gnu/packages/search.scm

   1 ;;; GNU Guix --- Functional package management for GNU
   2 ;;; Copyright © 2014 Mark H Weaver <mhw@netris.org>
   3 ;;; Copyright © 2015, 2016 Eric Bavier <bavier@member.fsf.org>
   4 ;;; Copyright © 2017 Thomas Danckaert <post@thomasdanckaert.be>
   5 ;;; Copyright © 2017 Ricardo Wurmus <rekado@elephly.net>
   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 search)
  23   #:use-module ((guix licenses)
  24                 #:select (gpl2 gpl2+ gpl3+ lgpl2.1+ bsd-3 x11))
  25   #:use-module (guix packages)
  26   #:use-module (guix download)
  27   #:use-module (guix utils)
  28   #:use-module (guix build-system gnu)
  29   #:use-module (gnu packages)
  30   #:use-module (gnu packages compression)
  31   #:use-module (gnu packages check)
  32   #:use-module (gnu packages databases)
  33   #:use-module (gnu packages linux)
  34   #:use-module (gnu packages perl)
  35   #:use-module (gnu packages python)
  36   #:use-module (gnu packages web)
  37   #:use-module (gnu packages xml))
  38
  39 (define-public xapian
  40   (package
  41     (name "xapian")
  42     (version "1.4.5")
  43     ;; Note: When updating Xapian, remember to update xapian-bindings below.
  44     (source (origin
  45               (method url-fetch)
  46               (uri (string-append "https://oligarchy.co.uk/xapian/" version
  47                                   "/xapian-core-" version ".tar.xz"))
  48               (sha256
  49                (base32 "0axhqrj202hbll9mcx1qdm8gsqj19216w3z02gyjbycxvr9gkdc5"))))
  50     (build-system gnu-build-system)
  51     (inputs `(("zlib" ,zlib)
  52               ("util-linux" ,util-linux)))
  53     (arguments
  54      `(#:phases
  55        (modify-phases %standard-phases
  56          (replace 'check
  57            ;; As of Xapian 1.3.3, the TCP server implementation uses
  58            ;; getaddrinfo(). This does not work in the build environment,
  59            ;; so exclude those tests. See HACKING for the list of targets.
  60            (lambda _
  61              (zero? (system* "make"
  62                              "check-inmemory"
  63                              "check-remoteprog"
  64                              ;"check-remotetcp"
  65                              "check-multi"
  66                              "check-glass"
  67                              "check-chert")))))))
  68     (synopsis "Search Engine Library")
  69     (description
  70      "Xapian is a highly adaptable toolkit which allows developers to easily
  71 add advanced indexing and search facilities to their own applications.  It
  72 supports the Probabilistic Information Retrieval model and also supports a
  73 rich set of boolean query operators.")
  74     (home-page "https://xapian.org/")
  75     (license (list gpl2+ bsd-3 x11))))
  76
  77 (define-public python-xapian-bindings
  78   (package (inherit xapian)
  79     (name "python-xapian-bindings")
  80     (version (package-version xapian))
  81     (source (origin
  82               (method url-fetch)
  83               (uri (string-append "https://oligarchy.co.uk/xapian/" version
  84                                   "/xapian-bindings-" version ".tar.xz"))
  85               (sha256
  86                (base32
  87                 "0cwx39764w24xd25w271had4w78lnw1dgz36yvlw1g3i19rqcy34"))))
  88     (build-system gnu-build-system)
  89     (arguments
  90      `(#:configure-flags '("--with-python3")
  91        #:make-flags
  92        (list (string-append "pkgpylibdir="
  93                             (assoc-ref %outputs "out")
  94                             "/lib/python" ,(version-major+minor
  95                                             (package-version python))
  96                             "/site-packages/xapian"))))
  97     (inputs
  98      `(("python" ,python)
  99        ("python-sphinx" ,python-sphinx) ; for documentation
 100        ("xapian" ,xapian)
 101        ("zlib" ,zlib)))
 102     (synopsis "Python bindings for the Xapian search engine library")
 103     (license gpl2+)))
 104
 105 (define-public libtocc
 106   (package
 107     (name "libtocc")
 108     (version "1.0.1")
 109     (source
 110      (origin
 111        (method url-fetch)
 112        (uri (string-append "https://github.com/aidin36/tocc/releases/download/"
 113                            "v" version "/tocc-" version ".tar.gz"))
 114        (sha256
 115         (base32
 116          "1kd2jd74m8ksc8s7hh0haz0q0c3n0mr39bbky262kk4l58f1g068"))))
 117     (build-system gnu-build-system)
 118     (native-inputs `(("catch" ,catch-framework)))
 119     (inputs `(("unqlite" ,unqlite)))
