gnu/packages/datastructures.scm

   1 ;;; GNU Guix --- Functional package management for GNU
   2 ;;; Copyright © 2015, 2016, 2018, 2019 Ricardo Wurmus <rekado@elephly.net>
   3 ;;; Copyright © 2016, 2017, 2019, 2020 Tobias Geerinckx-Rice <me@tobias.gr>
   4 ;;; Copyright © 2018 Meiyo Peng <meiyo.peng@gmail.com>
   5 ;;; Copyright © 2019 Efraim Flashner <efraim@flashner.co.il>
   6 ;;; Copyright © 2020 Mark H Weaver <mhw@netris.org>
   7 ;;;
   8 ;;; This file is part of GNU Guix.
   9 ;;;
  10 ;;; GNU Guix is free software; you can redistribute it and/or modify it
  11 ;;; under the terms of the GNU General Public License as published by
  12 ;;; the Free Software Foundation; either version 3 of the License, or (at
  13 ;;; your option) any later version.
  14 ;;;
  15 ;;; GNU Guix is distributed in the hope that it will be useful, but
  16 ;;; WITHOUT ANY WARRANTY; without even the implied warranty of
  17 ;;; MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  18 ;;; GNU General Public License for more details.
  19 ;;;
  20 ;;; You should have received a copy of the GNU General Public License
  21 ;;; along with GNU Guix.  If not, see <http://www.gnu.org/licenses/>.
  22
  23 (define-module (gnu packages datastructures)
  24   #:use-module (gnu packages)
  25   #:use-module (gnu packages autotools)
  26   #:use-module (gnu packages boost)
  27   #:use-module (gnu packages perl)
  28   #:use-module ((guix licenses) #:prefix license:)
  29   #:use-module (guix packages)
  30   #:use-module (guix download)
  31   #:use-module (guix git-download)
  32   #:use-module (guix build-system cmake)
  33   #:use-module (guix build-system gnu))
  34
  35 (define-public gdsl
  36   (package
  37     (name "gdsl")
  38     (version "1.8")
  39     (source (origin
  40               (method url-fetch)
  41               (uri (string-append "http://download.gna.org/gdsl/"
  42                                   "gdsl-" version ".tar.gz"))
  43               (sha256
  44                (base32
  45                 "1v64jvlnj8jfpphphgjgb36p0kv50kwfyqncf0y12f16v8ydyiaw"))))
  46     (build-system gnu-build-system)
  47     (home-page "http://home.gna.org/gdsl/")
  48     (synopsis "Generic data structures library")
  49     (description "The Generic Data Structures Library (GDSL) is a collection
  50 of routines for generic data structures manipulation.  It is a re-entrant
  51 library fully written from scratch in pure ANSI C.  It is designed to offer
  52 for C programmers common data structures with powerful algorithms, and hidden
  53 implementation.  Available structures are lists, queues, stacks, hash tables,
  54 binary trees, binary search trees, red-black trees, 2D arrays, permutations
  55 and heaps.")
  56     (license license:gpl2+)))
  57
  58 (define-public marisa
  59   (package
  60     (name "marisa")
  61     (version "0.2.6")
  62     (source
  63      (origin
  64        (method url-fetch)
  65        (uri (string-append "https://github.com/s-yata/marisa-trie/files/"
  66                            "4832504/marisa-" version ".tar.gz"))
  67        (sha256
  68         (base32 "1pk6wmi28pa8srb4szybrwfn71jldb61c5vgxsiayxcyg1ya4qqh"))))
  69     (build-system gnu-build-system)
  70     (native-inputs
  71      `(("autoconf" ,autoconf)
  72        ("automake" ,automake)
  73        ("libtool" ,libtool)))
  74     (home-page "https://github.com/s-yata/marisa-trie")
  75     (synopsis "Trie data structure C++ library")
  76     (description "@acronym{MARISA, Matching Algorithm with Recursively
  77 Implemented StorAge} is a static and space-efficient trie data structure C++
  78 library.")
  79
  80     ;; Dual-licensed, according to docs/readme.en.html (source files lack
  81     ;; copyright/license headers.)
  82     (license (list license:bsd-2 license:lgpl2.1+))))
  83
  84 (define-public sparsehash
  85   (package
  86     (name "sparsehash")
  87     (version "2.0.3")
  88     (source (origin
  89               (method git-fetch)
  90               (uri (git-reference
  91                      (url "https://github.com/sparsehash/sparsehash")
  92                      (commit (string-append name "-" version))))
  93               (file-name (git-file-name name version))
  94               (sha256
  95                (base32
  96                 "0m3f0cnpnpf6aak52wn8xbrrdw8p0yhq8csgc8nlvf9zp8c402na"))))
  97     (build-system gnu-build-system)
  98     (synopsis "Memory-efficient hashtable implementations")
  99     (description
 100      "This library contains several hash-map implementations, similar in API
 101 to SGI's @code{hash_map} class, but with different performance
 102 characteristics.  @code{sparse_hash_map} uses very little space overhead, 1-2
 103 bits per entry.  @code{dense_hash_map} is very fast, particularly on lookup.
