Re: More New Functions: findfigs.m, isfloat.m, genvarname.m

[Bill Denney]

Robert Platt wrote:
> Sorry for not replying sooner, I was away on a course.
>
> The version submitted by Bill assumes that the input strings are already
> valid. e.g.:
>
> genvarname('4example', '4example')
> ==> 4example1 --- invalid variable name
>
> compared this to '_4example', which is generated by the other version.
>
> Guaranteeing the variable names generated is an important feature of the
> Matlab genvarname function. Are you happy to submit the other version of
> genvarname, Bill?
>
> Rob
Hi Rob and John,

Rob, your version of genvarname.m will be significantly slower due to 
the use of sub-functions and regexp.  I incorporated all the features of 
your version (and your tests and documentation) into the version I 
created.  What do you think of the attached version?

Have a good day,

Bill

# HG changeset patch
# User address@hidden
# Date 1205589149 14400
# Node ID 5e794cfcbde1a74db488c5d266c67a2fbb09dea2
# Parent  bf704d1c5e4375253408bbb855f99138cee2bb5c
genvarname.m: new function

diff -r bf704d1c5e43 -r 5e794cfcbde1 scripts/ChangeLog
--- a/scripts/ChangeLog Fri Mar 14 08:00:50 2008 -0400
+++ b/scripts/ChangeLog Sat Mar 15 09:52:29 2008 -0400
@@ -1,3 +1,7 @@ 2008-03-14  Kai Habel  <address@hidden
+2008-03-15  Bill Denney  <address@hidden>
+
+       * general/genvarname.m: new function
+
 2008-03-14  Kai Habel  <address@hidden>
 
         * plot/__go_draw_axes__.m: Expicitly set gnuplot user
diff -r bf704d1c5e43 -r 5e794cfcbde1 scripts/general/Makefile.in
--- a/scripts/general/Makefile.in       Fri Mar 14 08:00:50 2008 -0400
+++ b/scripts/general/Makefile.in       Sat Mar 15 09:52:29 2008 -0400
@@ -33,17 +33,18 @@ INSTALL_PROGRAM = @INSTALL_PROGRAM@
 INSTALL_PROGRAM = @INSTALL_PROGRAM@
 INSTALL_DATA = @INSTALL_DATA@
 
-SOURCES = __isequal__.m __splinen__.m accumarray.m arrayfun.m bicubic.m \
-  bitcmp.m bitget.m bitset.m blkdiag.m cart2pol.m cart2sph.m cell2mat.m \
-  celldisp.m circshift.m common_size.m cplxpair.m cumtrapz.m deal.m del2.m \
-  diff.m flipdim.m fliplr.m flipud.m gradient.m ind2sub.m int2str.m interp1.m \
-  interp2.m interp3.m interpn.m interpft.m is_duplicate_entry.m isa.m \
-  isdefinite.m isdir.m isequal.m isequalwithequalnans.m \
-  isscalar.m issquare.m issymmetric.m isvector.m logical.m logspace.m \
-  lookup.m mod.m nargchk.m nextpow2.m nthroot.m num2str.m perror.m \
-  pol2cart.m polyarea.m postpad.m prepad.m quadl.m randperm.m rat.m rem.m \
-  repmat.m rot90.m rotdim.m shift.m shiftdim.m sortrows.m \
-  sph2cart.m strerror.m structfun.m sub2ind.m trapz.m tril.m triu.m
+SOURCES = __isequal__.m __splinen__.m accumarray.m arrayfun.m \
+  bicubic.m bitcmp.m bitget.m bitset.m blkdiag.m cart2pol.m cart2sph.m \
+  cell2mat.m celldisp.m circshift.m common_size.m cplxpair.m \
+  cumtrapz.m deal.m del2.m diff.m flipdim.m fliplr.m flipud.m \
+  genvarname.m gradient.m ind2sub.m int2str.m interp1.m interp2.m \
+  interp3.m interpn.m interpft.m is_duplicate_entry.m isa.m \
+  isdefinite.m isdir.m isequal.m isequalwithequalnans.m isscalar.m \
+  issquare.m issymmetric.m isvector.m logical.m logspace.m lookup.m \
+  mod.m nargchk.m nextpow2.m nthroot.m num2str.m perror.m pol2cart.m \
+  polyarea.m postpad.m prepad.m quadl.m randperm.m rat.m rem.m \
+  repmat.m rot90.m rotdim.m shift.m shiftdim.m sortrows.m sph2cart.m \
+  strerror.m structfun.m sub2ind.m trapz.m tril.m triu.m
 