 120     (arguments
 121      `(#:phases (modify-phases %standard-phases
 122                   (add-before
 123                    'configure 'chdir-source
 124                    (lambda _ (chdir "libtocc/src")))
 125                   (replace
 126                    'check
 127                    (lambda _
 128                      (with-directory-excursion "../tests"
 129                        (and (zero? (system* "./configure"
 130                                             (string-append "CONFIG_SHELL="
 131                                                            (which "sh"))
 132                                             (string-append "SHELL="
 133                                                            (which "sh"))
 134                                             "CPPFLAGS=-I../src"
 135                                             (string-append
 136                                              "LDFLAGS=-L../src/.libs "
 137                                              "-Wl,-rpath=../src/.libs")))
 138                             (zero? (system* "make"))
 139                             (zero? (system* "./libtocctests")))))))))
 140     (home-page "https://t-o-c-c.com/")
 141     (synopsis "Tool for Obsessive Compulsive Classifiers")
 142     (description
 143      "libtocc is the engine of the Tocc project, a tag-based file management
 144 system.  The goal of Tocc is to provide a better system for classifying files
 145 that is more flexible than classic file systems that are based on a tree of
 146 files and directories.")
 147     (license gpl3+)))
 148
 149 (define-public tocc
 150   (package
 151     (name "tocc")
 152     (version (package-version libtocc))
 153     (source (package-source libtocc))
 154     (build-system gnu-build-system)
 155     (inputs
 156      `(("libtocc" ,libtocc)
 157        ("unqlite" ,unqlite)))
 158     (arguments
 159      `(#:tests? #f                      ;No tests
 160        #:phases (modify-phases %standard-phases
 161                   (add-after
 162                    'unpack 'chdir-source
 163                    (lambda _ (chdir "cli/src"))))))
 164     (home-page "https://t-o-c-c.com/")
 165     (synopsis "Command-line interface to libtocc")
 166     (description
 167      "Tocc is a tag-based file management system.  This package contains the
 168 command line tool for interacting with libtocc.")
 169     (license gpl3+)))
 170
 171 (define-public bool
 172   (package
 173     (name "bool")
 174     (version "0.2.2")
 175     (source
 176      (origin
 177        (method url-fetch)
 178        (uri (string-append "mirror://gnu/bool/bool-"
 179                            version ".tar.xz"))
 180        (sha256
 181         (base32
 182          "1frdmgrmb509fxbdpsxxw3lvvwv7xm1pavqrqgm4jg698iix6xfw"))))
 183     (build-system gnu-build-system)
 184     (home-page "https://www.gnu.org/software/bool/")
 185     (synopsis "Finding text and HTML files that match boolean expressions")
 186     (description
 187      "GNU Bool is a utility to perform text searches on files using Boolean
 188 expressions.  For example, a search for \"hello AND world\" would return a
 189 file containing the phrase \"Hello, world!\".  It supports both AND and OR
 190 statements, as well as the NEAR statement to search for the occurrence of
 191 words in close proximity to each other.  It handles context gracefully,
 192 accounting for new lines and paragraph changes.  It also has robust support
 193 for parsing HTML files.")
 194     (license gpl3+)))
 195
 196 (define-public hyperestraier
 197   (package
 198     (name "hyperestraier")
 199     (version "1.4.13")
 200     (source
 201       (origin
 202         (method url-fetch)
 203         (uri (string-append "http://fallabs.com/" name "/"
 204                             name "-" version ".tar.gz"))
 205         (sha256
 206          (base32
 207           "1qk3pxgzyrpcz5qfyd5xs2hw9q1cbb7j5zd4kp1diq501wcj2vs9"))))
 208     (inputs
 209      `(("qdbm" ,qdbm)
 210        ("zlib" ,zlib)))
 211     (build-system gnu-build-system)
 212     (arguments
 213      `(#:configure-flags (list (string-append "LDFLAGS=-Wl,-rpath="
 214                                               (assoc-ref %outputs "out")
 215                                               "/lib"))))
 216     (home-page "http://fallabs.com/hyperestraier")
 217     (synopsis "Full-text search system")
 218     (description "Hyper Estraier can be used to integrate full-text
 219 search into applications, using either the provided command line and CGI
 220 interfaces, or a C API.")
