4/iconv_open.m4
@@ -0,0 +1,237 @@
+# iconv_open.m4 serial 5
+dnl Copyright (C) 2007-2009 Free Software Foundation, Inc.
+dnl This file is free software; the Free Software Foundation
+dnl gives unlimited permission to copy and/or distribute it,
+dnl with or without modifications, as long as this notice is preserved.
+
+AC_DEFUN([gl_FUNC_ICONV_OPEN],
+[
+ AC_REQUIRE([AM_ICONV])
+ AC_REQUIRE([AC_CANONICAL_HOST])
+ AC_REQUIRE([gl_ICONV_H_DEFAULTS])
+ if test "$am_cv_func_iconv" = yes; then
+ dnl Test whether iconv_open accepts standardized encoding names.
+ dnl We know that GNU libiconv and GNU libc do.
+ AC_EGREP_CPP([gnu_iconv], [
+ #include <iconv.h>
+ #if defined _LIBICONV_VERSION || defined __GLIBC__
+ gnu_iconv
+ #endif
+ ], [gl_func_iconv_gnu=yes], [gl_func_iconv_gnu=no])
+ if test $gl_func_iconv_gnu = no; then
+ iconv_flavor=
+ case "$host_os" in
+ aix*) iconv_flavor=ICONV_FLAVOR_AIX ;;
+ irix*) iconv_flavor=ICONV_FLAVOR_IRIX ;;
+ hpux*) iconv_flavor=ICONV_FLAVOR_HPUX ;;
+ osf*) iconv_flavor=ICONV_FLAVOR_OSF ;;
+ esac
+ if test -n "$iconv_flavor"; then
+ AC_DEFINE_UNQUOTED([ICONV_FLAVOR], [$iconv_flavor],
+ [Define to a symbolic name denoting the flavor of iconv_open()
+ implementation.])
+ gl_REPLACE_ICONV_OPEN
+ fi
+ fi
+ fi
+])
+
+AC_DEFUN([gl_REPLACE_ICONV_OPEN],
+[
+ gl_REPLACE_ICONV_H
+ REPLACE_ICONV_OPEN=1
+ AC_LIBOBJ([iconv_open])
+])
+
+AC_DEFUN([gl_FUNC_ICONV_OPEN_UTF],
+[
+ AC_REQUIRE([gl_FUNC_ICONV_OPEN])
+ AC_REQUIRE([AC_CANONICAL_HOST]) dnl for cross-compiles
+ AC_REQUIRE([gl_ICONV_H_DEFAULTS])
+ if test "$am_cv_func_iconv" = yes; then
+ if test -n "$am_cv_proto_iconv_arg1"; then
+ ICONV_CONST="const"
+ else
+ ICONV_CONST=
+ fi
+ AC_SUBST([ICONV_CONST])
+ AC_CACHE_CHECK([whether iconv supports conversion between UTF-8 and
UTF-{16,32}{BE,LE}],
+ [gl_cv_func_iconv_supports_utf],
+ [
+ save_LIBS="$LIBS"
+ LIBS="$LIBS $LIBICONV"
+ AC_TRY_RUN([
+#include <iconv.h>
+#include <errno.h>
+#include <stdio.h>
+#include <stdlib.h>
+#include <string.h>
+#define ASSERT(expr) if (!(expr)) return 1;
+int main ()
+{
+ /* Test conversion from UTF-8 to UTF-16BE with no errors. */
+ {
+ static const char input[] =
+ "Japanese (\346\227\245\346\234\254\350\252\236)
[\360\235\224\215\360\235\224\236\360\235\224\255]";
+ static const char expected[] =
+ "\000J\000a\000p\000a\000n\000e\000s\000e\000
\000(\145\345\147\054\212\236\000)\000
\000[\330\065\335\015\330\065\335\036\330\065\335\055\000]";
+ iconv_t cd;
+ char buf[100];
+ const char *inptr;
+ size_t inbytesleft;
+ char *outptr;
+ size_t outbytesleft;
+ size_t res;
+ cd = iconv_open ("UTF-16BE", "UTF-8");
+ ASSERT (cd != (iconv_t)(-1));
+ inptr = input;
+ inbytesleft = sizeof (input) - 1;
+ outptr = buf;
+ outbytesleft = sizeof (buf);
+ res = iconv (cd,
+ (ICONV_CONST char **) &inptr, &inbytesleft,
+ &outptr, &outbytesleft);
+ ASSERT (res == 0 && inbytesleft == 0);
+ ASSERT (outptr == buf + (sizeof (expected) - 1));
+ ASSERT (memcmp (buf, expected, sizeof (expected) - 1) == 0);
+ ASSERT (iconv_close (cd) == 0);
+ }
+ /* Test conversion from UTF-8 to UTF-16LE with no errors. */
+ {
+ static const char input[] =
+ "Japanese (\346\227\245\346\234\254\350\252\236)
[\360\235\224\215\360\235\224\236\360\235\224\255]";
+ static const char expected[] =
+ "J\000a\000p\000a\000n\000e\000s\000e\000
\000(\000\345\145\054\147\236\212)\000
\000[\000\065\330\015\335\065\330\036\335\065\330\055\335]\000";
+ iconv_t cd;
+ char buf[100];
+ const char *inptr;
+ size_t inbytesleft;
+ char *outptr;
+ size_t outbytesleft;
+ size_t res;
+ cd = iconv_open ("UTF-16LE", "UTF-8");
+ ASSERT (cd != (iconv_t)(-1));
+ inptr = input;
+ inbytesleft = sizeof (input) - 1;
+ outptr = buf;
+ outbytesleft = sizeof (buf);
+ res = iconv (cd,
+ (ICONV_CONST char **) &inptr, &inbytesleft,
+ &outptr, &outbytesleft);
+ ASSERT (res == 0 && inbytesleft == 0);
+ ASSERT (outptr == buf + (sizeof (expected) - 1));
+ ASSERT (memcmp (buf, expected, sizeof (expected) - 1) == 0);
+ ASSERT (iconv_close (cd) == 0);
+ }
+ /* Test conversion from UTF-8 to UTF-32BE with no errors. */
+ {
+ static const char input[] =
+ "Japanese (\346\227\245\346\234\254\350\252\236)
[\360\235\224\215\360\235\224\236\360\235\224\255]";
+ static const char expected[] =
+
"\000\000\000J\000\000\000a\000\000\000p\000\000\000a\000\000\000n\000\000\000e\000\000\000s\000\000\000e\000\000\000
\000\000\000(\000\000\145\345\000\000\147\054\000\000\212\236\000\000\000)\000\000\000
\000\000\000[\000\001\325\015\000\001\325\036\000\001\325\055\000\000\000]";
+ iconv_t cd;
+ char buf[100];
+ const char *inptr;
+ size_t inbytesleft;
+ char *outptr;
+ size_t outbytesleft;
+ size_t res;
+ cd = iconv_open ("UTF-32BE", "UTF-8");
+ ASSERT (cd != (iconv_t)(-1));
+ inptr = input;
+ inbytesleft = sizeof (input) - 1;
+ outptr = buf;
+ outbytesleft = sizeof (buf);
+ res = iconv (cd,
+ (ICONV_CONST char **) &inptr, &inbytesleft,
+ &outptr, &outbytesleft);
+ ASSERT (res == 0 && inbytesleft == 0);
+ ASSERT (outptr == buf + (sizeof (expected) - 1));
+ ASSERT (memcmp (buf, expected, sizeof (expected) - 1) == 0);
+ ASSERT (iconv_close (cd) == 0);
+ }
+ /* Test conversion from UTF-8 to UTF-32LE with no errors. */
+ {
+ static const char input[] =
+ "Japanese (\346\227\245\346\234\254\350\252\236)
[\360\235\224\215\360\235\224\236\360\235\224\255]";
+ static const char expected[] =
+
"J\000\000\000a\000\000\000p\000\000\000a\000\000\000n\000\000\000e\000\000\000s\000\000\000e\000\000\000
\000\000\000(\000\000\000\345\145\000\000\054\147\000\000\236\212\000\000)\000\000\000
\000\000\000[\000\000\000\015\325\001\000\036\325\001\000\055\325\001\000]\000\000\000";
+ iconv_t cd;
+ char buf[100];
+ const char *inptr;
+ size_t inbytesleft;
+ char *outptr;
+ size_t outbytesleft;
+ size_t res;
+ cd = iconv_open ("UTF-32LE", "UTF-8");
+ ASSERT (cd != (iconv_t)(-1));
+ inptr = input;
+ inbytesleft = sizeof (input) - 1;
+ outptr = buf;
+ outbytesleft = sizeof (buf);
+ res = iconv (cd,
+ (ICONV_CONST char **) &inptr, &inbytesleft,
+ &outptr, &outbytesleft);
+ ASSERT (res == 0 && inbytesleft == 0);
+ ASSERT (outptr == buf + (sizeof (expected) - 1));
+ ASSERT (memcmp (buf, expected, sizeof (expected) - 1) == 0);
+ ASSERT (iconv_close (cd) == 0);
+ }
+ /* Test conversion from UTF-16BE to UTF-8 with no errors.
+ This test fails on NetBSD 3.0. */
+ {
+ static const char input[] =
+ "\000J\000a\000p\000a\000n\000e\000s\000e\000
\000(\145\345\147\054\212\236\000)\000
\000[\330\065\335\015\330\065\335\036\330\065\335\055\000]";
+ static const char expected[] =
+ "Japanese (\346\227\245\346\234\254\350\252\236)
[\360\235\224\215\360\235\224\236\360\235\224\255]";
+ iconv_t cd;
+ char buf[100];
+ const char *inptr;
+ size_t inbytesleft;
+ char *outptr;
+ size_t outbytesleft;
+ size_t res;
+ cd = iconv_open ("UTF-8", "UTF-16BE");
+ ASSERT (cd != (iconv_t)(-1));
+ inptr = input;
+ inbytesleft = sizeof (input) - 1;
+ outptr = buf;
+ outbytesleft = sizeof (buf);
+ res = iconv (cd,
+ (ICONV_CONST char **) &inptr, &inbytesleft,
+ &outptr, &outbytesleft);
+ ASSERT (res == 0 && inbytesleft == 0);
+ ASSERT (outptr == buf + (sizeof (expected) - 1));
+ ASSERT (memcmp (buf, expected, sizeof (expected) - 1) == 0);
+ ASSERT (iconv_close (cd) == 0);
+ }
+ return 0;
+}], [gl_cv_func_iconv_supports_utf=yes], [gl_cv_func_iconv_supports_utf=no],
+ [
+ dnl We know that GNU libiconv, GNU libc, and Solaris >= 9 do.
+ dnl OSF/1 5.1 has these encodings, but inserts a BOM in the "to"
+ dnl direction.
+ gl_cv_func_iconv_supports_utf=no
+ if test $gl_func_iconv_gnu = yes; then
+ gl_cv_func_iconv_supports_utf=yes
+ else
+changequote(,)dnl
+ case "$host_os" in
+ solaris2.9 | solaris2.1[0-9]) gl_cv_func_iconv_supports_utf=yes
;;
+ esac
+changequote([,])dnl
+ fi
+ ])
+ LIBS="$save_LIBS"
+ ])
+ if test $gl_cv_func_iconv_supports_utf = no; then
+ REPLACE_ICONV_UTF=1
+ AC_DEFINE([REPLACE_ICONV_UTF], [1],
+ [Define if the iconv() functions are enhanced to handle the
UTF-{16,32}{BE,LE} encodings.])
+ REPLACE_ICONV=1
+ gl_REPLACE_ICONV_OPEN
+ AC_LIBOBJ([iconv])
+ AC_LIBOBJ([iconv_close])
+ fi
+ fi
+])
diff --git a/m4/lib-ld.m4 b/m4/lib-ld.m4
new file mode 100644
index 0000000..e4863f2
--- /dev/null
+++ b/m4/lib-ld.m4
@@ -0,0 +1,110 @@
+# lib-ld.m4 serial 4 (gettext-0.18)
+dnl Copyright (C) 1996-2003, 2009 Free Software Foundation, Inc.
+dnl This file is free software; the Free Software Foundation
+dnl gives unlimited permission to copy and/or distribute it,
+dnl with or without modifications, as long as this notice is preserved.
+
+dnl Subroutines of libtool.m4,
+dnl with replacements s/AC_/AC_LIB/ and s/lt_cv/acl_cv/ to avoid collision
+dnl with libtool.m4.
+
+dnl From libtool-1.4. Sets the variable with_gnu_ld to yes or no.
+AC_DEFUN([AC_LIB_PROG_LD_GNU],
+[AC_CACHE_CHECK([if the linker ($LD) is GNU ld], [acl_cv_prog_gnu_ld],
+[# I'd rather use --version here, but apparently some GNU ld's only accept -v.
+case `$LD -v 2>&1 </dev/null` in
+*GNU* | *'with BFD'*)
+ acl_cv_prog_gnu_ld=yes ;;
+*)
+ acl_cv_prog_gnu_ld=no ;;
+esac])
+with_gnu_ld=$acl_cv_prog_gnu_ld
+])
+
+dnl From libtool-1.4. Sets the variable LD.
+AC_DEFUN([AC_LIB_PROG_LD],
+[AC_ARG_WITH([gnu-ld],
+[ --with-gnu-ld assume the C compiler uses GNU ld [default=no]],
+test "$withval" = no || with_gnu_ld=yes, with_gnu_ld=no)
+AC_REQUIRE([AC_PROG_CC])dnl
+AC_REQUIRE([AC_CANONICAL_HOST])dnl
+# Prepare PATH_SEPARATOR.
+# The user is always right.
+if test "${PATH_SEPARATOR+set}" != set; then
+ echo "#! /bin/sh" >conf$$.sh
+ echo "exit 0" >>conf$$.sh
+ chmod +x conf$$.sh
+ if (PATH="/nonexistent;."; conf$$.sh) >/dev/null 2>&1; then
+ PATH_SEPARATOR=';'
+ else
+ PATH_SEPARATOR=:
+ fi
+ rm -f conf$$.sh
+fi
+ac_prog=ld
+if test "$GCC" = yes; then
+ # Check if gcc -print-prog-name=ld gives a path.
+ AC_MSG_CHECKING([for ld used by GCC])
+ case $host in
+ *-*-mingw*)
+ # gcc leaves a trailing carriage return which upsets mingw
+ ac_prog=`($CC -print-prog-name=ld) 2>&5 | tr -d '\015'` ;;
+ *)
+ ac_prog=`($CC -print-prog-name=ld) 2>&5` ;;
+ esac
+ case $ac_prog in
+ # Accept absolute paths.
+ [[\\/]* | [A-Za-z]:[\\/]*)]
+ [re_direlt='/[^/][^/]*/\.\./']
+ # Canonicalize the path of ld
+ ac_prog=`echo $ac_prog| sed 's%\\\\%/%g'`
+ while echo $ac_prog | grep "$re_direlt" > /dev/null 2>&1; do
+ ac_prog=`echo $ac_prog| sed "s%$re_direlt%/%"`
+ done
+ test -z "$LD" && LD="$ac_prog"
+ ;;
+ "")
+ # If it fails, then pretend we aren't using GCC.
