gnu/packages/unicode.scm

   1 ;;; GNU Guix --- Functional package management for GNU
   2 ;;; Copyright © 2020 Liliana Marie Prikler <liliana.prikler@gmail.com>
   3 ;;; Copyright © 2020 Efraim Flashner <efraim@flashner.co.il>
   4 ;;; Copyright © 2021 Felix Gruber <felgru@posteo.net>
   5 ;;;
   6 ;;; This file is part of GNU Guix.
   7 ;;;
   8 ;;; GNU Guix is free software; you can redistribute it and/or modify it
   9 ;;; under the terms of the GNU General Public License as published by
  10 ;;; the Free Software Foundation; either version 3 of the License, or (at
  11 ;;; your option) any later version.
  12 ;;;
  13 ;;; GNU Guix is distributed in the hope that it will be useful, but
  14 ;;; WITHOUT ANY WARRANTY; without even the implied warranty of
  15 ;;; MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  16 ;;; GNU General Public License for more details.
  17 ;;;
  18 ;;; You should have received a copy of the GNU General Public License
  19 ;;; along with GNU Guix.  If not, see <http://www.gnu.org/licenses/>.
  20
  21 (define-module (gnu packages unicode)
  22   #:use-module (gnu packages autotools)
  23   #:use-module (guix git-download)
  24   #:use-module (guix licenses)
  25   #:use-module (guix packages)
  26   #:use-module (guix download)
  27   #:use-module (guix utils)
  28   #:use-module (guix build-system copy)
  29   #:use-module (guix build-system gnu)
  30   #:use-module (guix build-system trivial))
  31
  32 (define-public libunibreak
  33   (package
  34     (name "libunibreak")
  35     (version "5.0")
  36     (source (origin
  37               (method git-fetch)
  38               (uri (git-reference
  39                     (url "https://github.com/adah1972/libunibreak")
  40                     (commit (string-append "libunibreak_"
  41                               (string-replace-substring version "." "_")))))
  42               (file-name (git-file-name name version))
  43               (sha256
  44                (base32
  45                 "0r5dndhwsiy65lmavz3vdgal9nl8g97hbmdjg6zyq3zh5hs87vwf"))))
  46     (build-system gnu-build-system)
  47     (native-inputs
  48       (list autoconf-wrapper
  49             automake
  50             libtool
  51             ucd-next ; required for tests
  52             ))
  53     (arguments
  54      `(#:parallel-tests? #f  ; parallel tests cause non-deterministic
  55                              ; build failures
  56        #:phases
  57        (modify-phases %standard-phases
  58          (add-before 'check 'pre-check
  59            (lambda* (#:key inputs #:allow-other-keys)
  60              (for-each (lambda (file)
  61                          (copy-file
  62                            (search-input-file inputs
  63                              (string-append "/share/ucd/auxiliary/"
  64                                             file))
  65                            (string-append "src/" file)))
  66                        '("LineBreakTest.txt"
  67                          "WordBreakTest.txt"
  68                          "GraphemeBreakTest.txt")))))))
  69     (home-page "http://vimgadgets.sourceforge.net/libunibreak/")
  70     (synopsis "Unicode line breaking and word breaking algorithms")
  71     (description
  72      "Libunibreak is an implementation of the line breaking and word
  73 breaking algorithms as described in Unicode Standard Annex 14 and
  74 Unicode Standard Annex 29.  It is designed to be used in a generic text
  75 renderer.")
  76     (license zlib)))
  77
  78 (define-public ucd
  79   (package
  80     (name "ucd")
  81     (version "12.0.0")
  82     (source
  83      (origin
  84        (method url-fetch/zipbomb)
  85        (uri (string-append "https://www.unicode.org/Public/zipped/" version
  86                            "/UCD.zip"))
  87        (sha256
  88         (base32
  89          "1ighy39cjkmqnv1797wrxjz76mv1fdw7zp5j04q55bkwxsdkvrmh"))))
  90     (build-system copy-build-system)
  91     (arguments
  92      '(#:install-plan
  93        '(("." "share/ucd/"))))
  94     (home-page "https://www.unicode.org")
  95     (synopsis "Unicode Character Database")
  96     (description
  97      "The @dfn{Unicode Character Database} (UCD) consists of a number of data
  98 files listing Unicode character properties and related data.  It also includes
  99 test data for conformance to several important Unicode algorithms.")
