gnu/packages/textutils.scm

   1 ;;; GNU Guix --- Functional package management for GNU
   2 ;;; Copyright © 2015 Taylan Ulrich Bayırlı/Kammer <taylanbayirli@gmail.com>
   3 ;;; Copyright © 2015, 2016 Ricardo Wurmus <rekado@elephly.net>
   4 ;;; Copyright © 2015 Ben Woodcroft <donttrustben@gmail.com>
   5 ;;; Copyright © 2015 Roel Janssen <roel@gnu.org>
   6 ;;;
   7 ;;; This file is part of GNU Guix.
   8 ;;;
   9 ;;; GNU Guix is free software; you can redistribute it and/or modify it
  10 ;;; under the terms of the GNU General Public License as published by
  11 ;;; the Free Software Foundation; either version 3 of the License, or (at
  12 ;;; your option) any later version.
  13 ;;;
  14 ;;; GNU Guix is distributed in the hope that it will be useful, but
  15 ;;; WITHOUT ANY WARRANTY; without even the implied warranty of
  16 ;;; MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  17 ;;; GNU General Public License for more details.
  18 ;;;
  19 ;;; You should have received a copy of the GNU General Public License
  20 ;;; along with GNU Guix.  If not, see <http://www.gnu.org/licenses/>.
  21
  22 (define-module (gnu packages textutils)
  23   #:use-module ((guix licenses) #:prefix license:)
  24   #:use-module (guix packages)
  25   #:use-module (guix download)
  26   #:use-module (guix git-download)
  27   #:use-module (guix build-system gnu)
  28   #:use-module (guix build-system cmake)
  29   #:use-module (gnu packages autotools)
  30   #:use-module (gnu packages python))
  31
  32 (define-public recode
  33   (package
  34     (name "recode")
  35     ;; Last beta release (3.7-beta2) is from 2008; last commit from Feb 2014.
  36     ;; So we use that commit instead.
  37     (version "3.7.0.201402")
  38     (source
  39      (origin
  40        (method git-fetch)
  41        (uri (git-reference
  42              (url "https://github.com/pinard/Recode.git")
  43              (commit "2d7092a9999194fc0e9449717a8048c8d8e26c18")))
  44        (sha256
  45         (base32 "1wssv8z6g3ryrw33sksz4rjhlnhgvvdqszw1ggl4rcwks34n86zm"))
  46        (file-name (string-append name "-" version "-checkout"))))
  47     (build-system gnu-build-system)
  48     (native-inputs `(("python" ,python-2)))
  49     (arguments
  50      '(#:phases
  51        (alist-cons-before
  52         'check 'pre-check
  53         (lambda _
  54           (substitute* "tests/setup.py"
  55             (("([[:space:]]*)include_dirs=.*" all space)
  56              (string-append all space "library_dirs=['../src/.libs'],\n")))
  57           ;; The test extension 'Recode.so' lacks RUNPATH for 'librecode.so'.
  58           (setenv "LD_LIBRARY_PATH" (string-append (getcwd) "/src/.libs")))
  59         %standard-phases)))
  60     (home-page "https://github.com/pinard/Recode")
  61     (synopsis "Text encoding converter")
  62     (description "The Recode library converts files between character sets and
  63 usages.  It recognises or produces over 200 different character sets (or about
  64 300 if combined with an iconv library) and transliterates files between almost
  65 any pair.  When exact transliteration are not possible, it gets rid of
  66 offending characters or falls back on approximations.  The recode program is a
  67 handy front-end to the library.")
  68     (license license:gpl2+)))
  69
  70 (define-public enca
  71   (package
  72     (name "enca")
  73     (version "1.16")
  74     (source
  75      (origin
  76        (method url-fetch)
  77        (uri (string-append
  78              "https://github.com/nijel/enca/archive/" version ".tar.gz"))
  79        (sha256
  80         (base32 "1xik00x0yvhswsw2isnclabhv536xk1s42cf5z54gfbpbhc7ni8l"))
  81        (file-name (string-append name "-" version ".tar.gz"))))
  82     (build-system gnu-build-system)
  83     (inputs `(("recode" ,recode)))
  84
  85     ;; Both 'test-convert-64.sh' and 'test-convert-filter.sh' manipulate a
  86     ;; 'test.tmp' file, so they have to run in sequence.
  87     (arguments '(#:parallel-tests? #f))
  88
  89     (home-page "https://github.com/nijel/enca")
  90     (synopsis "Text encoding detection tool")
  91     (description "Enca (Extremely Naive Charset Analyser) consists of libenca,
  92 an encoding detection library, and enca, a command line frontend, integrating
  93 libenca and several charset conversion libraries and tools.")