+ ac_prog=ld
+ ;;
+ *)
+ # If it is relative, then search for the first ld in PATH.
+ with_gnu_ld=unknown
+ ;;
+ esac
+elif test "$with_gnu_ld" = yes; then
+ AC_MSG_CHECKING([for GNU ld])
+else
+ AC_MSG_CHECKING([for non-GNU ld])
+fi
+AC_CACHE_VAL([acl_cv_path_LD],
+[if test -z "$LD"; then
+ IFS="${IFS= }"; ac_save_ifs="$IFS"; IFS="${IFS}${PATH_SEPARATOR-:}"
+ for ac_dir in $PATH; do
+ test -z "$ac_dir" && ac_dir=.
+ if test -f "$ac_dir/$ac_prog" || test -f "$ac_dir/$ac_prog$ac_exeext"; then
+ acl_cv_path_LD="$ac_dir/$ac_prog"
+ # Check to see if the program is GNU ld. I'd rather use --version,
+ # but apparently some GNU ld's only accept -v.
+ # Break only if it was the GNU/non-GNU ld that we prefer.
+ case `"$acl_cv_path_LD" -v 2>&1 < /dev/null` in
+ *GNU* | *'with BFD'*)
+ test "$with_gnu_ld" != no && break ;;
+ *)
+ test "$with_gnu_ld" != yes && break ;;
+ esac
+ fi
+ done
+ IFS="$ac_save_ifs"
+else
+ acl_cv_path_LD="$LD" # Let the user override the test with a path.
+fi])
+LD="$acl_cv_path_LD"
+if test -n "$LD"; then
+ AC_MSG_RESULT([$LD])
+else
+ AC_MSG_RESULT([no])
+fi
+test -z "$LD" && AC_MSG_ERROR([no acceptable ld found in \$PATH])
+AC_LIB_PROG_LD_GNU
+])
diff --git a/m4/lib-link.m4 b/m4/lib-link.m4
new file mode 100644
index 0000000..2144203
--- /dev/null
+++ b/m4/lib-link.m4
@@ -0,0 +1,761 @@
+# lib-link.m4 serial 19 (gettext-0.18)
+dnl Copyright (C) 2001-2009 Free Software Foundation, Inc.
+dnl This file is free software; the Free Software Foundation
+dnl gives unlimited permission to copy and/or distribute it,
+dnl with or without modifications, as long as this notice is preserved.
+
+dnl From Bruno Haible.
+
+AC_PREREQ([2.54])
+
+dnl AC_LIB_LINKFLAGS(name [, dependencies]) searches for libname and
+dnl the libraries corresponding to explicit and implicit dependencies.
+dnl Sets and AC_SUBSTs the LIB${NAME} and LTLIB${NAME} variables and
+dnl augments the CPPFLAGS variable.
+dnl Sets and AC_SUBSTs the LIB${NAME}_PREFIX variable to nonempty if libname
+dnl was found in ${LIB${NAME}_PREFIX}/$acl_libdirstem.
+AC_DEFUN([AC_LIB_LINKFLAGS],
+[
+ AC_REQUIRE([AC_LIB_PREPARE_PREFIX])
+ AC_REQUIRE([AC_LIB_RPATH])
+ pushdef([Name],[translit([$1],[./-], [___])])
+ pushdef([NAME],[translit([$1],[abcdefghijklmnopqrstuvwxyz./-],
+ [ABCDEFGHIJKLMNOPQRSTUVWXYZ___])])
+ AC_CACHE_CHECK([how to link with lib[]$1], [ac_cv_lib[]Name[]_libs], [
+ AC_LIB_LINKFLAGS_BODY([$1], [$2])
+ ac_cv_lib[]Name[]_libs="$LIB[]NAME"
+ ac_cv_lib[]Name[]_ltlibs="$LTLIB[]NAME"
+ ac_cv_lib[]Name[]_cppflags="$INC[]NAME"
+ ac_cv_lib[]Name[]_prefix="$LIB[]NAME[]_PREFIX"
+ ])
+ LIB[]NAME="$ac_cv_lib[]Name[]_libs"
+ LTLIB[]NAME="$ac_cv_lib[]Name[]_ltlibs"
+ INC[]NAME="$ac_cv_lib[]Name[]_cppflags"
+ LIB[]NAME[]_PREFIX="$ac_cv_lib[]Name[]_prefix"
+ AC_LIB_APPENDTOVAR([CPPFLAGS], [$INC]NAME)
+ AC_SUBST([LIB]NAME)
+ AC_SUBST([LTLIB]NAME)
+ AC_SUBST([LIB]NAME[_PREFIX])
+ dnl Also set HAVE_LIB[]NAME so that AC_LIB_HAVE_LINKFLAGS can reuse the
+ dnl results of this search when this library appears as a dependency.
+ HAVE_LIB[]NAME=yes
+ popdef([NAME])
+ popdef([Name])
+])
+
+dnl AC_LIB_HAVE_LINKFLAGS(name, dependencies, includes, testcode,
[missing-message])
+dnl searches for libname and the libraries corresponding to explicit and
+dnl implicit dependencies, together with the specified include files and
+dnl the ability to compile and link the specified testcode. The missing-message
+dnl defaults to 'no' and may contain additional hints for the user.
+dnl If found, it sets and AC_SUBSTs HAVE_LIB${NAME}=yes and the LIB${NAME}
+dnl and LTLIB${NAME} variables and augments the CPPFLAGS variable, and
+dnl #defines HAVE_LIB${NAME} to 1. Otherwise, it sets and AC_SUBSTs
+dnl HAVE_LIB${NAME}=no and LIB${NAME} and LTLIB${NAME} to empty.
+dnl Sets and AC_SUBSTs the LIB${NAME}_PREFIX variable to nonempty if libname
+dnl was found in ${LIB${NAME}_PREFIX}/$acl_libdirstem.
+AC_DEFUN([AC_LIB_HAVE_LINKFLAGS],
+[
+ AC_REQUIRE([AC_LIB_PREPARE_PREFIX])
+ AC_REQUIRE([AC_LIB_RPATH])
+ pushdef([Name],[translit([$1],[./-], [___])])
+ pushdef([NAME],[translit([$1],[abcdefghijklmnopqrstuvwxyz./-],
+ [ABCDEFGHIJKLMNOPQRSTUVWXYZ___])])
+
+ dnl Search for lib[]Name and define LIB[]NAME, LTLIB[]NAME and INC[]NAME
+ dnl accordingly.
+ AC_LIB_LINKFLAGS_BODY([$1], [$2])
+
+ dnl Add $INC[]NAME to CPPFLAGS before performing the following checks,
+ dnl because if the user has installed lib[]Name and not disabled its use
+ dnl via --without-lib[]Name-prefix, he wants to use it.
+ ac_save_CPPFLAGS="$CPPFLAGS"
+ AC_LIB_APPENDTOVAR([CPPFLAGS], [$INC]NAME)
+
+ AC_CACHE_CHECK([for lib[]$1], [ac_cv_lib[]Name], [
+ ac_save_LIBS="$LIBS"
+ LIBS="$LIBS $LIB[]NAME"
+ AC_TRY_LINK([$3], [$4],
+ [ac_cv_lib[]Name=yes],
+ [ac_cv_lib[]Name='m4_if([$5], [], [no], [[$5]])'])
+ LIBS="$ac_save_LIBS"
+ ])
+ if test "$ac_cv_lib[]Name" = yes; then
+ HAVE_LIB[]NAME=yes
+ AC_DEFINE([HAVE_LIB]NAME, 1, [Define if you have the lib[]$1 library.])
+ AC_MSG_CHECKING([how to link with lib[]$1])
+ AC_MSG_RESULT([$LIB[]NAME])
+ else
+ HAVE_LIB[]NAME=no
+ dnl If $LIB[]NAME didn't lead to a usable library, we don't need
+ dnl $INC[]NAME either.
+ CPPFLAGS="$ac_save_CPPFLAGS"
+ LIB[]NAME=
+ LTLIB[]NAME=
+ LIB[]NAME[]_PREFIX=
+ fi
+ AC_SUBST([HAVE_LIB]NAME)
+ AC_SUBST([LIB]NAME)
+ AC_SUBST([LTLIB]NAME)
+ AC_SUBST([LIB]NAME[_PREFIX])
+ popdef([NAME])
+ popdef([Name])
+])
+
+dnl Determine the platform dependent parameters needed to use rpath:
+dnl acl_libext,
+dnl acl_shlibext,
+dnl acl_hardcode_libdir_flag_spec,
+dnl acl_hardcode_libdir_separator,
+dnl acl_hardcode_direct,
+dnl acl_hardcode_minus_L.
+AC_DEFUN([AC_LIB_RPATH],
+[
+ dnl Tell automake >= 1.10 to complain if config.rpath is missing.
+ m4_ifdef([AC_REQUIRE_AUX_FILE], [AC_REQUIRE_AUX_FILE([config.rpath])])
+ AC_REQUIRE([AC_PROG_CC]) dnl we use $CC, $GCC, $LDFLAGS
+ AC_REQUIRE([AC_LIB_PROG_LD]) dnl we use $LD, $with_gnu_ld
+ AC_REQUIRE([AC_CANONICAL_HOST]) dnl we use $host
+ AC_REQUIRE([AC_CONFIG_AUX_DIR_DEFAULT]) dnl we use $ac_aux_dir
+ AC_CACHE_CHECK([for shared library run path origin], [acl_cv_rpath], [
+ CC="$CC" GCC="$GCC" LDFLAGS="$LDFLAGS" LD="$LD" with_gnu_ld="$with_gnu_ld"
\
+ ${CONFIG_SHELL-/bin/sh} "$ac_aux_dir/config.rpath" "$host" > conftest.sh
+ . ./conftest.sh
+ rm -f ./conftest.sh
+ acl_cv_rpath=done
+ ])
+ wl="$acl_cv_wl"
+ acl_libext="$acl_cv_libext"
+ acl_shlibext="$acl_cv_shlibext"
+ acl_libname_spec="$acl_cv_libname_spec"
+ acl_library_names_spec="$acl_cv_library_names_spec"
+ acl_hardcode_libdir_flag_spec="$acl_cv_hardcode_libdir_flag_spec"
+ acl_hardcode_libdir_separator="$acl_cv_hardcode_libdir_separator"
+ acl_hardcode_direct="$acl_cv_hardcode_direct"
+ acl_hardcode_minus_L="$acl_cv_hardcode_minus_L"
+ dnl Determine whether the user wants rpath handling at all.
+ AC_ARG_ENABLE([rpath],
+ [ --disable-rpath do not hardcode runtime library paths],
+ :, enable_rpath=yes)
+])
+
+dnl AC_LIB_FROMPACKAGE(name, package)
+dnl declares that libname comes from the given package. The configure file
+dnl will then not have a --with-libname-prefix option but a
+dnl --with-package-prefix option. Several libraries can come from the same
+dnl package. This declaration must occur before an AC_LIB_LINKFLAGS or similar
+dnl macro call that searches for libname.
+AC_DEFUN([AC_LIB_FROMPACKAGE],
+[
+ pushdef([NAME],[translit([$1],[abcdefghijklmnopqrstuvwxyz./-],
+ [ABCDEFGHIJKLMNOPQRSTUVWXYZ___])])
+ define([acl_frompackage_]NAME, [$2])
+ popdef([NAME])
+ pushdef([PACK],[$2])
+ pushdef([PACKUP],[translit(PACK,[abcdefghijklmnopqrstuvwxyz./-],
+ [ABCDEFGHIJKLMNOPQRSTUVWXYZ___])])
+ define([acl_libsinpackage_]PACKUP,
+ m4_ifdef([acl_libsinpackage_]PACKUP, [acl_libsinpackage_]PACKUP[[,
]],)[lib$1])
+ popdef([PACKUP])
+ popdef([PACK])
+])
+
+dnl AC_LIB_LINKFLAGS_BODY(name [, dependencies]) searches for libname and
+dnl the libraries corresponding to explicit and implicit dependencies.
+dnl Sets the LIB${NAME}, LTLIB${NAME} and INC${NAME} variables.
+dnl Also, sets the LIB${NAME}_PREFIX variable to nonempty if libname was found
+dnl in ${LIB${NAME}_PREFIX}/$acl_libdirstem.
+AC_DEFUN([AC_LIB_LINKFLAGS_BODY],
+[
+ AC_REQUIRE([AC_LIB_PREPARE_MULTILIB])
+ pushdef([NAME],[translit([$1],[abcdefghijklmnopqrstuvwxyz./-],
+ [ABCDEFGHIJKLMNOPQRSTUVWXYZ___])])
+ pushdef([PACK],[m4_ifdef([acl_frompackage_]NAME, [acl_frompackage_]NAME,
lib[$1])])
+ pushdef([PACKUP],[translit(PACK,[abcdefghijklmnopqrstuvwxyz./-],
+ [ABCDEFGHIJKLMNOPQRSTUVWXYZ___])])
+ pushdef([PACKLIBS],[m4_ifdef([acl_frompackage_]NAME,
[acl_libsinpackage_]PACKUP, lib[$1])])
+ dnl Autoconf >= 2.61 supports dots in --with options.
+
pushdef([P_A_C_K],[m4_if(m4_version_compare(m4_defn([m4_PACKAGE_VERSION]),[2.61]),[-1],[translit(PACK,[.],[_])],PACK)])
+ dnl By default, look in $includedir and $libdir.
+ use_additional=yes
+ AC_LIB_WITH_FINAL_PREFIX([
+ eval additional_includedir=\"$includedir\"
+ eval additional_libdir=\"$libdir\"
+ ])
+ AC_ARG_WITH(P_A_C_K[-prefix],
+[[ --with-]]P_A_C_K[[-prefix[=DIR] search for ]PACKLIBS[ in DIR/include and
DIR/lib
+ --without-]]P_A_C_K[[-prefix don't search for ]PACKLIBS[ in includedir
and libdir]],
+[
+ if test "X$withval" = "Xno"; then
+ use_additional=no
+ else
+ if test "X$withval" = "X"; then
+ AC_LIB_WITH_FINAL_PREFIX([
+ eval additional_includedir=\"$includedir\"
+ eval additional_libdir=\"$libdir\"
+ ])
+ else
+ additional_includedir="$withval/include"
+ additional_libdir="$withval/$acl_libdirstem"
+ if test "$acl_libdirstem2" != "$acl_libdirstem" \
+ && ! test -d "$withval/$acl_libdirstem"; then
+ additional_libdir="$withval/$acl_libdirstem2"
+ fi
+ fi
+ fi
+])
+ dnl Search the library and its dependencies in $additional_libdir and
+ dnl $LDFLAGS. Using breadth-first-seach.