 DISTFILES = $(addprefix $(srcdir)/, Makefile.in $(SOURCES))
 
diff -r bf704d1c5e43 -r 5e794cfcbde1 scripts/general/genvarname.m
--- /dev/null   Thu Jan 01 00:00:00 1970 +0000
+++ b/scripts/general/genvarname.m      Sat Mar 15 09:52:29 2008 -0400
@@ -0,0 +1,209 @@
+## Copyright (C) 2008 Bill Denney, Robert Platt
+##
+## This file is part of Octave.
+##
+## Octave 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 3 of the License, or (at
+## your option) any later version.
+##
+## Octave 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 Octave; see the file COPYING.  If not, see
+## <http://www.gnu.org/licenses/>.
+
+## -*- texinfo -*-
+## @deftypefn {Function File} address@hidden =} findfigs (@var{str})
+## @deftypefnx {Function File} address@hidden =} findfigs (@var{str}, 
@var{exclusions})
+## Create unique variable(s) from @var{str}.  If @var{exclusions} is
+## given, then the variable(s) will be unique to each other and to
+## @var{exclusions} (@var{exclusions} may be either a string or a cellstr).
+##
+## If @var{str} is a cellstr, then a unique variable is created for each
+## cell in @var{str}.
+##
+## @example
+## @group
+## x = 3.141;
+## genvarname ("x", who ())
+## @result{} x1
+## @end group
+## @end example
+##
+## If @var{wanted} is a cell array, genvarname will make sure the returned
+## strings are distinct:
+##
+## @example
+## @group
+## genvarname (@{"foo", "foo"@})
+## @result{}
+## @{
+##   [1,1] = foo
+##   [1,2] = foo1
+## @}
+## @end group
+## @end example
+##
+## Note that the result is a char array/cell array of strings, not the
+## variables themselves. To define a variable, @code{eval()} can be
+## used. The following trivial example sets @code{x} to @code{42}.
+##
+## @example
+## @group
+## name = genvarname ("x");
+## eval([name " = 42"]);
+## @result{} x =  42
+## @end group
+## @end example
+##
+## Also, this can be useful for creating unique struct field names.
+##
+## @example
+## @group
+## x = struct ();
+## for i = 1:3
+##   x.(genvarname ("a", fieldnames (x))) = i;
+## endfor
+## @result{}
+## x =
+## @{
+##   a =  1
+##   a1 =  2
+##   a2 =  3
+## @}
+## @end group
+## @end example
+##
+## Since variable names may only contain letters, digits and underscores,
+## genvarname replaces any sequence of disallowed characters with
+## an underscore. Also, variables may not begin with a digit; in this
+## case an underscore is added before the variable name.
+##
+## Variable names beginning and ending with two underscores "__" are valid but
+## they are used internally by octave and should generally be avoided, 
therefore
+## genvarname will not generate such names.
+##
+## genvarname will also make sure that returned names do not clash with
+## keywords such as "for" and "if". A number will be appended if necessary.
+## Note, however, that this does @strong{not} include function names,
+## such as "sin". Such names should be included in @var{avoid} if necessary.
+## @seealso{isvarname, exist, tmpnam, eval}
+## @end deftypefn
+
+## Authors: Rob Platt <address@hidden>
+##          Bill Denney <address@hidden>
+
+function varname = genvarname (str, exclusions)
+
+  strinput = ischar (str);
+  ## Process the inputs
+  if (nargin < 2)
+    exclusions = {};
+  elseif ischar (exclusions)
+    if (rows (exclusions) != 1)
+      error ("genvarname: if more than one exclusion is given, it must be a 
cellstr")
+    endif
+    exclusions = {exclusions};
+  elseif (! iscellstr (exclusions))