 100     (license unicode)))
 101
 102 (define-public ucd-next
 103   (package
 104     (inherit ucd)
 105     (name "ucd-next")
 106     (version "14.0.0")
 107     (source
 108      (origin
 109        (method url-fetch/zipbomb)
 110        (uri (string-append "https://www.unicode.org/Public/zipped/" version
 111                            "/UCD.zip"))
 112        (sha256
 113         (base32
 114          "001nq9w52ijma0vps40xwy2q6ylpyf1393lzb128ibypnmv54fh3"))))))
 115
 116 (define (unicode-emoji-file name version hash)
 117   (origin
 118     (method url-fetch)
 119     (uri (string-append "https://www.unicode.org/Public/emoji/"
 120                         version
 121                         "/emoji-" name ".txt"))
 122     (sha256 (base32 hash))))
 123
 124 (define-public unicode-emoji
 125   (package
 126     (name "unicode-emoji")
 127     (version "12.0")
 128     (source #f)
 129     (build-system trivial-build-system)
 130     (arguments
 131      `(#:modules ((guix build utils))
 132        #:builder
 133        (let ((out (string-append %output "/share/unicode/emoji")))
 134          (use-modules (guix build utils))
 135          (mkdir-p out)
 136          (for-each
 137           (lambda (input)
 138             (copy-file
 139              (cdr input)
 140              (string-append out "/"
 141                             (substring (car input) 8) ; strip "unicode-"
 142                             ".txt")))
 143           %build-inputs)
 144          #t)))
 145     (inputs
 146      `(("unicode-emoji-data"
 147         ,(unicode-emoji-file
 148           "data" version
 149           "03sf7h1d6kb9m5s02lif88jsi5kjszpkfvcymaqxj8ds70ar9pgv"))
 150        ("unicode-emoji-sequences"
 151         ,(unicode-emoji-file
 152           "sequences" version
 153           "1hghki2rn3n7m4lwpwi2a5wrsf2nij4bxga9ldabx4g0g2k23svs"))
 154        ("unicode-emoji-test"
 155         ,(unicode-emoji-file
 156           "test" version
 157           "1dqd0fh999mh6naj816ni113m9dimfy3ih9nffjq2lrv9mmlgdck"))
 158        ("unicode-emoji-variation-sequences"
 159         ,(unicode-emoji-file
 160           "variation-sequences" version
 161           "1cccwx5bl79w4c19vi5dhjqxrph92s8hihp9y8s2cqvdzmgbln7a"))
 162        ("unicode-emoji-zwj-sequences"
 163         ,(unicode-emoji-file
 164           "zwj-sequences" version
 165           "1l791nbijmmhwa7kmvfn8gp26ban512l6mgqpz1mnbq3xm19181n"))))
 166     (home-page "https://www.unicode.org")
 167     (synopsis "Unicode Emoji data")
 168     (description
 169      "This package includes data files listing characters and sequences, that
 170 Unicode emoji supporting fonts or keyboards should support according to the
 171 Unicode Technological Standard #51.")
 172     (license unicode)))
 173
 174 (define-public unicode-cldr-common
 175   (package
 176     (name "unicode-cldr-common")
 177     (version "36.0")
 178     (source
 179      (origin
 180        (method url-fetch/zipbomb)
 181        (uri (string-append "https://unicode.org/Public/cldr/"
 182                            (version-major version)
 183                            "/cldr-common-" version ".zip"))
 184        (sha256
 185         (base32
 186          "0hxsc3j5zb32hmiaj0r3ajchmklx6zng6zlh1ca6s9plq5b9w9q7"))))
 187     (build-system copy-build-system)
 188     (arguments
 189      '(#:install-plan
 190        '(("common" "share/unicode/cldr/"))))
 191     (home-page "https://www.unicode.org")
 192     (synopsis "Locale data repository")
 193     (description
 194      "The Unicode Common Locale Data Repository (CLDR) is a large repository
 195 of locale data, including among others
 196
 197 @itemize
 198 @item patterns for formatting and parsing,
 199 @item name translations,
 200 @item and various information on languages, scripts and country-specific
 201   conventions.
 202 @end itemize\n")
 203     (license unicode)))