+ LIB[]NAME=
+ LTLIB[]NAME=
+ INC[]NAME=
+ LIB[]NAME[]_PREFIX=
+ rpathdirs=
+ ltrpathdirs=
+ names_already_handled=
+ names_next_round='$1 $2'
+ while test -n "$names_next_round"; do
+ names_this_round="$names_next_round"
+ names_next_round=
+ for name in $names_this_round; do
+ already_handled=
+ for n in $names_already_handled; do
+ if test "$n" = "$name"; then
+ already_handled=yes
+ break
+ fi
+ done
+ if test -z "$already_handled"; then
+ names_already_handled="$names_already_handled $name"
+ dnl See if it was already located by an earlier AC_LIB_LINKFLAGS
+ dnl or AC_LIB_HAVE_LINKFLAGS call.
+ uppername=`echo "$name" | sed -e
'y|abcdefghijklmnopqrstuvwxyz./-|ABCDEFGHIJKLMNOPQRSTUVWXYZ___|'`
+ eval value=\"\$HAVE_LIB$uppername\"
+ if test -n "$value"; then
+ if test "$value" = yes; then
+ eval value=\"\$LIB$uppername\"
+ test -z "$value" || LIB[]NAME="${LIB[]NAME}${LIB[]NAME:+ }$value"
+ eval value=\"\$LTLIB$uppername\"
+ test -z "$value" || LTLIB[]NAME="${LTLIB[]NAME}${LTLIB[]NAME:+
}$value"
+ else
+ dnl An earlier call to AC_LIB_HAVE_LINKFLAGS has determined
+ dnl that this library doesn't exist. So just drop it.
+ :
+ fi
+ else
+ dnl Search the library lib$name in $additional_libdir and $LDFLAGS
+ dnl and the already constructed $LIBNAME/$LTLIBNAME.
+ found_dir=
+ found_la=
+ found_so=
+ found_a=
+ eval libname=\"$acl_libname_spec\" # typically: libname=lib$name
+ if test -n "$acl_shlibext"; then
+ shrext=".$acl_shlibext" # typically: shrext=.so
+ else
+ shrext=
+ fi
+ if test $use_additional = yes; then
+ dir="$additional_libdir"
+ dnl The same code as in the loop below:
+ dnl First look for a shared library.
+ if test -n "$acl_shlibext"; then
+ if test -f "$dir/$libname$shrext"; then
+ found_dir="$dir"
+ found_so="$dir/$libname$shrext"
+ else
+ if test "$acl_library_names_spec" =
'$libname$shrext$versuffix'; then
+ ver=`(cd "$dir" && \
+ for f in "$libname$shrext".*; do echo "$f"; done \
+ | sed -e "s,^$libname$shrext\\\\.,," \
+ | sort -t '.' -n -r -k1,1 -k2,2 -k3,3 -k4,4 -k5,5 \
+ | sed 1q ) 2>/dev/null`
+ if test -n "$ver" && test -f "$dir/$libname$shrext.$ver";
then
+ found_dir="$dir"
+ found_so="$dir/$libname$shrext.$ver"
+ fi
+ else
+ eval library_names=\"$acl_library_names_spec\"
+ for f in $library_names; do
+ if test -f "$dir/$f"; then
+ found_dir="$dir"
+ found_so="$dir/$f"
+ break
+ fi
+ done
+ fi
+ fi
+ fi
+ dnl Then look for a static library.
+ if test "X$found_dir" = "X"; then
+ if test -f "$dir/$libname.$acl_libext"; then
+ found_dir="$dir"
+ found_a="$dir/$libname.$acl_libext"
+ fi
+ fi
+ if test "X$found_dir" != "X"; then
+ if test -f "$dir/$libname.la"; then
+ found_la="$dir/$libname.la"
+ fi
+ fi
+ fi
+ if test "X$found_dir" = "X"; then
+ for x in $LDFLAGS $LTLIB[]NAME; do
+ AC_LIB_WITH_FINAL_PREFIX([eval x=\"$x\"])
+ case "$x" in
+ -L*)
+ dir=`echo "X$x" | sed -e 's/^X-L//'`
+ dnl First look for a shared library.
+ if test -n "$acl_shlibext"; then
+ if test -f "$dir/$libname$shrext"; then
+ found_dir="$dir"
+ found_so="$dir/$libname$shrext"
+ else
+ if test "$acl_library_names_spec" =
'$libname$shrext$versuffix'; then
+ ver=`(cd "$dir" && \
+ for f in "$libname$shrext".*; do echo "$f"; done
\
+ | sed -e "s,^$libname$shrext\\\\.,," \
+ | sort -t '.' -n -r -k1,1 -k2,2 -k3,3 -k4,4
-k5,5 \
+ | sed 1q ) 2>/dev/null`
+ if test -n "$ver" && test -f
"$dir/$libname$shrext.$ver"; then
+ found_dir="$dir"
+ found_so="$dir/$libname$shrext.$ver"
+ fi
+ else
+ eval library_names=\"$acl_library_names_spec\"
+ for f in $library_names; do
+ if test -f "$dir/$f"; then
+ found_dir="$dir"
+ found_so="$dir/$f"
+ break
+ fi
+ done
+ fi
+ fi
+ fi
+ dnl Then look for a static library.
+ if test "X$found_dir" = "X"; then
+ if test -f "$dir/$libname.$acl_libext"; then
+ found_dir="$dir"
+ found_a="$dir/$libname.$acl_libext"
+ fi
+ fi
+ if test "X$found_dir" != "X"; then
+ if test -f "$dir/$libname.la"; then
+ found_la="$dir/$libname.la"
+ fi
+ fi
+ ;;
+ esac
+ if test "X$found_dir" != "X"; then
+ break
+ fi
+ done
+ fi
+ if test "X$found_dir" != "X"; then
+ dnl Found the library.
+ LTLIB[]NAME="${LTLIB[]NAME}${LTLIB[]NAME:+ }-L$found_dir -l$name"
+ if test "X$found_so" != "X"; then
+ dnl Linking with a shared library. We attempt to hardcode its
+ dnl directory into the executable's runpath, unless it's the
+ dnl standard /usr/lib.
+ if test "$enable_rpath" = no \
+ || test "X$found_dir" = "X/usr/$acl_libdirstem" \
+ || test "X$found_dir" = "X/usr/$acl_libdirstem2"; then
+ dnl No hardcoding is needed.
+ LIB[]NAME="${LIB[]NAME}${LIB[]NAME:+ }$found_so"
+ else
+ dnl Use an explicit option to hardcode DIR into the resulting
+ dnl binary.
+ dnl Potentially add DIR to ltrpathdirs.
+ dnl The ltrpathdirs will be appended to $LTLIBNAME at the end.
+ haveit=
+ for x in $ltrpathdirs; do
+ if test "X$x" = "X$found_dir"; then
+ haveit=yes
+ break
+ fi
+ done
+ if test -z "$haveit"; then
+ ltrpathdirs="$ltrpathdirs $found_dir"
+ fi
+ dnl The hardcoding into $LIBNAME is system dependent.
+ if test "$acl_hardcode_direct" = yes; then
+ dnl Using DIR/libNAME.so during linking hardcodes DIR into
the
+ dnl resulting binary.
+ LIB[]NAME="${LIB[]NAME}${LIB[]NAME:+ }$found_so"
+ else
+ if test -n "$acl_hardcode_libdir_flag_spec" && test
"$acl_hardcode_minus_L" = no; then
+ dnl Use an explicit option to hardcode DIR into the
resulting
+ dnl binary.
+ LIB[]NAME="${LIB[]NAME}${LIB[]NAME:+ }$found_so"
+ dnl Potentially add DIR to rpathdirs.
+ dnl The rpathdirs will be appended to $LIBNAME at the end.
+ haveit=
+ for x in $rpathdirs; do
+ if test "X$x" = "X$found_dir"; then
+ haveit=yes
+ break
+ fi
+ done
+ if test -z "$haveit"; then
+ rpathdirs="$rpathdirs $found_dir"
+ fi
+ else
+ dnl Rely on "-L$found_dir".
+ dnl But don't add it if it's already contained in the
LDFLAGS
+ dnl or the already constructed $LIBNAME
+ haveit=
+ for x in $LDFLAGS $LIB[]NAME; do
+ AC_LIB_WITH_FINAL_PREFIX([eval x=\"$x\"])
+ if test "X$x" = "X-L$found_dir"; then
+ haveit=yes
+ break
+ fi
+ done
+ if test -z "$haveit"; then
+ LIB[]NAME="${LIB[]NAME}${LIB[]NAME:+ }-L$found_dir"
+ fi
+ if test "$acl_hardcode_minus_L" != no; then
+ dnl FIXME: Not sure whether we should use
+ dnl "-L$found_dir -l$name" or "-L$found_dir $found_so"
+ dnl here.
+ LIB[]NAME="${LIB[]NAME}${LIB[]NAME:+ }$found_so"
+ else
+ dnl We cannot use $acl_hardcode_runpath_var and
LD_RUN_PATH
+ dnl here, because this doesn't fit in flags passed to the
+ dnl compiler. So give up. No hardcoding. This affects
only
+ dnl very old systems.
+ dnl FIXME: Not sure whether we should use
+ dnl "-L$found_dir -l$name" or "-L$found_dir $found_so"
+ dnl here.
+ LIB[]NAME="${LIB[]NAME}${LIB[]NAME:+ }-l$name"
+ fi
+ fi
+ fi
+ fi
+ else
+ if test "X$found_a" != "X"; then
+ dnl Linking with a static library.
+ LIB[]NAME="${LIB[]NAME}${LIB[]NAME:+ }$found_a"
+ else
+ dnl We shouldn't come here, but anyway it's good to have a
+ dnl fallback.
+ LIB[]NAME="${LIB[]NAME}${LIB[]NAME:+ }-L$found_dir -l$name"
+ fi
+ fi
+ dnl Assume the include files are nearby.
+ additional_includedir=
+ case "$found_dir" in
+ */$acl_libdirstem | */$acl_libdirstem/)
+ basedir=`echo "X$found_dir" | sed -e 's,^X,,' -e
"s,/$acl_libdirstem/"'*$,,'`
+ if test "$name" = '$1'; then
+ LIB[]NAME[]_PREFIX="$basedir"
+ fi
+ additional_includedir="$basedir/include"
+ ;;
+ */$acl_libdirstem2 | */$acl_libdirstem2/)
+ basedir=`echo "X$found_dir" | sed -e 's,^X,,' -e
"s,/$acl_libdirstem2/"'*$,,'`
+ if test "$name" = '$1'; then
+ LIB[]NAME[]_PREFIX="$basedir"
+ fi
+ additional_includedir="$basedir/include"
+ ;;
+ esac
+ if test "X$additional_includedir" != "X"; then
+ dnl Potentially add $additional_includedir to $INCNAME.
+ dnl But don't add it
+ dnl 1. if it's the standard /usr/include,
+ dnl 2. if it's /usr/local/include and we are using GCC on
Linux,
+ dnl 3. if it's already present in $CPPFLAGS or the already
+ dnl constructed $INCNAME,
+ dnl 4. if it doesn't exist as a directory.
+ if test "X$additional_includedir" != "X/usr/include"; then
+ haveit=
+ if test "X$additional_includedir" = "X/usr/local/include"; then
+ if test -n "$GCC"; then
+ case $host_os in
+ linux* | gnu* | k*bsd*-gnu) haveit=yes;;
+ esac
+ fi
+ fi
+ if test -z "$haveit"; then
+ for x in $CPPFLAGS $INC[]NAME; do
+ AC_LIB_WITH_FINAL_PREFIX([eval x=\"$x\"])
+ if test "X$x" = "X-I$additional_includedir"; then
+ haveit=yes
+ break
+ fi
+ done
+ if test -z "$haveit"; then
+ if test -d "$additional_includedir"; then
+ dnl Really add $additional_includedir to $INCNAME.
+ INC[]NAME="${INC[]NAME}${INC[]NAME:+
}-I$additional_includedir"
+ fi
+ fi
+ fi
+ fi
+ fi
+ dnl Look for dependencies.
+ if test -n "$found_la"; then
+ dnl Read the .la file. It defines the variables
+ dnl dlname, library_names, old_library, dependency_libs, current,
+ dnl age, revision, installed, dlopen, dlpreopen, libdir.
+ save_libdir="$libdir"
+ case "$found_la" in
+ */* | *\\*) . "$found_la" ;;
+ *) . "./$found_la" ;;
+ esac
+ libdir="$save_libdir"
+ dnl We use only dependency_libs.
+ for dep in $dependency_libs; do
+ case "$dep" in
+ -L*)
+ additional_libdir=`echo "X$dep" | sed -e 's/^X-L//'`
+ dnl Potentially add $additional_libdir to $LIBNAME and
$LTLIBNAME.
+ dnl But don't add it
+ dnl 1. if it's the standard /usr/lib,
+ dnl 2. if it's /usr/local/lib and we are using GCC on
Linux,
+ dnl 3. if it's already present in $LDFLAGS or the already
+ dnl constructed $LIBNAME,
+ dnl 4. if it doesn't exist as a directory.
+ if test "X$additional_libdir" != "X/usr/$acl_libdirstem" \
+ && test "X$additional_libdir" !=
"X/usr/$acl_libdirstem2"; then
+ haveit=
+ if test "X$additional_libdir" =
"X/usr/local/$acl_libdirstem" \
+ || test "X$additional_libdir" =
"X/usr/local/$acl_libdirstem2"; then
+ if test -n "$GCC"; then
+ case $host_os in
+ linux* | gnu* | k*bsd*-gnu) haveit=yes;;
+ esac
+ fi
+ fi
+ if test -z "$haveit"; then
+ haveit=
+ for x in $LDFLAGS $LIB[]NAME; do
+ AC_LIB_WITH_FINAL_PREFIX([eval x=\"$x\"])
+ if test "X$x" = "X-L$additional_libdir"; then
+ haveit=yes
+ break
+ fi
+ done
+ if test -z "$haveit"; then
+ if test -d "$additional_libdir"; then
+ dnl Really add $additional_libdir to $LIBNAME.
+ LIB[]NAME="${LIB[]NAME}${LIB[]NAME:+
}-L$additional_libdir"
+ fi
+ fi
+ haveit=
+ for x in $LDFLAGS $LTLIB[]NAME; do
+ AC_LIB_WITH_FINAL_PREFIX([eval x=\"$x\"])
+ if test "X$x" = "X-L$additional_libdir"; then
+ haveit=yes
+ break
+ fi
+ done
+ if test -z "$haveit"; then
+ if test -d "$additional_libdir"; then
+ dnl Really add $additional_libdir to $LTLIBNAME.
+ LTLIB[]NAME="${LTLIB[]NAME}${LTLIB[]NAME:+
}-L$additional_libdir"
+ fi
+ fi
+ fi
+ fi
+ ;;
+ -R*)
+ dir=`echo "X$dep" | sed -e 's/^X-R//'`
+ if test "$enable_rpath" != no; then
+ dnl Potentially add DIR to rpathdirs.
+ dnl The rpathdirs will be appended to $LIBNAME at the
end.