 221     (license lgpl2.1+)))
 222
 223 (define-public mlocate
 224   (package
 225     (name "mlocate")
 226     (version "0.26")
 227     (source (origin
 228               (method url-fetch)
 229               (uri (string-append "http://releases.pagure.org/mlocate/"
 230                                   "mlocate-" version ".tar.xz"))
 231               (sha256
 232                (base32
 233                 "0gi6y52gkakhhlnzy0p6izc36nqhyfx5830qirhvk3qrzrwxyqrh"))))
 234     (build-system gnu-build-system)
 235     (home-page "https://pagure.io/mlocate")
 236     (synopsis "Locate files on the file system")
 237     (description
 238      "mlocate is a locate/updatedb implementation.  The 'm' stands for
 239 \"merging\": @code{updatedb} reuses the existing database to avoid rereading
 240 most of the file system, which makes it faster and does not trash the system
 241 caches as much.  The locate(1) utility is intended to be completely compatible
 242 with slocate, and attempts to be compatible to GNU locate when it does not
 243 conflict with slocate compatibility.")
 244     (license gpl2)))
 245
 246 (define-public swish-e
 247   (package
 248     (name "swish-e")
 249     (version "2.4.7")
 250     (source (origin
 251               (method url-fetch)
 252               (uri (list (string-append
 253                           "https://web.archive.org/web/20160730145202/"
 254                           "http://swish-e.org/distribution/"
 255                           "swish-e-" version ".tar.gz")
 256                          (string-append "http://http.debian.net/debian/pool/"
 257                                         "main/s/swish-e/swish-e_" version
 258                                         ".orig.tar.gz")))
 259               (file-name (string-append name "-" version ".tar.gz"))
 260               (sha256
 261                (base32
 262                 "0qkrk7z25yp9hynj21vxkyn7yi8gcagcfxnass5cgczcz0gm9pax"))
 263               (patches (search-patches "swish-e-search.patch"
 264                                        "swish-e-format-security.patch"))))
 265     (build-system gnu-build-system)
 266     ;; Several other packages and perl modules may be installed alongside
 267     ;; swish-e to extend its features at runtime, but are not required for
 268     ;; building: xpdf, catdoc, MP3::Tag, Spreadsheet::ParseExcel,
 269     ;; HTML::Entities.
 270     (inputs
 271      `(("perl" ,perl)
 272        ("perl-uri" ,perl-uri)
 273        ("perl-html-parser" ,perl-html-parser)
 274        ("perl-html-tagset" ,perl-html-tagset)
 275        ("perl-mime-types" ,perl-mime-types)))
 276     (arguments
 277      `(;; XXX: This fails to build with zlib (API mismatch) and tests fail
 278        ;; with libxml2, so disable both.
 279        #:configure-flags (list (string-append "--without-zlib")
 280                                (string-append "--without-libxml2"))
 281        #:phases (modify-phases %standard-phases
 282                   (add-after 'install 'wrap-programs
 283                     (lambda* (#:key inputs outputs #:allow-other-keys)
 284                       (let* ((out (assoc-ref outputs "out")))
 285                         (for-each
 286                          (lambda (program)
 287                            (wrap-program program
 288                              `("PERL5LIB" ":" prefix
 289                                ,(map (lambda (i)
 290                                        (string-append (assoc-ref inputs i)
 291                                                       "/lib/perl5/site_perl"))
 292                                      ;; These perl modules have no propagated
 293                                      ;; inputs, so no further analysis needed.
 294                                      '("perl-uri"
 295                                        "perl-html-parser"
 296                                        "perl-html-tagset"
 297                                        "perl-mime-types")))))
 298                          (list (string-append out "/lib/swish-e/swishspider")
 299                                (string-append out "/bin/swish-filter-test")))
 300                         #t))))))
 301     (home-page (string-append "https://web.archive.org/web/20160730145202/"
 302                               "http://swish-e.org"))
 303     (synopsis "Web indexing system")
 304     (description
 305      "Swish-e is Simple Web Indexing System for Humans - Enhanced.  Swish-e
 306 can quickly and easily index directories of files or remote web sites and
 307 search the generated indexes.")
 308     (license gpl2+)))                   ;with exception
 309
 310 ;;; search.scm ends here