+    error ("genvarname: exclusions must be a string or a cellstr")
+  endif
+  if ischar (str)
+    if (rows (str) != 1)
+      error ("genvarname: if more than one str is given, it must be a cellstr")
+    endif
+    str = {str};
+  elseif (! iscellstr (str))
+    error ("genvarname: str must be a string or a cellstr")
+  endif
+
+  validchars = ["A":"Z" "a":"z" "0":"9" "_"];
+
+  varname = cell (size (str));
+  for i = 1:numel (str)
+    ## Perform any modifications to the varname to make sure that it is
+    ## a valid variable name.
+
+    ## remove invalid characters
+    str{i}(! ismember (str{i}, validchars)) = "_";
+    ## do not use keywords
+    if iskeyword (str{i})
+      str{i} = ["_" str{i}];
+    endif
+    ## double underscores at the beginning and end are reserved variables
+    underscores = (str{i} == "_");
+    if any (underscores)
+      firstnon = find (!underscores, 1);
+      lastnon = find (!underscores, 1, "last");
+      str{i}([1:firstnon-2 lastnon+2:end]) = [];
+    endif
+    ## The variable cannot be empty
+    if isempty (str{i})
+      str{i} = "x";
+    endif
+    ## it cannot start with a number
+    if ismember (str{i}(1), "0":"9")
+      str{i} = ["_" str{i}];
+    endif
+
+    ## make sure that the variable is unique relative to other variables
+    ## and the exclusions list
+    excluded = any (strcmp (str{i}, exclusions));
+    if (excluded && ismember (str{i}(end), "0":"9"))
+      ## if it is not unique and ends with a digit, add an underscore to
+      ## make the variable name more readable ("x1_1" instead of "x11")
+      str{i}(end+1) = "_";
+    endif
+    varname(i) = str(i);
+    idx = 0;
+    while excluded
+      idx++;
+      varname{i} = sprintf("%s%d", str{i}, idx);
+      excluded = any (strcmp (varname{i}, exclusions));
+    endwhile
+    exclusions(end+1) = varname(i);
+  endfor
+
+  if strinput
+    varname = varname{1};
+  endif
+
+endfunction
+
+## Tests
+## a single argument
+%!assert(genvarname("a"), "a")
+## a single argument with a non-conflicting exception
+%!assert(genvarname("a", "b"), "a")
+## a single argument with a conflicting exception
+%!assert(genvarname("a", "a"), "a1")
+## a single argument as a cell
+%!assert(genvarname({"a"}), {"a"})
+%!assert(genvarname({"a"}, "b"), {"a"})
+%!assert(genvarname({"a"}, {"b"}), {"a"})
+%!assert(genvarname({"a"}, "a"), {"a1"})
+%!assert(genvarname({"a"}, {"a"}), {"a1"})
+## Test different arguments
+## orientation
+%!assert(genvarname({"a" "b"}), {"a" "b"})
+%!assert(genvarname({"a";"b"}), {"a";"b"})
+%!assert(genvarname({"a" "a"}), {"a" "a1"})
+%!assert(genvarname({"a" "b";"c" "d"}), {"a" "b";"c" "d"})
+%!assert(genvarname({"a" "a" "a";"a" "a" "a"}), {"a" "a2" "a4";"a1" "a3" "a5"})
+## more than one repetition
+%!assert(genvarname({"a" "a" "a"}), {"a" "a1" "a2"})
+%!assert(genvarname({"a" "a" "a"}, {"a" "a1" "a2"}), {"a3" "a4" "a5"})
+## more than one repetition not in order
+%!assert(genvarname({"a" "b" "a" "b" "a"}), {"a" "b" "a1" "b1" "a2"})
+## Variable name munging
+%!assert (genvarname ("__x__"), "_x_")
+%!assert (genvarname ("123456789"), "_123456789")
+%!assert (genvarname ("_$1__"), "_1_")
+%!assert (genvarname ("__foo__", "_foo_"), "_foo_1")
+%!assert (genvarname ("1million_and1", "_1million_and1"), "_1million_and1_1")
+%!assert (genvarname ({"", "", ""}), {"x", "x1", "x2"})
+%!assert (genvarname ("if"), "_if")
+%!assert (genvarname ({"if", "if", "if"}), {"_if", "_if1", "_if2"})