+ haveit=
+ for x in $rpathdirs; do
+ if test "X$x" = "X$dir"; then
+ haveit=yes
+ break
+ fi
+ done
+ if test -z "$haveit"; then
+ rpathdirs="$rpathdirs $dir"
+ fi
+ dnl Potentially add DIR to ltrpathdirs.
+ dnl The ltrpathdirs will be appended to $LTLIBNAME at
the end.
+ haveit=
+ for x in $ltrpathdirs; do
+ if test "X$x" = "X$dir"; then
+ haveit=yes
+ break
+ fi
+ done
+ if test -z "$haveit"; then
+ ltrpathdirs="$ltrpathdirs $dir"
+ fi
+ fi
+ ;;
+ -l*)
+ dnl Handle this in the next round.
+ names_next_round="$names_next_round "`echo "X$dep" | sed
-e 's/^X-l//'`
+ ;;
+ *.la)
+ dnl Handle this in the next round. Throw away the .la's
+ dnl directory; it is already contained in a preceding -L
+ dnl option.
+ names_next_round="$names_next_round "`echo "X$dep" | sed
-e 's,^X.*/,,' -e 's,^lib,,' -e 's,\.la$,,'`
+ ;;
+ *)
+ dnl Most likely an immediate library name.
+ LIB[]NAME="${LIB[]NAME}${LIB[]NAME:+ }$dep"
+ LTLIB[]NAME="${LTLIB[]NAME}${LTLIB[]NAME:+ }$dep"
+ ;;
+ esac
+ done
+ fi
+ else
+ dnl Didn't find the library; assume it is in the system directories
+ dnl known to the linker and runtime loader. (All the system
+ dnl directories known to the linker should also be known to the
+ dnl runtime loader, otherwise the system is severely
misconfigured.)
+ LIB[]NAME="${LIB[]NAME}${LIB[]NAME:+ }-l$name"
+ LTLIB[]NAME="${LTLIB[]NAME}${LTLIB[]NAME:+ }-l$name"
+ fi
+ fi
+ fi
+ done
+ done
+ if test "X$rpathdirs" != "X"; then
+ if test -n "$acl_hardcode_libdir_separator"; then
+ dnl Weird platform: only the last -rpath option counts, the user must
+ dnl pass all path elements in one option. We can arrange that for a
+ dnl single library, but not when more than one $LIBNAMEs are used.
+ alldirs=
+ for found_dir in $rpathdirs; do
+
alldirs="${alldirs}${alldirs:+$acl_hardcode_libdir_separator}$found_dir"
+ done
+ dnl Note: acl_hardcode_libdir_flag_spec uses $libdir and $wl.
+ acl_save_libdir="$libdir"
+ libdir="$alldirs"
+ eval flag=\"$acl_hardcode_libdir_flag_spec\"
+ libdir="$acl_save_libdir"
+ LIB[]NAME="${LIB[]NAME}${LIB[]NAME:+ }$flag"
+ else
+ dnl The -rpath options are cumulative.
+ for found_dir in $rpathdirs; do
+ acl_save_libdir="$libdir"
+ libdir="$found_dir"
+ eval flag=\"$acl_hardcode_libdir_flag_spec\"
+ libdir="$acl_save_libdir"
+ LIB[]NAME="${LIB[]NAME}${LIB[]NAME:+ }$flag"
+ done
+ fi
+ fi
+ if test "X$ltrpathdirs" != "X"; then
+ dnl When using libtool, the option that works for both libraries and
+ dnl executables is -R. The -R options are cumulative.
+ for found_dir in $ltrpathdirs; do
+ LTLIB[]NAME="${LTLIB[]NAME}${LTLIB[]NAME:+ }-R$found_dir"
+ done
+ fi
+ popdef([P_A_C_K])
+ popdef([PACKLIBS])
+ popdef([PACKUP])
+ popdef([PACK])
+ popdef([NAME])
+])
+
+dnl AC_LIB_APPENDTOVAR(VAR, CONTENTS) appends the elements of CONTENTS to VAR,
+dnl unless already present in VAR.
+dnl Works only for CPPFLAGS, not for LIB* variables because that sometimes
+dnl contains two or three consecutive elements that belong together.
+AC_DEFUN([AC_LIB_APPENDTOVAR],
+[
+ for element in [$2]; do
+ haveit=
+ for x in $[$1]; do
+ AC_LIB_WITH_FINAL_PREFIX([eval x=\"$x\"])
+ if test "X$x" = "X$element"; then
+ haveit=yes
+ break
+ fi
+ done
+ if test -z "$haveit"; then
+ [$1]="${[$1]}${[$1]:+ }$element"
+ fi
+ done
+])
+
+dnl For those cases where a variable contains several -L and -l options
+dnl referring to unknown libraries and directories, this macro determines the
+dnl necessary additional linker options for the runtime path.
+dnl AC_LIB_LINKFLAGS_FROM_LIBS([LDADDVAR], [LIBSVALUE], [USE-LIBTOOL])
+dnl sets LDADDVAR to linker options needed together with LIBSVALUE.
+dnl If USE-LIBTOOL evaluates to non-empty, linking with libtool is assumed,
+dnl otherwise linking without libtool is assumed.
+AC_DEFUN([AC_LIB_LINKFLAGS_FROM_LIBS],
+[
+ AC_REQUIRE([AC_LIB_RPATH])
+ AC_REQUIRE([AC_LIB_PREPARE_MULTILIB])
+ $1=
+ if test "$enable_rpath" != no; then
+ if test -n "$acl_hardcode_libdir_flag_spec" && test
"$acl_hardcode_minus_L" = no; then
+ dnl Use an explicit option to hardcode directories into the resulting
+ dnl binary.
+ rpathdirs=
+ next=
+ for opt in $2; do
+ if test -n "$next"; then
+ dir="$next"
+ dnl No need to hardcode the standard /usr/lib.
+ if test "X$dir" != "X/usr/$acl_libdirstem" \
+ && test "X$dir" != "X/usr/$acl_libdirstem2"; then
+ rpathdirs="$rpathdirs $dir"
+ fi
+ next=
+ else
+ case $opt in
+ -L) next=yes ;;
+ -L*) dir=`echo "X$opt" | sed -e 's,^X-L,,'`
+ dnl No need to hardcode the standard /usr/lib.
+ if test "X$dir" != "X/usr/$acl_libdirstem" \
+ && test "X$dir" != "X/usr/$acl_libdirstem2"; then
+ rpathdirs="$rpathdirs $dir"
+ fi
+ next= ;;
+ *) next= ;;
+ esac
+ fi
+ done
+ if test "X$rpathdirs" != "X"; then
+ if test -n ""$3""; then
+ dnl libtool is used for linking. Use -R options.
+ for dir in $rpathdirs; do
+ $1="${$1}${$1:+ }-R$dir"
+ done
+ else
+ dnl The linker is used for linking directly.
+ if test -n "$acl_hardcode_libdir_separator"; then
+ dnl Weird platform: only the last -rpath option counts, the user
+ dnl must pass all path elements in one option.
+ alldirs=
+ for dir in $rpathdirs; do
+
alldirs="${alldirs}${alldirs:+$acl_hardcode_libdir_separator}$dir"
+ done
+ acl_save_libdir="$libdir"
+ libdir="$alldirs"
+ eval flag=\"$acl_hardcode_libdir_flag_spec\"
+ libdir="$acl_save_libdir"
+ $1="$flag"
+ else
+ dnl The -rpath options are cumulative.
+ for dir in $rpathdirs; do
+ acl_save_libdir="$libdir"
+ libdir="$dir"
+ eval flag=\"$acl_hardcode_libdir_flag_spec\"
+ libdir="$acl_save_libdir"
+ $1="${$1}${$1:+ }$flag"
+ done
+ fi
+ fi
+ fi
+ fi
+ fi
+ AC_SUBST([$1])
+])
diff --git a/m4/lib-prefix.m4 b/m4/lib-prefix.m4
new file mode 100644
index 0000000..4b7ee33
--- /dev/null
+++ b/m4/lib-prefix.m4
@@ -0,0 +1,224 @@
+# lib-prefix.m4 serial 7 (gettext-0.18)
+dnl Copyright (C) 2001-2005, 2008-2009 Free Software Foundation, Inc.
+dnl This file is free software; the Free Software Foundation
+dnl gives unlimited permission to copy and/or distribute it,
+dnl with or without modifications, as long as this notice is preserved.
+
+dnl From Bruno Haible.
+
+dnl AC_LIB_ARG_WITH is synonymous to AC_ARG_WITH in autoconf-2.13, and
+dnl similar to AC_ARG_WITH in autoconf 2.52...2.57 except that is doesn't
+dnl require excessive bracketing.
+ifdef([AC_HELP_STRING],
+[AC_DEFUN([AC_LIB_ARG_WITH], [AC_ARG_WITH([$1],[[$2]],[$3],[$4])])],
+[AC_DEFUN([AC_][LIB_ARG_WITH], [AC_ARG_WITH([$1],[$2],[$3],[$4])])])
+
+dnl AC_LIB_PREFIX adds to the CPPFLAGS and LDFLAGS the flags that are needed
+dnl to access previously installed libraries. The basic assumption is that
+dnl a user will want packages to use other packages he previously installed
+dnl with the same --prefix option.
+dnl This macro is not needed if only AC_LIB_LINKFLAGS is used to locate
+dnl libraries, but is otherwise very convenient.
+AC_DEFUN([AC_LIB_PREFIX],
+[
+ AC_BEFORE([$0], [AC_LIB_LINKFLAGS])
+ AC_REQUIRE([AC_PROG_CC])
+ AC_REQUIRE([AC_CANONICAL_HOST])
+ AC_REQUIRE([AC_LIB_PREPARE_MULTILIB])
+ AC_REQUIRE([AC_LIB_PREPARE_PREFIX])
+ dnl By default, look in $includedir and $libdir.
+ use_additional=yes
+ AC_LIB_WITH_FINAL_PREFIX([
+ eval additional_includedir=\"$includedir\"
+ eval additional_libdir=\"$libdir\"
+ ])
+ AC_LIB_ARG_WITH([lib-prefix],
+[ --with-lib-prefix[=DIR] search for libraries in DIR/include and DIR/lib
+ --without-lib-prefix don't search for libraries in includedir and libdir],
+[
+ if test "X$withval" = "Xno"; then
+ use_additional=no
+ else
+ if test "X$withval" = "X"; then
+ AC_LIB_WITH_FINAL_PREFIX([
+ eval additional_includedir=\"$includedir\"
+ eval additional_libdir=\"$libdir\"
+ ])
+ else
+ additional_includedir="$withval/include"
+ additional_libdir="$withval/$acl_libdirstem"
+ fi
+ fi
+])
+ if test $use_additional = yes; then
+ dnl Potentially add $additional_includedir to $CPPFLAGS.
+ dnl But don't add it
+ dnl 1. if it's the standard /usr/include,
+ dnl 2. if it's already present in $CPPFLAGS,
+ dnl 3. if it's /usr/local/include and we are using GCC on Linux,
+ dnl 4. if it doesn't exist as a directory.
+ if test "X$additional_includedir" != "X/usr/include"; then
+ haveit=
+ for x in $CPPFLAGS; do
+ AC_LIB_WITH_FINAL_PREFIX([eval x=\"$x\"])
+ if test "X$x" = "X-I$additional_includedir"; then
+ haveit=yes
+ break
+ fi
+ done
+ if test -z "$haveit"; then
+ if test "X$additional_includedir" = "X/usr/local/include"; then
+ if test -n "$GCC"; then
+ case $host_os in
+ linux* | gnu* | k*bsd*-gnu) haveit=yes;;
+ esac
+ fi
+ fi
+ if test -z "$haveit"; then
+ if test -d "$additional_includedir"; then
+ dnl Really add $additional_includedir to $CPPFLAGS.
+ CPPFLAGS="${CPPFLAGS}${CPPFLAGS:+ }-I$additional_includedir"
+ fi
+ fi
+ fi
+ fi
+ dnl Potentially add $additional_libdir to $LDFLAGS.
+ dnl But don't add it
+ dnl 1. if it's the standard /usr/lib,
+ dnl 2. if it's already present in $LDFLAGS,
+ dnl 3. if it's /usr/local/lib and we are using GCC on Linux,
+ dnl 4. if it doesn't exist as a directory.
+ if test "X$additional_libdir" != "X/usr/$acl_libdirstem"; then
+ haveit=
+ for x in $LDFLAGS; do
+ AC_LIB_WITH_FINAL_PREFIX([eval x=\"$x\"])
+ if test "X$x" = "X-L$additional_libdir"; then
+ haveit=yes
+ break
+ fi
+ done
+ if test -z "$haveit"; then
+ if test "X$additional_libdir" = "X/usr/local/$acl_libdirstem"; then
+ if test -n "$GCC"; then
+ case $host_os in
+ linux*) haveit=yes;;
+ esac
+ fi
+ fi
+ if test -z "$haveit"; then
+ if test -d "$additional_libdir"; then
+ dnl Really add $additional_libdir to $LDFLAGS.
+ LDFLAGS="${LDFLAGS}${LDFLAGS:+ }-L$additional_libdir"
+ fi
+ fi
+ fi
+ fi
+ fi
+])
+
+dnl AC_LIB_PREPARE_PREFIX creates variables acl_final_prefix,
+dnl acl_final_exec_prefix, containing the values to which $prefix and
+dnl $exec_prefix will expand at the end of the configure script.
+AC_DEFUN([AC_LIB_PREPARE_PREFIX],
+[
+ dnl Unfortunately, prefix and exec_prefix get only finally determined
+ dnl at the end of configure.
+ if test "X$prefix" = "XNONE"; then
+ acl_final_prefix="$ac_default_prefix"
+ else
+ acl_final_prefix="$prefix"
+ fi
+ if test "X$exec_prefix" = "XNONE"; then
+ acl_final_exec_prefix='${prefix}'
+ else
+ acl_final_exec_prefix="$exec_prefix"
+ fi
+ acl_save_prefix="$prefix"
+ prefix="$acl_final_prefix"
+ eval acl_final_exec_prefix=\"$acl_final_exec_prefix\"
+ prefix="$acl_save_prefix"
+])
+
+dnl AC_LIB_WITH_FINAL_PREFIX([statement]) evaluates statement, with the
+dnl variables prefix and exec_prefix bound to the values they will have
+dnl at the end of the configure script.
+AC_DEFUN([AC_LIB_WITH_FINAL_PREFIX],
+[
+ acl_save_prefix="$prefix"
+ prefix="$acl_final_prefix"
+ acl_save_exec_prefix="$exec_prefix"
+ exec_prefix="$acl_final_exec_prefix"
+ $1
+ exec_prefix="$acl_save_exec_prefix"
+ prefix="$acl_save_prefix"
+])
+
+dnl AC_LIB_PREPARE_MULTILIB creates
+dnl - a variable acl_libdirstem, containing the basename of the libdir, either
+dnl "lib" or "lib64" or "lib/64",
+dnl - a variable acl_libdirstem2, as a secondary possible value for
+dnl acl_libdirstem, either the same as acl_libdirstem or "lib/sparcv9" or
+dnl "lib/amd64".