  94     (license license:gpl2)))
  95
  96 (define-public utf8proc
  97   (package
  98     (name "utf8proc")
  99     (version "1.3.1")
 100     (source
 101      (origin
 102        (method url-fetch)
 103        (uri (string-append
 104              "https://github.com/JuliaLang/utf8proc/archive/v"
 105              version ".tar.gz"))
 106        (file-name (string-append name "-" version ".tar.gz"))
 107        (sha256
 108         (base32 "1k48as5kjkar4yj3dwxyll8ykj4k723ib5a6mnw1g86q3zi0zdl3"))))
 109     (build-system gnu-build-system)
 110     (arguments
 111      '(#:tests? #f ;no "check" target
 112        #:make-flags (list "CC=gcc"
 113                           (string-append "prefix=" (assoc-ref %outputs "out")))
 114        #:phases
 115        (modify-phases %standard-phases
 116          (delete 'configure))))
 117     (home-page "http://julialang.org/utf8proc/")
 118     (synopsis "C library for processing UTF-8 Unicode data")
 119     (description "utf8proc is a small C library that provides Unicode
 120 normalization, case-folding, and other operations for data in the UTF-8
 121 encoding, supporting Unicode version 7.0.")
 122     (license license:expat)))
 123
 124 (define-public libgtextutils
 125   (package
 126     (name "libgtextutils")
 127     (version "0.7")
 128     (source
 129      (origin
 130        (method url-fetch)
 131        (uri (string-append
 132              "https://github.com/agordon/libgtextutils/releases/download/"
 133              version "/libgtextutils-" version ".tar.gz"))
 134        (sha256
 135         (base32 "0jiybkb2z58wa2msvllnphr4js2hvjvh988pavb3mzkgr6ihwbkr"))))
 136     (build-system gnu-build-system)
 137     (arguments
 138      '(#:phases
 139        (alist-cons-after
 140         'unpack 'autoreconf
 141         (lambda _ (zero? (system* "autoreconf" "-vif")))
 142         %standard-phases)))
 143     (native-inputs
 144      `(("autoconf" ,autoconf)
 145        ("automake" ,automake)
 146        ("libtool" ,libtool)))
 147     (home-page "https://github.com/agordon/libgtextutils")
 148     (synopsis "Gordon's text utils library")
 149     (description
 150      "libgtextutils is a text utilities library used by the fastx toolkit from
 151 the Hannon Lab.")
 152     (license license:agpl3+)))
 153
 154 (define-public cityhash
 155   (let ((commit "8af9b8c")
 156         (revision "1"))
 157     (package
 158       (name "cityhash")
 159       (version (string-append "1.1." revision "." commit))
 160       (source (origin
 161                 (method git-fetch)
 162                 (uri (git-reference
 163                       (url "https://github.com/google/cityhash.git")
 164                       (commit commit)))
 165                 (file-name (string-append name "-" version ".tar.gz"))
 166                 (sha256
 167                  (base32
 168                   "0n6skf5dv8yfl1ckax8dqhvsbslkwc9158zf2ims0xqdvzsahbi6"))))
 169     (build-system gnu-build-system)
 170     (home-page "https://github.com/google/cityhash")
 171     (synopsis "C++ hash functions for strings")
 172     (description
 173      "CityHash provides hash functions for strings.  The functions mix the
 174 input bits thoroughly but are not suitable for cryptography.")
 175     (license license:expat))))
 176
 177 (define-public libconfig
 178   (package
 179     (name "libconfig")
 180     (version "1.5")
 181     (source (origin
 182               (method url-fetch)
 183               (uri (string-append "http://www.hyperrealm.com/libconfig/"
 184                                   "libconfig-" version ".tar.gz"))
 185               (sha256
 186                (base32
 187                 "1xh3hzk63v4y8815lc5209m3s6ms2cpgw4h5hg462i4f1lwsl7g3"))))
 188     (build-system gnu-build-system)
 189     (home-page "http://www.hyperrealm.com/libconfig/")
 190     (synopsis "C/C++ configuration file library")
 191     (description
 192      "Libconfig is a simple library for manipulating structured configuration
 193 files.  This file format is more compact and more readable than XML.  And
 194 unlike XML, it is type-aware, so it is not necessary to do string parsing in
 195 application code.")
 196     (license license:lgpl2.1+)))
 197
 198 (define-public pfff
 199   (package
 200     (name "pfff")
 201     (version "1.0")
 202     (source (origin
 203               (method url-fetch)
 204               (uri (string-append "https://github.com/pfff/pfff/archive/v"
 205                                   version ".tar.gz"))
 206               (file-name (string-append name "-" version ".tar.gz"))
 207               (sha256
 208                (base32
 209                 "00m553aa277iarxj6dalmklyb64r7ias49bfwzbacsfg8h3kar8m"))))
 210     (build-system cmake-build-system)
 211     (home-page "http://biit.cs.ut.ee/pfff/")
 212     (synopsis "Probabilistic fast file fingerprinting tool")
 213     (description
 214      "pfff is a tool for calculating a compact digital fingerprint of a file
 215 by sampling randomly from the file instead of reading it in full.
 216 Consequently, the computation has a flat performance characteristic,
 217 correlated with data variation rather than file size.  pfff can be as reliable
 218 as existing hashing techniques, with provably negligible risk of collisions.")
 219     (license license:bsd-3)))