bug#26938: [PATCH] gnu: Add catdoc.

[Hartmut Goebel]

* gnu/packages/textutils.scm (catdoc): New variable.
---
 gnu/packages/textutils.scm | 39 +++++++++++++++++++++++++++++++++++++++
 1 file changed, 39 insertions(+)

diff --git a/gnu/packages/textutils.scm b/gnu/packages/textutils.scm
index dbd71c2..7945df3 100644
--- a/gnu/packages/textutils.scm
+++ b/gnu/packages/textutils.scm
@@ -10,6 +10,7 @@
 ;;; Copyright © 2016 Marius Bakke <address@hidden>
 ;;; Copyright © 2017 Eric Bavier <address@hidden>
 ;;; Copyright © 2017 Rene Saavedra <address@hidden>
+;;; Copyright © 2017 Hartmut Goebel <address@hidden>
 ;;;
 ;;; This file is part of GNU Guix.
 ;;;
@@ -368,6 +369,44 @@ to everybody, because they believe that everybody runs 
Windows and therefore
 runs Word\".")
     (license license:gpl2+)))
 
+(define-public catdoc
+  (package
+    (name "catdoc")
+    (version "0.95")
+    (source (origin
+              (method url-fetch)
+              (uri (string-append "http://ftp.wagner.pp.ru/pub/catdoc/";
+                                  "catdoc-" version ".tar.gz"))
+              (sha256
+               (base32
+                "15h7v3bmwfk4z8r78xs5ih6vd0pskn0rj90xghvbzdjj0cc88jji"))))
+    (build-system gnu-build-system)
+    ;; TODO: Also build `wordview` which requires `tk` – make a separate
+    ;; package for this.
+    (arguments
+     '(#:tests? #f ; There are no tests
+       #:configure-flags '("--disable-wordview")
+       #:phases
+       (modify-phases %standard-phases
+         (add-before 'install 'fix-install
+           (lambda* (#:key outputs #:allow-other-keys)
+             (let ((out (assoc-ref outputs "out")))
+               (mkdir-p (string-append out "/share/man/man1"))))))))
+    (home-page "http://www.wagner.pp.ru/~vitus/software/catdoc/";)
+    (synopsis "MS-Word to TeX or plain text converter")
+    (description "@command{catdoc} extracts text from MS-Word files, trying to
+preserve as many special printable characters as possible.  catdoc supports
+everything up to Word-97. Also supported are MS Write documents and RTF files.
+
+It doesn't even try to preserve fancy Word formatting, because Word users
+usually don't care about document structure, and it is this very thing which
+is important to LaTeX users.
+
+This package also provides @command{xls2csv}, which extracts data from Excel
+spreadsheets and outputs it in comma-separated-value format, and
address@hidden, which extracts data from PowerPoint presentations.")
+    (license license:gpl2+)))
+
 (define-public utfcpp
   (package
     (name "utfcpp")
-- 
2.7.4