+AC_DEFUN([AC_LIB_PREPARE_MULTILIB],
+[
+ dnl There is no formal standard regarding lib and lib64.
+ dnl On glibc systems, the current practice is that on a system supporting
+ dnl 32-bit and 64-bit instruction sets or ABIs, 64-bit libraries go under
+ dnl $prefix/lib64 and 32-bit libraries go under $prefix/lib. We determine
+ dnl the compiler's default mode by looking at the compiler's library search
+ dnl path. If at least one of its elements ends in /lib64 or points to a
+ dnl directory whose absolute pathname ends in /lib64, we assume a 64-bit ABI.
+ dnl Otherwise we use the default, namely "lib".
+ dnl On Solaris systems, the current practice is that on a system supporting
+ dnl 32-bit and 64-bit instruction sets or ABIs, 64-bit libraries go under
+ dnl $prefix/lib/64 (which is a symlink to either $prefix/lib/sparcv9 or
+ dnl $prefix/lib/amd64) and 32-bit libraries go under $prefix/lib.
+ AC_REQUIRE([AC_CANONICAL_HOST])
+ acl_libdirstem=lib
+ acl_libdirstem2=
+ case "$host_os" in
+ solaris*)
+ dnl See Solaris 10 Software Developer Collection > Solaris 64-bit
Developer's Guide > The Development Environment
+ dnl <http://docs.sun.com/app/docs/doc/816-5138/dev-env?l=en&a=view>.
+ dnl "Portable Makefiles should refer to any library directories using
the 64 symbolic link."
+ dnl But we want to recognize the sparcv9 or amd64 subdirectory also if
the
+ dnl symlink is missing, so we set acl_libdirstem2 too.
+ AC_CACHE_CHECK([for 64-bit host], [gl_cv_solaris_64bit],
+ [AC_EGREP_CPP([sixtyfour bits], [
+#ifdef _LP64
+sixtyfour bits
+#endif
+ ], [gl_cv_solaris_64bit=yes], [gl_cv_solaris_64bit=no])
+ ])
+ if test $gl_cv_solaris_64bit = yes; then
+ acl_libdirstem=lib/64
+ case "$host_cpu" in
+ sparc*) acl_libdirstem2=lib/sparcv9 ;;
+ i*86 | x86_64) acl_libdirstem2=lib/amd64 ;;
+ esac
+ fi
+ ;;
+ *)
+ searchpath=`(LC_ALL=C $CC -print-search-dirs) 2>/dev/null | sed -n -e
's,^libraries: ,,p' | sed -e 's,^=,,'`
+ if test -n "$searchpath"; then
+ acl_save_IFS="${IFS= }"; IFS=":"
+ for searchdir in $searchpath; do
+ if test -d "$searchdir"; then
+ case "$searchdir" in
+ */lib64/ | */lib64 ) acl_libdirstem=lib64 ;;
+ */../ | */.. )
+ # Better ignore directories of this form. They are misleading.
+ ;;
+ *) searchdir=`cd "$searchdir" && pwd`
+ case "$searchdir" in
+ */lib64 ) acl_libdirstem=lib64 ;;
+ esac ;;
+ esac
+ fi
+ done
+ IFS="$acl_save_IFS"
+ fi
+ ;;
+ esac
+ test -n "$acl_libdirstem2" || acl_libdirstem2="$acl_libdirstem"
+])
diff --git a/m4/libunistring.m4 b/m4/libunistring.m4
new file mode 100644
index 0000000..52ff06b
--- /dev/null
+++ b/m4/libunistring.m4
@@ -0,0 +1,37 @@
+# libunistring.m4 serial 1
+dnl Copyright (C) 2009 Free Software Foundation, Inc.
+dnl This file is free software; the Free Software Foundation
+dnl gives unlimited permission to copy and/or distribute it,
+dnl with or without modifications, as long as this notice is preserved.
+
+dnl gl_LIBUNISTRING
+dnl Searches for an installed libunistring.
+dnl If found, it sets and AC_SUBSTs HAVE_LIBUNISTRING=yes and the LIBUNISTRING
+dnl and LTLIBUNISTRING variables and augments the CPPFLAGS variable, and
+dnl #defines HAVE_LIBUNISTRING to 1. Otherwise, it sets and AC_SUBSTs
+dnl HAVE_LIBUNISTRING=no and LIBUNINSTRING and LTLIBUNISTRING to empty.
+
+AC_DEFUN([gl_LIBUNISTRING],
+[
+ dnl First, try to link without -liconv. libunistring often depends on
+ dnl libiconv, but we don't know (and often don't need to know) where
+ dnl libiconv is installed.
+ AC_LIB_HAVE_LINKFLAGS([unistring], [],
+ [#include <uniconv.h>], [u8_strconv_from_locale((char*)0);],
+ [no, consider installing GNU libunistring])
+ if test "$ac_cv_libunistring" != yes; then
+ dnl Second try, with -liconv.
+ AC_REQUIRE([AM_ICONV])
+ if test -n "$LIBICONV"; then
+ glus_save_LIBS="$LIBS"
+ LIBS="$LIBS $LIBICONV"
+ AC_LIB_HAVE_LINKFLAGS([unistring], [],
+ [#include <uniconv.h>], [u8_strconv_from_locale((char*)0);],
+ [no, consider installing GNU libunistring])
+ if test -n "$LIBUNISTRING"; then
+ LIBUNISTRING="$LIBUNISTRING $LIBICONV"
+ fi
+ LIBS="$glus_save_LIBS"
+ fi
+ fi
+])
diff --git a/m4/string_h.m4 b/m4/string_h.m4
new file mode 100644
index 0000000..2d5553c
--- /dev/null
+++ b/m4/string_h.m4
@@ -0,0 +1,92 @@
+# Configure a GNU-like replacement for <string.h>.
+
+# Copyright (C) 2007, 2008 Free Software Foundation, Inc.
+# This file is free software; the Free Software Foundation
+# gives unlimited permission to copy and/or distribute it,
+# with or without modifications, as long as this notice is preserved.
+
+# serial 6
+
+# Written by Paul Eggert.
+
+AC_DEFUN([gl_HEADER_STRING_H],
+[
+ dnl Use AC_REQUIRE here, so that the default behavior below is expanded
+ dnl once only, before all statements that occur in other macros.
+ AC_REQUIRE([gl_HEADER_STRING_H_BODY])
+])
+
+AC_DEFUN([gl_HEADER_STRING_H_BODY],
+[
+ AC_REQUIRE([AC_C_RESTRICT])
+ AC_REQUIRE([gl_HEADER_STRING_H_DEFAULTS])
+ gl_CHECK_NEXT_HEADERS([string.h])
+])
+
+AC_DEFUN([gl_STRING_MODULE_INDICATOR],
+[
+ dnl Use AC_REQUIRE here, so that the default settings are expanded once only.
+ AC_REQUIRE([gl_HEADER_STRING_H_DEFAULTS])
+
GNULIB_[]m4_translit([$1],[abcdefghijklmnopqrstuvwxyz./-],[ABCDEFGHIJKLMNOPQRSTUVWXYZ___])=1
+])
+
+AC_DEFUN([gl_HEADER_STRING_H_DEFAULTS],
+[
+ GNULIB_MEMMEM=0; AC_SUBST([GNULIB_MEMMEM])
+ GNULIB_MEMPCPY=0; AC_SUBST([GNULIB_MEMPCPY])
+ GNULIB_MEMRCHR=0; AC_SUBST([GNULIB_MEMRCHR])
+ GNULIB_RAWMEMCHR=0; AC_SUBST([GNULIB_RAWMEMCHR])
+ GNULIB_STPCPY=0; AC_SUBST([GNULIB_STPCPY])
+ GNULIB_STPNCPY=0; AC_SUBST([GNULIB_STPNCPY])
+ GNULIB_STRCHRNUL=0; AC_SUBST([GNULIB_STRCHRNUL])
+ GNULIB_STRDUP=0; AC_SUBST([GNULIB_STRDUP])
+ GNULIB_STRNDUP=0; AC_SUBST([GNULIB_STRNDUP])
+ GNULIB_STRNLEN=0; AC_SUBST([GNULIB_STRNLEN])
+ GNULIB_STRPBRK=0; AC_SUBST([GNULIB_STRPBRK])
+ GNULIB_STRSEP=0; AC_SUBST([GNULIB_STRSEP])
+ GNULIB_STRSTR=0; AC_SUBST([GNULIB_STRSTR])
+ GNULIB_STRCASESTR=0; AC_SUBST([GNULIB_STRCASESTR])
+ GNULIB_STRTOK_R=0; AC_SUBST([GNULIB_STRTOK_R])
+ GNULIB_MBSLEN=0; AC_SUBST([GNULIB_MBSLEN])
+ GNULIB_MBSNLEN=0; AC_SUBST([GNULIB_MBSNLEN])
+ GNULIB_MBSCHR=0; AC_SUBST([GNULIB_MBSCHR])
+ GNULIB_MBSRCHR=0; AC_SUBST([GNULIB_MBSRCHR])
+ GNULIB_MBSSTR=0; AC_SUBST([GNULIB_MBSSTR])
+ GNULIB_MBSCASECMP=0; AC_SUBST([GNULIB_MBSCASECMP])
+ GNULIB_MBSNCASECMP=0; AC_SUBST([GNULIB_MBSNCASECMP])
+ GNULIB_MBSPCASECMP=0; AC_SUBST([GNULIB_MBSPCASECMP])
+ GNULIB_MBSCASESTR=0; AC_SUBST([GNULIB_MBSCASESTR])
+ GNULIB_MBSCSPN=0; AC_SUBST([GNULIB_MBSCSPN])
+ GNULIB_MBSPBRK=0; AC_SUBST([GNULIB_MBSPBRK])
+ GNULIB_MBSSPN=0; AC_SUBST([GNULIB_MBSSPN])
+ GNULIB_MBSSEP=0; AC_SUBST([GNULIB_MBSSEP])
+ GNULIB_MBSTOK_R=0; AC_SUBST([GNULIB_MBSTOK_R])
+ GNULIB_STRERROR=0; AC_SUBST([GNULIB_STRERROR])
+ GNULIB_STRSIGNAL=0; AC_SUBST([GNULIB_STRSIGNAL])
+ GNULIB_STRVERSCMP=0; AC_SUBST([GNULIB_STRVERSCMP])
+ dnl Assume proper GNU behavior unless another module says otherwise.
+ HAVE_DECL_MEMMEM=1; AC_SUBST([HAVE_DECL_MEMMEM])
+ HAVE_MEMPCPY=1; AC_SUBST([HAVE_MEMPCPY])
+ HAVE_DECL_MEMRCHR=1; AC_SUBST([HAVE_DECL_MEMRCHR])
+ HAVE_RAWMEMCHR=1; AC_SUBST([HAVE_RAWMEMCHR])
+ HAVE_STPCPY=1; AC_SUBST([HAVE_STPCPY])
+ HAVE_STPNCPY=1; AC_SUBST([HAVE_STPNCPY])
+ HAVE_STRCHRNUL=1; AC_SUBST([HAVE_STRCHRNUL])
+ HAVE_DECL_STRDUP=1; AC_SUBST([HAVE_DECL_STRDUP])
+ HAVE_STRNDUP=1; AC_SUBST([HAVE_STRNDUP])
+ HAVE_DECL_STRNDUP=1; AC_SUBST([HAVE_DECL_STRNDUP])
+ HAVE_DECL_STRNLEN=1; AC_SUBST([HAVE_DECL_STRNLEN])
+ HAVE_STRPBRK=1; AC_SUBST([HAVE_STRPBRK])
+ HAVE_STRSEP=1; AC_SUBST([HAVE_STRSEP])
+ HAVE_STRCASESTR=1; AC_SUBST([HAVE_STRCASESTR])
+ HAVE_DECL_STRTOK_R=1; AC_SUBST([HAVE_DECL_STRTOK_R])
+ HAVE_DECL_STRERROR=1; AC_SUBST([HAVE_DECL_STRERROR])
+ HAVE_DECL_STRSIGNAL=1; AC_SUBST([HAVE_DECL_STRSIGNAL])
+ HAVE_STRVERSCMP=1; AC_SUBST([HAVE_STRVERSCMP])
+ REPLACE_MEMMEM=0; AC_SUBST([REPLACE_MEMMEM])
+ REPLACE_STRDUP=0; AC_SUBST([REPLACE_STRDUP])
+ REPLACE_STRSTR=0; AC_SUBST([REPLACE_STRSTR])
+ REPLACE_STRCASESTR=0; AC_SUBST([REPLACE_STRCASESTR])
+ REPLACE_STRERROR=0; AC_SUBST([REPLACE_STRERROR])
+ REPLACE_STRSIGNAL=0; AC_SUBST([REPLACE_STRSIGNAL])
+])
diff --git a/module/Makefile.am b/module/Makefile.am
index 95dc75a..d149bb6 100644
--- a/module/Makefile.am
+++ b/module/Makefile.am
@@ -31,7 +31,7 @@ modpath =
# putting these core modules first.
SOURCES = \
- ice-9/psyntax-pp.scm \
+ ice-9/psyntax-pp.scm \
system/base/pmatch.scm system/base/syntax.scm
\
system/base/compile.scm system/base/language.scm \
\
@@ -53,6 +53,7 @@ SOURCES =
\
\
$(ICE_9_SOURCES) \
$(SRFI_SOURCES) \
+ $(RNRS_SOURCES) \
$(OOP_SOURCES) \
\
$(SCRIPTS_SOURCES)
@@ -209,6 +210,10 @@ SRFI_SOURCES = \
srfi/srfi-69.scm \
srfi/srfi-88.scm
+RNRS_SOURCES = \
+ rnrs/bytevector.scm \
+ rnrs/io/ports.scm
+
EXTRA_DIST += scripts/ChangeLog-2008
EXTRA_DIST += scripts/README
diff --git a/module/rnrs/bytevector.scm b/module/rnrs/bytevector.scm
new file mode 100644
index 0000000..793cbc0
--- /dev/null
+++ b/module/rnrs/bytevector.scm
@@ -0,0 +1,84 @@
+;;;; bytevector.scm --- R6RS bytevector API
+
+;;;; Copyright (C) 2009 Free Software Foundation, Inc.
+;;;;
+;;;; This library is free software; you can redistribute it and/or
+;;;; modify it under the terms of the GNU Lesser General Public
+;;;; License as published by the Free Software Foundation; either
+;;;; version 2.1 of the License, or (at your option) any later version.
+;;;;
+;;;; This library is distributed in the hope that it will be useful,
+;;;; but WITHOUT ANY WARRANTY; without even the implied warranty of
+;;;; MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+;;;; Lesser General Public License for more details.
+;;;;
+;;;; You should have received a copy of the GNU Lesser General Public
+;;;; License along with this library; if not, write to the Free Software
+;;;; Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301
USA
+
+;;; Author: Ludovic Courtès <address@hidden>
+
+;;; Commentary:
+;;;
+;;; A "bytevector" is a raw bit string. This module provides procedures to
+;;; manipulate bytevectors and interpret their contents in a number of ways:
+;;; bytevector contents can be accessed as signed or unsigned integer of
+;;; various sizes and endianness, as IEEE-754 floating point numbers, or as
+;;; strings. It is a useful tool to decode binary data.
+;;;
+;;; Code:
+
+(define-module (rnrs bytevector)
+ :export-syntax (endianness)
+ :export (native-endianness bytevector?
+ make-bytevector bytevector-length bytevector=? bytevector-fill!
+ bytevector-copy! bytevector-copy bytevector-u8-ref
+ bytevector-s8-ref
+ bytevector-u8-set! bytevector-s8-set! bytevector->u8-list
+ u8-list->bytevector
+ bytevector-uint-ref bytevector-uint-set!
+ bytevector-sint-ref bytevector-sint-set!
+ bytevector->sint-list bytevector->uint-list
+ uint-list->bytevector sint-list->bytevector
+
+ bytevector-u16-ref bytevector-s16-ref
+ bytevector-u16-set! bytevector-s16-set!
+ bytevector-u16-native-ref bytevector-s16-native-ref
+ bytevector-u16-native-set! bytevector-s16-native-set!
+
+ bytevector-u32-ref bytevector-s32-ref
+ bytevector-u32-set! bytevector-s32-set!
+ bytevector-u32-native-ref bytevector-s32-native-ref
+ bytevector-u32-native-set! bytevector-s32-native-set!
+
+ bytevector-u64-ref bytevector-s64-ref
+ bytevector-u64-set! bytevector-s64-set!
+ bytevector-u64-native-ref bytevector-s64-native-ref
+ bytevector-u64-native-set! bytevector-s64-native-set!
+
+ bytevector-ieee-single-ref
+ bytevector-ieee-single-set!
+ bytevector-ieee-single-native-ref
+ bytevector-ieee-single-native-set!
+
+ bytevector-ieee-double-ref
+ bytevector-ieee-double-set!
+ bytevector-ieee-double-native-ref
+ bytevector-ieee-double-native-set!
+
+ string->utf8 string->utf16 string->utf32
+ utf8->string utf16->string utf32->string))
+
+
+(load-extension "libguile" "scm_init_bytevectors")
+
+(define-macro (endianness sym)
+ (if (memq sym '(big little))
+ `(quote ,sym)
+ (error "unsupported endianness" sym)))
+
+;;; Local Variables:
+;;; coding: latin-1
+;;; End:
+
+;;; bytevector.scm ends here
diff --git a/module/rnrs/io/ports.scm b/module/rnrs/io/ports.scm
new file mode 100644
index 0000000..73843ee
--- /dev/null
+++ b/module/rnrs/io/ports.scm
@@ -0,0 +1,111 @@
+;;;; ports.scm --- R6RS port API
+
+;;;; Copyright (C) 2009 Free Software Foundation, Inc.
+;;;;
+;;;; This library is free software; you can redistribute it and/or
+;;;; modify it under the terms of the GNU Lesser General Public
+;;;; License as published by the Free Software Foundation; either
+;;;; version 2.1 of the License, or (at your option) any later version.
+;;;;
+;;;; This library is distributed in the hope that it will be useful,
+;;;; but WITHOUT ANY WARRANTY; without even the implied warranty of
+;;;; MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+;;;; Lesser General Public License for more details.
+;;;;
+;;;; You should have received a copy of the GNU Lesser General Public
+;;;; License along with this library; if not, write to the Free Software
+;;;; Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301
USA
+
+;;; Author: Ludovic Courtès <address@hidden>
+
+;;; Commentary:
+;;;
+;;; The I/O port API of the R6RS is provided by this module. In many areas
+;;; it complements or refines Guile's own historical port API. For instance,
+;;; it allows for binary I/O with bytevectors.
+;;;
+;;; Code:
+
+(define-module (rnrs io ports)
+ :re-export (eof-object? port? input-port? output-port?)
+ :export (eof-object
+
+ ;; input & output ports
+ port-transcoder binary-port? transcoded-port
+ port-position set-port-position!
+ port-has-port-position? port-has-set-port-position!?
+ call-with-port
+
+ ;; input ports
+ open-bytevector-input-port
+ make-custom-binary-input-port
+
+ ;; binary input
+ get-u8 lookahead-u8
+ get-bytevector-n get-bytevector-n!
+ get-bytevector-some get-bytevector-all
+
+ ;; output ports
+ open-bytevector-output-port
+ make-custom-binary-output-port
+
+ ;; binary output
+ put-u8 put-bytevector))
+
+(load-extension "libguile" "scm_init_r6rs_ports")
+
+
+
+;;;
+;;; Input and output ports.
+;;;
+
+(define (port-transcoder port)
+ (error "port transcoders are not supported" port))
+
+(define (binary-port? port)
+ ;; So far, we don't support transcoders other than the binary transcoder.
+ #t)
+
+(define (transcoded-port port)
+ (error "port transcoders are not supported" port))
+
+(define (port-position port)
+ "Return the offset (an integer) indicating where the next octet will be
+read from/written to in @var{port}."
+
+ ;; FIXME: We should raise an `&assertion' error when not supported.
+ (seek port 0 SEEK_CUR))
+
+(define (set-port-position! port offset)
+ "Set the position where the next octet will be read from/written to
address@hidden"
+
+ ;; FIXME: We should raise an `&assertion' error when not supported.
+ (seek port offset SEEK_SET))
+
+(define (port-has-port-position? port)
+ "Return @code{#t} is @var{port} supports @code{port-position}."
+ (and (false-if-exception (port-position port)) #t))
+
+(define (port-has-set-port-position!? port)
+ "Return @code{#t} is @var{port} supports @code{set-port-position!}."
+ (and (false-if-exception (set-port-position! port (port-position port)))
+ #t))
+
+(define (call-with-port port proc)
+ "Call @var{proc}, passing it @var{port} and closing @var{port} upon exit of
address@hidden Return the return values of @var{proc}."
+ (dynamic-wind
+ (lambda ()
+ #t)
+ (lambda ()
+ (proc port))
+ (lambda ()
+ (close-port port))))
+
+;;; Local Variables:
+;;; coding: latin-1
+;;; End:
+
+;;; ports.scm ends here
diff --git a/test-suite/Makefile.am b/test-suite/Makefile.am
index 3854d4a..0b986d4 100644
--- a/test-suite/Makefile.am
+++ b/test-suite/Makefile.am
@@ -26,6 +26,7 @@ SCM_TESTS = tests/alist.test \
tests/arbiters.test \
tests/asm-to-bytecode.test \
tests/bit-operations.test \
+ tests/bytevectors.test \
tests/c-api.test \
tests/chars.test \
tests/common-list.test \
@@ -62,6 +63,7 @@ SCM_TESTS = tests/alist.test \
tests/q.test \
tests/r4rs.test \
tests/r5rs_pitfall.test \
+ tests/r6rs-ports.test \
tests/ramap.test \
tests/reader.test \
tests/receive.test \
diff --git a/test-suite/tests/bytevectors.test
b/test-suite/tests/bytevectors.test
new file mode 100644
index 0000000..b2ae65c
--- /dev/null
+++ b/test-suite/tests/bytevectors.test
@@ -0,0 +1,531 @@
+;;;; bytevectors.test --- Exercise the R6RS bytevector API.
+;;;;
+;;;; Copyright (C) 2009 Free Software Foundation, Inc.
+;;;; Ludovic Courtès
+;;;;
+;;;; This library is free software; you can redistribute it and/or
+;;;; modify it under the terms of the GNU Lesser General Public
+;;;; License as published by the Free Software Foundation; either
+;;;; version 2.1 of the License, or (at your option) any later version.
+;;;;
+;;;; This library is distributed in the hope that it will be useful,
+;;;; but WITHOUT ANY WARRANTY; without even the implied warranty of
+;;;; MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+;;;; Lesser General Public License for more details.
+;;;;
+;;;; You should have received a copy of the GNU Lesser General Public
+;;;; License along with this library; if not, write to the Free Software
+;;;; Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301
USA
+
+(define-module (test-bytevector)
+ :use-module (test-suite lib)
+ :use-module (rnrs bytevector))
+
+;;; Some of the tests in here are examples taken from the R6RS Standard
+;;; Libraries document.
+
+
+(with-test-prefix "2.2 General Operations"
+
+ (pass-if "native-endianness"
+ (not (not (memq (native-endianness) '(big little)))))
+
+ (pass-if "make-bytevector"
+ (and (bytevector? (make-bytevector 20))
+ (bytevector? (make-bytevector 20 3))))
+
+ (pass-if "bytevector-length"
+ (= (bytevector-length (make-bytevector 20)) 20))
+
+ (pass-if "bytevector=?"
+ (and (bytevector=? (make-bytevector 20 7)
+ (make-bytevector 20 7))
+ (not (bytevector=? (make-bytevector 20 7)
+ (make-bytevector 20 0))))))
+
+
+(with-test-prefix "2.3 Operations on Bytes and Octets"
+
+ (pass-if "bytevector-{u8,s8}-ref"
+ (equal? '(-127 129 -1 255)
+ (let ((b1 (make-bytevector 16 -127))
+ (b2 (make-bytevector 16 255)))
+ (list (bytevector-s8-ref b1 0)
+ (bytevector-u8-ref b1 0)
+ (bytevector-s8-ref b2 0)
+ (bytevector-u8-ref b2 0)))))
+
+ (pass-if "bytevector-{u8,s8}-set!"
+ (equal? '(-126 130 -10 246)
+ (let ((b (make-bytevector 16 -127)))
+
+ (bytevector-s8-set! b 0 -126)
+ (bytevector-u8-set! b 1 246)
+
+ (list (bytevector-s8-ref b 0)
+ (bytevector-u8-ref b 0)
+ (bytevector-s8-ref b 1)
+ (bytevector-u8-ref b 1)))))
+
+ (pass-if "bytevector->u8-list"
+ (let ((lst '(1 2 3 128 150 255)))
+ (equal? lst
+ (bytevector->u8-list
+ (let ((b (make-bytevector 6)))
+ (for-each (lambda (i v)
+ (bytevector-u8-set! b i v))
+ (iota 6)
+ lst)
+ b)))))
+
+ (pass-if "u8-list->bytevector"
+ (let ((lst '(1 2 3 128 150 255)))
+ (equal? lst
+ (bytevector->u8-list (u8-list->bytevector lst)))))
+
+ (pass-if "bytevector-uint-{ref,set!} [small]"
+ (let ((b (make-bytevector 15)))
+ (bytevector-uint-set! b 0 #x1234
+ (endianness little) 2)
+ (equal? (bytevector-uint-ref b 0 (endianness big) 2)
+ #x3412)))
+
+ (pass-if "bytevector-uint-set! [large]"
+ (let ((b (make-bytevector 16)))
+ (bytevector-uint-set! b 0 (- (expt 2 128) 3)
+ (endianness little) 16)
+ (equal? (bytevector->u8-list b)
+ '(253 255 255 255 255 255 255 255
+ 255 255 255 255 255 255 255 255))))
+
+ (pass-if "bytevector-uint-{ref,set!} [large]"
+ (let ((b (make-bytevector 120)))
+ (bytevector-uint-set! b 0 (- (expt 2 128) 3)
+ (endianness little) 16)
+ (equal? (bytevector-uint-ref b 0 (endianness little) 16)
+ #xfffffffffffffffffffffffffffffffd)))
+
+ (pass-if "bytevector-sint-ref [small]"
+ (let ((b (u8-list->bytevector '(#xff #xf0 #xff))))
+ (equal? (bytevector-sint-ref b 0 (endianness big) 2)
+ (bytevector-sint-ref b 1 (endianness little) 2)
+ -16)))
+
+ (pass-if "bytevector-sint-ref [large]"
+ (let ((b (make-bytevector 50)))
+ (bytevector-uint-set! b 0 (- (expt 2 128) 3)
+ (endianness little) 16)
+ (equal? (bytevector-sint-ref b 0 (endianness little) 16)
+ -3)))
+
+ (pass-if "bytevector-sint-set! [small]"
+ (let ((b (make-bytevector 3)))
+ (bytevector-sint-set! b 0 -16 (endianness big) 2)
+ (bytevector-sint-set! b 1 -16 (endianness little) 2)
+ (equal? (bytevector->u8-list b)
+ '(#xff #xf0 #xff)))))
+
+
+(with-test-prefix "2.4 Operations on Integers of Arbitrary Size"
+
+ (pass-if "bytevector->sint-list"
+ (let ((b (u8-list->bytevector '(1 2 3 255 1 2 1 2))))
+ (equal? (bytevector->sint-list b (endianness little) 2)
+ '(513 -253 513 513))))
+
+ (pass-if "bytevector->uint-list"
+ (let ((b (u8-list->bytevector '(2 1 255 3 2 1 2 1))))
+ (equal? (bytevector->uint-list b (endianness big) 2)
+ '(513 65283 513 513))))
+
+ (pass-if "bytevector->uint-list [empty]"
+ (let ((b (make-bytevector 0)))
+ (null? (bytevector->uint-list b (endianness big) 2))))
+
+ (pass-if-exception "bytevector->sint-list [out-of-range]"
+ exception:out-of-range
+ (bytevector->sint-list (make-bytevector 6) (endianness little) 8))
+
+ (pass-if "bytevector->sint-list [off-by-one]"
+ (equal? (bytevector->sint-list (make-bytevector 31 #xff)
+ (endianness little) 8)
+ '(-1 -1 -1)))
+
+ (pass-if "{sint,uint}-list->bytevector"
+ (let ((b1 (sint-list->bytevector '(513 -253 513 513)
+ (endianness little) 2))
+ (b2 (uint-list->bytevector '(513 65283 513 513)
+ (endianness little) 2))
+ (b3 (u8-list->bytevector '(1 2 3 255 1 2 1 2))))
+ (and (bytevector=? b1 b2)
+ (bytevector=? b2 b3))))
+
+ (pass-if "sint-list->bytevector [limits]"
+ (bytevector=? (sint-list->bytevector '(-32768 32767)
+ (endianness big) 2)
+ (let ((bv (make-bytevector 4)))
+ (bytevector-u8-set! bv 0 #x80)
+ (bytevector-u8-set! bv 1 #x00)
+ (bytevector-u8-set! bv 2 #x7f)
+ (bytevector-u8-set! bv 3 #xff)
+ bv)))
+
+ (pass-if-exception "sint-list->bytevector [out-of-range]"
+ exception:out-of-range
+ (sint-list->bytevector (list 0 0 (expt 2 16)) (endianness big)
+ 2))
+
+ (pass-if-exception "uint-list->bytevector [out-of-range]"
+ exception:out-of-range
+ (uint-list->bytevector '(0 -1) (endianness big) 2)))
+
+
+(with-test-prefix "2.5 Operations on 16-Bit Integers"
+
+ (pass-if "bytevector-u16-ref"
+ (let ((b (u8-list->bytevector
+ '(255 255 255 255 255 255 255 255
+ 255 255 255 255 255 255 255 253))))
+ (and (equal? (bytevector-u16-ref b 14 (endianness little))
+ #xfdff)
+ (equal? (bytevector-u16-ref b 14 (endianness big))
+ #xfffd))))
+
+ (pass-if "bytevector-s16-ref"
+ (let ((b (u8-list->bytevector
+ '(255 255 255 255 255 255 255 255
+ 255 255 255 255 255 255 255 253))))
+ (and (equal? (bytevector-s16-ref b 14 (endianness little))
+ -513)
+ (equal? (bytevector-s16-ref b 14 (endianness big))
+ -3))))
+
+ (pass-if "bytevector-s16-ref [unaligned]"
+ (let ((b (u8-list->bytevector '(#xff #xf0 #xff))))
+ (equal? (bytevector-s16-ref b 1 (endianness little))
+ -16)))
+
+ (pass-if "bytevector-{u16,s16}-ref"
+ (let ((b (make-bytevector 2)))
+ (bytevector-u16-set! b 0 44444 (endianness little))
+ (and (equal? (bytevector-u16-ref b 0 (endianness little))
+ 44444)
+ (equal? (bytevector-s16-ref b 0 (endianness little))
+ (- 44444 65536)))))
+
+ (pass-if "bytevector-native-{u16,s16}-{ref,set!}"
+ (let ((b (make-bytevector 2)))
+ (bytevector-u16-native-set! b 0 44444)
+ (and (equal? (bytevector-u16-native-ref b 0)
+ 44444)
+ (equal? (bytevector-s16-native-ref b 0)
+ (- 44444 65536)))))
+
+ (pass-if "bytevector-s16-{ref,set!} [unaligned]"
+ (let ((b (make-bytevector 3)))
+ (bytevector-s16-set! b 1 -77 (endianness little))
+ (equal? (bytevector-s16-ref b 1 (endianness little))
+ -77))))
+
+
+(with-test-prefix "2.6 Operations on 32-bit Integers"
+
+ (pass-if "bytevector-u32-ref"
+ (let ((b (u8-list->bytevector
+ '(255 255 255 255 255 255 255 255
+ 255 255 255 255 255 255 255 253))))
+ (and (equal? (bytevector-u32-ref b 12 (endianness little))
+ #xfdffffff)
+ (equal? (bytevector-u32-ref b 12 (endianness big))
+ #xfffffffd))))
+
+ (pass-if "bytevector-s32-ref"
+ (let ((b (u8-list->bytevector
+ '(255 255 255 255 255 255 255 255
+ 255 255 255 255 255 255 255 253))))
+ (and (equal? (bytevector-s32-ref b 12 (endianness little))
+ -33554433)
+ (equal? (bytevector-s32-ref b 12 (endianness big))
+ -3))))
+
+ (pass-if "bytevector-{u32,s32}-ref"
+ (let ((b (make-bytevector 4)))
+ (bytevector-u32-set! b 0 2222222222 (endianness little))
+ (and (equal? (bytevector-u32-ref b 0 (endianness little))
+ 2222222222)
+ (equal? (bytevector-s32-ref b 0 (endianness little))
+ (- 2222222222 (expt 2 32))))))
+
+ (pass-if "bytevector-{u32,s32}-native-{ref,set!}"
+ (let ((b (make-bytevector 4)))
+ (bytevector-u32-native-set! b 0 2222222222)
+ (and (equal? (bytevector-u32-native-ref b 0)
+ 2222222222)
+ (equal? (bytevector-s32-native-ref b 0)
+ (- 2222222222 (expt 2 32)))))))
+
+
+(with-test-prefix "2.7 Operations on 64-bit Integers"
+
+ (pass-if "bytevector-u64-ref"
+ (let ((b (u8-list->bytevector
+ '(255 255 255 255 255 255 255 255
+ 255 255 255 255 255 255 255 253))))
+ (and (equal? (bytevector-u64-ref b 8 (endianness little))
+ #xfdffffffffffffff)
+ (equal? (bytevector-u64-ref b 8 (endianness big))
+ #xfffffffffffffffd))))
+
+ (pass-if "bytevector-s64-ref"
+ (let ((b (u8-list->bytevector
+ '(255 255 255 255 255 255 255 255
+ 255 255 255 255 255 255 255 253))))
+ (and (equal? (bytevector-s64-ref b 8 (endianness little))
+ -144115188075855873)
+ (equal? (bytevector-s64-ref b 8 (endianness big))
+ -3))))
+
+ (pass-if "bytevector-{u64,s64}-ref"
+ (let ((b (make-bytevector 8))
+ (big 9333333333333333333))
+ (bytevector-u64-set! b 0 big (endianness little))
+ (and (equal? (bytevector-u64-ref b 0 (endianness little))
+ big)
+ (equal? (bytevector-s64-ref b 0 (endianness little))
+ (- big (expt 2 64))))))
+
+ (pass-if "bytevector-{u64,s64}-native-{ref,set!}"
+ (let ((b (make-bytevector 8))
+ (big 9333333333333333333))
+ (bytevector-u64-native-set! b 0 big)
+ (and (equal? (bytevector-u64-native-ref b 0)
+ big)
+ (equal? (bytevector-s64-native-ref b 0)
+ (- big (expt 2 64))))))
+
+ (pass-if "ref/set! with zero"
+ (let ((b (make-bytevector 8)))
+ (bytevector-s64-set! b 0 -1 (endianness big))
+ (bytevector-u64-set! b 0 0 (endianness big))
+ (= 0 (bytevector-u64-ref b 0 (endianness big))))))
+
+
+(with-test-prefix "2.8 Operations on IEEE-754 Representations"
+
+ (pass-if "bytevector-ieee-single-native-{ref,set!}"
+ (let ((b (make-bytevector 4))
+ (number 3.00))
+ (bytevector-ieee-single-native-set! b 0 number)
+ (equal? (bytevector-ieee-single-native-ref b 0)
+ number)))
+
+ (pass-if "bytevector-ieee-single-{ref,set!}"
+ (let ((b (make-bytevector 8))
+ (number 3.14))
+ (bytevector-ieee-single-set! b 0 number (endianness little))
+ (bytevector-ieee-single-set! b 4 number (endianness big))
+ (equal? (bytevector-ieee-single-ref b 0 (endianness little))
+ (bytevector-ieee-single-ref b 4 (endianness big)))))
+
+ (pass-if "bytevector-ieee-single-{ref,set!} [unaligned]"
+ (let ((b (make-bytevector 9))
+ (number 3.14))
+ (bytevector-ieee-single-set! b 1 number (endianness little))
+ (bytevector-ieee-single-set! b 5 number (endianness big))
+ (equal? (bytevector-ieee-single-ref b 1 (endianness little))
+ (bytevector-ieee-single-ref b 5 (endianness big)))))
+
+ (pass-if "bytevector-ieee-double-native-{ref,set!}"
+ (let ((b (make-bytevector 8))
+ (number 3.14))
+ (bytevector-ieee-double-native-set! b 0 number)
+ (equal? (bytevector-ieee-double-native-ref b 0)
+ number)))
+
+ (pass-if "bytevector-ieee-double-{ref,set!}"
+ (let ((b (make-bytevector 16))
+ (number 3.14))
+ (bytevector-ieee-double-set! b 0 number (endianness little))
+ (bytevector-ieee-double-set! b 8 number (endianness big))
+ (equal? (bytevector-ieee-double-ref b 0 (endianness little))
+ (bytevector-ieee-double-ref b 8 (endianness big))))))
+
+
+(define (with-locale locale thunk)
+ ;; Run THUNK under LOCALE.
+ (let ((original-locale (setlocale LC_ALL)))
+ (catch 'system-error
+ (lambda ()
+ (setlocale LC_ALL locale))
+ (lambda (key . args)
+ (throw 'unresolved)))
+
+ (dynamic-wind
+ (lambda ()
+ #t)
+ thunk
+ (lambda ()
+ (setlocale LC_ALL original-locale)))))
+
+(define (with-latin1-locale thunk)
+ ;; Try out several ISO-8859-1 locales and run THUNK under the one that
+ ;; works (if any).
+ (define %locales
+ (map (lambda (name)
+ (string-append name ".ISO-8859-1"))
+ '("fr_FR" "es_ES" "en_GB" "en_US" "de_DE" "pt_PT")))
+
+ (let loop ((locales %locales))
+ (if (null? locales)
+ (throw 'unresolved)
+ (catch 'unresolved
+ (lambda ()
+ (with-locale (car locales) thunk))
+ (lambda (key . args)
+ (loop (cdr locales)))))))
+
+
+;; Default to the C locale for the following tests.
+(setlocale LC_ALL "C")
+
+
+(with-test-prefix "2.9 Operations on Strings"
+
+ (pass-if "string->utf8"
+ (let* ((str "hello, world")
+ (utf8 (string->utf8 str)))
+ (and (bytevector? utf8)
+ (= (bytevector-length utf8)
+ (string-length str))
+ (equal? (string->list str)
+ (map integer->char (bytevector->u8-list utf8))))))
+
+ (pass-if "string->utf8 [latin-1]"
+ (with-latin1-locale
+ (lambda ()
+ (let* ((str "hé, ça va bien ?")
+ (utf8 (string->utf8 str)))
+ (and (bytevector? utf8)
+ (= (bytevector-length utf8)
+ (+ 2 (string-length str))))))))
+
+ (pass-if "string->utf16"
+ (let* ((str "hello, world")
+ (utf16 (string->utf16 str)))
+ (and (bytevector? utf16)
+ (= (bytevector-length utf16)
+ (* 2 (string-length str)))
+ (equal? (string->list str)
+ (map integer->char
+ (bytevector->uint-list utf16
+ (endianness big) 2))))))
+
+ (pass-if "string->utf16 [little]"
+ (let* ((str "hello, world")
+ (utf16 (string->utf16 str (endianness little))))
+ (and (bytevector? utf16)
+ (= (bytevector-length utf16)
+ (* 2 (string-length str)))
+ (equal? (string->list str)
+ (map integer->char
+ (bytevector->uint-list utf16
+ (endianness little) 2))))))
+
+
+ (pass-if "string->utf32"
+ (let* ((str "hello, world")
+ (utf32 (string->utf32 str)))
+ (and (bytevector? utf32)
+ (= (bytevector-length utf32)
+ (* 4 (string-length str)))
+ (equal? (string->list str)
+ (map integer->char
+ (bytevector->uint-list utf32
+ (endianness big) 4))))))
+
+ (pass-if "string->utf32 [little]"
+ (let* ((str "hello, world")
+ (utf32 (string->utf32 str (endianness little))))
+ (and (bytevector? utf32)
+ (= (bytevector-length utf32)
+ (* 4 (string-length str)))
+ (equal? (string->list str)
+ (map integer->char
+ (bytevector->uint-list utf32
+ (endianness little) 4))))))
+
+ (pass-if "utf8->string"
+ (let* ((utf8 (u8-list->bytevector (map char->integer
+ (string->list "hello, world"))))
+ (str (utf8->string utf8)))
+ (and (string? str)
+ (= (string-length str)
+ (bytevector-length utf8))
+ (equal? (string->list str)
+ (map integer->char (bytevector->u8-list utf8))))))
+
+ (pass-if "utf8->string [latin-1]"
+ (with-latin1-locale
+ (lambda ()
+ (let* ((utf8 (string->utf8 "hé, ça va bien ?"))
+ (str (utf8->string utf8)))
+ (and (string? str)
+ (= (string-length str)
+ (- (bytevector-length utf8) 2)))))))
+
+ (pass-if "utf16->string"
+ (let* ((utf16 (uint-list->bytevector (map char->integer
+ (string->list "hello, world"))
+ (endianness big) 2))
+ (str (utf16->string utf16)))
+ (and (string? str)
+ (= (* 2 (string-length str))
+ (bytevector-length utf16))
+ (equal? (string->list str)
+ (map integer->char
+ (bytevector->uint-list utf16 (endianness big)
+ 2))))))
+
+ (pass-if "utf16->string [little]"
+ (let* ((utf16 (uint-list->bytevector (map char->integer
+ (string->list "hello, world"))
+ (endianness little) 2))
+ (str (utf16->string utf16 (endianness little))))
+ (and (string? str)
+ (= (* 2 (string-length str))
+ (bytevector-length utf16))
+ (equal? (string->list str)
+ (map integer->char
+ (bytevector->uint-list utf16 (endianness little)
+ 2))))))
+ (pass-if "utf32->string"
+ (let* ((utf32 (uint-list->bytevector (map char->integer
+ (string->list "hello, world"))
+ (endianness big) 4))
+ (str (utf32->string utf32)))
+ (and (string? str)
+ (= (* 4 (string-length str))
+ (bytevector-length utf32))
+ (equal? (string->list str)
+ (map integer->char
+ (bytevector->uint-list utf32 (endianness big)
+ 4))))))
+
+ (pass-if "utf32->string [little]"
+ (let* ((utf32 (uint-list->bytevector (map char->integer
+ (string->list "hello, world"))
+ (endianness little) 4))
+ (str (utf32->string utf32 (endianness little))))
+ (and (string? str)
+ (= (* 4 (string-length str))
+ (bytevector-length utf32))
+ (equal? (string->list str)
+ (map integer->char
+ (bytevector->uint-list utf32 (endianness little)
+ 4)))))))
+
+
+;;; Local Variables:
+;;; coding: latin-1
+;;; mode: scheme
+;;; End:
diff --git a/test-suite/tests/r6rs-ports.test b/test-suite/tests/r6rs-ports.test
new file mode 100644
index 0000000..204f371
--- /dev/null
+++ b/test-suite/tests/r6rs-ports.test
@@ -0,0 +1,455 @@
+;;;; r6rs-ports.test --- Exercise the R6RS I/O port API.
+;;;;
+;;;; Copyright (C) 2009 Free Software Foundation, Inc.
+;;;; Ludovic Courtès
+;;;;
+;;;; This library is free software; you can redistribute it and/or
+;;;; modify it under the terms of the GNU Lesser General Public
+;;;; License as published by the Free Software Foundation; either
+;;;; version 2.1 of the License, or (at your option) any later version.
+;;;;
+;;;; This library is distributed in the hope that it will be useful,
+;;;; but WITHOUT ANY WARRANTY; without even the implied warranty of
+;;;; MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+;;;; Lesser General Public License for more details.
+;;;;
+;;;; You should have received a copy of the GNU Lesser General Public
+;;;; License along with this library; if not, write to the Free Software
+;;;; Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301
USA
+
+(define-module (test-io-ports)
+ :use-module (test-suite lib)
+ :use-module (srfi srfi-1)
+ :use-module (srfi srfi-11)
+ :use-module (rnrs io ports)
+ :use-module (rnrs bytevector))
+
+;;; All these tests assume Guile 1.8's port system, where characters are
+;;; treated as octets.
+
+
+(with-test-prefix "7.2.5 End-of-File Object"
+
+ (pass-if "eof-object"
+ (and (eqv? (eof-object) (eof-object))
+ (eq? (eof-object) (eof-object)))))
+
+
+(with-test-prefix "7.2.8 Binary Input"
+
+ (pass-if "get-u8"
+ (let ((port (open-input-string "A")))
+ (and (= (char->integer #\A) (get-u8 port))
+ (eof-object? (get-u8 port)))))
+
+ (pass-if "lookahead-u8"
+ (let ((port (open-input-string "A")))
+ (and (= (char->integer #\A) (lookahead-u8 port))
+ (not (eof-object? port))
+ (= (char->integer #\A) (get-u8 port))
+ (eof-object? (get-u8 port)))))
+
+ (pass-if "get-bytevector-n [short]"
+ (let* ((port (open-input-string "GNU Guile"))
+ (bv (get-bytevector-n port 4)))
+ (and (bytevector? bv)
+ (equal? (bytevector->u8-list bv)
+ (map char->integer (string->list "GNU "))))))
+
+ (pass-if "get-bytevector-n [long]"
+ (let* ((port (open-input-string "GNU Guile"))
+ (bv (get-bytevector-n port 256)))
+ (and (bytevector? bv)
+ (equal? (bytevector->u8-list bv)
+ (map char->integer (string->list "GNU Guile"))))))
+
+ (pass-if-exception "get-bytevector-n with closed port"
+ exception:wrong-type-arg
+
+ (let ((port (%make-void-port "r")))
+
+ (close-port port)
+ (get-bytevector-n port 3)))
+
+ (pass-if "get-bytevector-n! [short]"
+ (let* ((port (open-input-string "GNU Guile"))
+ (bv (make-bytevector 4))
+ (read (get-bytevector-n! port bv 0 4)))
+ (and (equal? read 4)
+ (equal? (bytevector->u8-list bv)
+ (map char->integer (string->list "GNU "))))))
+
+ (pass-if "get-bytevector-n! [long]"
+ (let* ((str "GNU Guile")
+ (port (open-input-string str))
+ (bv (make-bytevector 256))
+ (read (get-bytevector-n! port bv 0 256)))
+ (and (equal? read (string-length str))
+ (equal? (map (lambda (i)
+ (bytevector-u8-ref bv i))
+ (iota read))
+ (map char->integer (string->list str))))))
+
+ (pass-if "get-bytevector-some [simple]"
+ (let* ((str "GNU Guile")
+ (port (open-input-string str))
+ (bv (get-bytevector-some port)))
+ (and (bytevector? bv)
+ (equal? (bytevector->u8-list bv)
+ (map char->integer (string->list str))))))
+
+ (pass-if "get-bytevector-some [only-some]"
+ (let* ((str "GNU Guile")
+ (index 0)
+ (port (make-soft-port
+ (vector #f #f #f
+ (lambda ()
+ (if (>= index (string-length str))
+ (eof-object)
+ (let ((c (string-ref str index)))
+ (set! index (+ index 1))
+ c)))
+ (lambda () #t)
+ (lambda ()
+ ;; Number of readily available octets: falls to
+ ;; zero after 4 octets have been read.
+ (- 4 (modulo index 5))))
+ "r"))
+ (bv (get-bytevector-some port)))
+ (and (bytevector? bv)
+ (= index 4)
+ (= (bytevector-length bv) index)
+ (equal? (bytevector->u8-list bv)
+ (map char->integer (string->list "GNU "))))))
+
+ (pass-if "get-bytevector-all"
+ (let* ((str "GNU Guile")
+ (index 0)
+ (port (make-soft-port
+ (vector #f #f #f
+ (lambda ()
+ (if (>= index (string-length str))
+ (eof-object)
+ (let ((c (string-ref str index)))
+ (set! index (+ index 1))
+ c)))
+ (lambda () #t)
+ (let ((cont? #f))
+ (lambda ()
+ ;; Number of readily available octets: falls to
+ ;; zero after 4 octets have been read and then
+ ;; starts again.
+ (let ((a (if cont?
+ (- (string-length str) index)
+ (- 4 (modulo index 5)))))
+ (if (= 0 a) (set! cont? #t))
+ a))))
+ "r"))
+ (bv (get-bytevector-all port)))
+ (and (bytevector? bv)
+ (= index (string-length str))
+ (= (bytevector-length bv) (string-length str))
+ (equal? (bytevector->u8-list bv)
+ (map char->integer (string->list str)))))))
+
+
+(define (make-soft-output-port)
+ (let* ((bv (make-bytevector 1024))
+ (read-index 0)
+ (write-index 0)
+ (write-char (lambda (chr)
+ (bytevector-u8-set! bv write-index
+ (char->integer chr))
+ (set! write-index (+ 1 write-index)))))
+ (make-soft-port
+ (vector write-char
+ (lambda (str) ;; write-string
+ (for-each write-char (string->list str)))
+ (lambda () #t) ;; flush-output
+ (lambda () ;; read-char
+ (if (>= read-index (bytevector-length bv))
+ (eof-object)
+ (let ((c (bytevector-u8-ref bv read-index)))
+ (set! read-index (+ read-index 1))
+ (integer->char c))))
+ (lambda () #t)) ;; close-port
+ "rw")))
+
+(with-test-prefix "7.2.11 Binary Output"
+
+ (pass-if "put-u8"
+ (let ((port (make-soft-output-port)))
+ (put-u8 port 77)
+ (equal? (get-u8 port) 77)))
+
+ (pass-if "put-bytevector [2 args]"
+ (let ((port (make-soft-output-port))
+ (bv (make-bytevector 256)))
+ (put-bytevector port bv)
+ (equal? (bytevector->u8-list bv)
+ (bytevector->u8-list
+ (get-bytevector-n port (bytevector-length bv))))))
+
+ (pass-if "put-bytevector [3 args]"
+ (let ((port (make-soft-output-port))
+ (bv (make-bytevector 256))
+ (start 10))
+ (put-bytevector port bv start)
+ (equal? (drop (bytevector->u8-list bv) start)
+ (bytevector->u8-list
+ (get-bytevector-n port (- (bytevector-length bv) start))))))
+
+ (pass-if "put-bytevector [4 args]"
+ (let ((port (make-soft-output-port))
+ (bv (make-bytevector 256))
+ (start 10)
+ (count 77))
+ (put-bytevector port bv start count)
+ (equal? (take (drop (bytevector->u8-list bv) start) count)
+ (bytevector->u8-list
+ (get-bytevector-n port count)))))
+
+ (pass-if-exception "put-bytevector with closed port"
+ exception:wrong-type-arg
+
+ (let* ((bv (make-bytevector 4))
+ (port (%make-void-port "w")))
+
+ (close-port port)
+ (put-bytevector port bv))))
+
+
+(with-test-prefix "7.2.7 Input Ports"
+
+ ;; This section appears here so that it can use the binary input
+ ;; primitives.
+
+ (pass-if "open-bytevector-input-port [1 arg]"
+ (let* ((str "Hello Port!")
+ (bv (u8-list->bytevector (map char->integer
+ (string->list str))))
+ (port (open-bytevector-input-port bv))
+ (read-to-string
+ (lambda (port)
+ (let loop ((chr (read-char port))
+ (result '()))
+ (if (eof-object? chr)
+ (apply string (reverse! result))
+ (loop (read-char port)
+ (cons chr result)))))))
+
+ (equal? (read-to-string port) str)))
+
+ (pass-if-exception "bytevector-input-port is read-only"
+ exception:wrong-type-arg
+
+ (let* ((str "Hello Port!")
+ (bv (u8-list->bytevector (map char->integer
+ (string->list str))))
+ (port (open-bytevector-input-port bv #f)))
+
+ (write "hello" port)))
+
+ (pass-if "bytevector input port supports seeking"
+ (let* ((str "Hello Port!")
+ (bv (u8-list->bytevector (map char->integer
+ (string->list str))))
+ (port (open-bytevector-input-port bv #f)))
+
+ (and (port-has-port-position? port)
+ (= 0 (port-position port))
+ (port-has-set-port-position!? port)
+ (begin
+ (set-port-position! port 6)
+ (= 6 (port-position port)))
+ (bytevector=? (get-bytevector-all port)
+ (u8-list->bytevector
+ (map char->integer (string->list "Port!")))))))
+
+ (pass-if-exception "make-custom-binary-input-port [wrong-num-args]"
+ exception:wrong-num-args
+
+ ;; Prior to Guile-R6RS-Libs 0.2, the last 3 arguments were wrongfully
+ ;; optional.
+ (make-custom-binary-input-port "port" (lambda args #t)))
+
+ (pass-if "make-custom-binary-input-port"
+ (let* ((source (make-bytevector 7777))
+ (read! (let ((pos 0)
+ (len (bytevector-length source)))
+ (lambda (bv start count)
+ (let ((amount (min count (- len pos))))
+ (if (> amount 0)
+ (bytevector-copy! source pos
+ bv start amount))
+ (set! pos (+ pos amount))
+ amount))))
+ (port (make-custom-binary-input-port "the port" read!
+ #f #f #f)))
+
+ (bytevector=? (get-bytevector-all port) source)))
+
+ (pass-if "custom binary input port does not support `port-position'"
+ (let* ((str "Hello Port!")
+ (source (open-bytevector-input-port
+ (u8-list->bytevector
+ (map char->integer (string->list str)))))
+ (read! (lambda (bv start count)
+ (let ((r (get-bytevector-n! source bv start count)))
+ (if (eof-object? r)
+ 0
+ r))))
+ (port (make-custom-binary-input-port "the port" read!
+ #f #f #f)))
+ (not (or (port-has-port-position? port)
+ (port-has-set-port-position!? port)))))
+
+ (pass-if "custom binary input port supports `port-position'"
+ (let* ((str "Hello Port!")
+ (source (open-bytevector-input-port
+ (u8-list->bytevector
+ (map char->integer (string->list str)))))
+ (read! (lambda (bv start count)
+ (let ((r (get-bytevector-n! source bv start count)))
+ (if (eof-object? r)
+ 0
+ r))))
+ (get-pos (lambda ()
+ (port-position source)))
+ (set-pos! (lambda (pos)
+ (set-port-position! source pos)))
+ (port (make-custom-binary-input-port "the port" read!
+ get-pos set-pos! #f)))
+
+ (and (port-has-port-position? port)
+ (= 0 (port-position port))
+ (port-has-set-port-position!? port)
+ (begin
+ (set-port-position! port 6)
+ (= 6 (port-position port)))
+ (bytevector=? (get-bytevector-all port)
+ (u8-list->bytevector
+ (map char->integer (string->list "Port!")))))))
+
+ (pass-if "custom binary input port `close-proc' is called"
+ (let* ((closed? #f)
+ (read! (lambda (bv start count) 0))
+ (get-pos (lambda () 0))
+ (set-pos! (lambda (pos) #f))
+ (close! (lambda () (set! closed? #t)))
+ (port (make-custom-binary-input-port "the port" read!
+ get-pos set-pos!
+ close!)))
+
+ (close-port port)
+ closed?)))
+
+
+(with-test-prefix "8.2.10 Output ports"
+
+ (pass-if "open-bytevector-output-port"
+ (let-values (((port get-content)
+ (open-bytevector-output-port #f)))
+ (let ((source (make-bytevector 7777)))
+ (put-bytevector port source)
+ (and (bytevector=? (get-content) source)
+ (bytevector=? (get-content) (make-bytevector 0))))))
+
+ (pass-if "open-bytevector-output-port [put-u8]"
+ (let-values (((port get-content)
+ (open-bytevector-output-port)))
+ (put-u8 port 77)
+ (and (bytevector=? (get-content) (make-bytevector 1 77))
+ (bytevector=? (get-content) (make-bytevector 0)))))
+
+ (pass-if "open-bytevector-output-port [display]"
+ (let-values (((port get-content)
+ (open-bytevector-output-port)))
+ (display "hello" port)
+ (and (bytevector=? (get-content) (string->utf8 "hello"))
+ (bytevector=? (get-content) (make-bytevector 0)))))
+
+ (pass-if "bytevector output port supports `port-position'"
+ (let-values (((port get-content)
+ (open-bytevector-output-port)))
+ (let ((source (make-bytevector 7777))
+ (overwrite (make-bytevector 33)))
+ (and (port-has-port-position? port)
+ (port-has-set-port-position!? port)
+ (begin
+ (put-bytevector port source)
+ (= (bytevector-length source)
+ (port-position port)))
+ (begin
+ (set-port-position! port 10)
+ (= 10 (port-position port)))
+ (begin
+ (put-bytevector port overwrite)
+ (bytevector-copy! overwrite 0 source 10
+ (bytevector-length overwrite))
+ (= (port-position port)
+ (+ 10 (bytevector-length overwrite))))
+ (bytevector=? (get-content) source)
+ (bytevector=? (get-content) (make-bytevector 0))))))
+
+ (pass-if "make-custom-binary-output"
+ (let ((port (make-custom-binary-output-port "cbop"
+ (lambda (x y z) 0)
+ #f #f #f)))
+ (and (output-port? port)
+ (binary-port? port)
+ (not (port-has-port-position? port))
+ (not (port-has-set-port-position!? port)))))
+
+ (pass-if "make-custom-binary-output-port [partial writes]"
+ (let* ((source (uint-list->bytevector (iota 333)
+ (native-endianness) 2))
+ (sink (make-bytevector (bytevector-length source)))
+ (sink-pos 0)
+ (eof? #f)
+ (write! (lambda (bv start count)
+ (if (= 0 count)
+ (begin
+ (set! eof? #t)
+ 0)
+ (let ((u8 (bytevector-u8-ref bv start)))
+ ;; Get one byte at a time.
+ (bytevector-u8-set! sink sink-pos u8)
+ (set! sink-pos (+ 1 sink-pos))
+ 1))))
+ (port (make-custom-binary-output-port "cbop" write!
+ #f #f #f)))
+ (put-bytevector port source)
+ (and (= sink-pos (bytevector-length source))
+ (not eof?)
+ (bytevector=? sink source))))
+
+ (pass-if "make-custom-binary-output-port [full writes]"
+ (let* ((source (uint-list->bytevector (iota 333)
+ (native-endianness) 2))
+ (sink (make-bytevector (bytevector-length source)))
+ (sink-pos 0)
+ (eof? #f)
+ (write! (lambda (bv start count)
+ (if (= 0 count)
+ (begin
+ (set! eof? #t)
+ 0)
+ (begin
+ (bytevector-copy! bv start
+ sink sink-pos
+ count)
+ (set! sink-pos (+ sink-pos count))
+ count))))
+ (port (make-custom-binary-output-port "cbop" write!
+ #f #f #f)))
+ (put-bytevector port source)
+ (and (= sink-pos (bytevector-length source))
+ (not eof?)
+ (bytevector=? sink source)))))
+
+
+;;; Local Variables:
+;;; coding: latin-1
+;;; mode: scheme
+;;; End:
hooks/post-receive
--
GNU Guile
| [Prev in Thread] |
Current Thread |
[Next in Thread] |
- [Guile-commits] GNU Guile branch, master, updated. 1ee2c72eafaae5f91f4c899bc4b4853af5c16f28,
Ludovic Courtès <=