 104 @code{sparse_hash_set} and @code{dense_hash_set} are the set versions of these
 105 routines.  All these implementation use a hashtable with internal quadratic
 106 probing.  This method is space-efficient -- there is no pointer overhead --
 107 and time-efficient for good hash functions.")
 108     (home-page "https://github.com/sparsehash/sparsehash")
 109     (license license:bsd-3)))
 110
 111 (define-public ssdeep
 112   (package
 113     (name "ssdeep")
 114     (version "2.14.1")
 115     (source
 116      (origin
 117        (method url-fetch)
 118        (uri (string-append "https://github.com/ssdeep-project/ssdeep/"
 119                            "releases/download/release-" version "/"
 120                            "ssdeep-" version ".tar.gz"))
 121        (sha256
 122         (base32 "04qkjc6kksxkv7xbnk32rwmf3a8czdv2vvrdzfs0kw06h73snbpz"))))
 123     (build-system gnu-build-system)
 124     (arguments
 125      `(#:configure-flags
 126        (list "--disable-static")))
 127     (home-page "https://ssdeep-project.github.io")
 128     (synopsis "Context-triggered piecewise hashing algorithm")
 129     (description "ssdeep computes and matches context triggered piecewise
 130 hashes (CTPH), also called fuzzy checksums.  It can identify similar files
 131 that have sequences of identical bytes in the same order, even though bytes
 132 in between these sequences may be different in both content and length.")
 133     (license license:gpl2+)))
 134
 135 (define-public liburcu
 136   (package
 137     (name "liburcu")
 138     (version "0.12.1")
 139     (source (origin
 140               (method url-fetch)
 141               (uri (string-append "https://www.lttng.org/files/urcu/"
 142                                   "userspace-rcu-" version ".tar.bz2"))
 143               (sha256
 144                (base32
 145                 "03nd1gy2c3fdb6xwdrd5lr1jcjxbzffqh3z91mzbjhjn6k8fmymv"))))
 146     (build-system gnu-build-system)
 147     (native-inputs
 148      `(("perl" ,perl)))                 ; for tests
 149     (home-page "https://liburcu.org/")
 150     (synopsis "User-space RCU data synchronisation library")
 151     (description "liburcu is a user-space @dfn{Read-Copy-Update} (RCU) data
 152 synchronisation library.  It provides read-side access that scales linearly
 153 with the number of cores.  liburcu-cds provides efficient data structures
 154 based on RCU and lock-free algorithms.  These structures include hash tables,
 155 queues, stacks, and doubly-linked lists.")
 156     (license license:lgpl2.1+)))
 157
 158 (define-public uthash
 159   (package
 160     (name "uthash")
 161     (version "2.1.0")
 162     (source
 163      (origin
 164        (method git-fetch)
 165        (uri (git-reference
 166              (url "https://github.com/troydhanson/uthash")
 167              (commit (string-append "v" version))))
 168        (file-name (git-file-name name version))
 169        (sha256
 170         (base32 "0k80bjbb6ss5wpmfmfji6xbyjm990hg9kcshwwnhdnh73vxkcd1m"))))
 171     (build-system gnu-build-system)
 172     (native-inputs
 173      `(("perl" ,perl)))
 174     (arguments
 175      `(#:make-flags
 176        (list "CC=gcc")
 177        #:phases
 178        (modify-phases %standard-phases
 179          (delete 'configure)            ; nothing to configure
 180          (delete 'build)                ; nothing to build
 181          (replace 'check
 182            (lambda* (#:key make-flags #:allow-other-keys)
 183              (with-directory-excursion "tests"
 184                (apply invoke "make" make-flags))))
 185          (replace 'install
 186            ;; There is no top-level Makefile to do this for us.
 187            (lambda* (#:key outputs #:allow-other-keys)
 188              (let* ((out (assoc-ref outputs "out"))
 189                     (doc (string-append out "/share/doc/" ,name "-" ,version))
 190                     (include (string-append out "/include")))
 191                ;; Don't install HTML files: they're just the below .txt files
 192                ;; dolled up, can be stale, and regeneration requires asciidoc.
 193                (for-each (λ (file) (install-file file doc))
 194                          (find-files "doc" "\\.txt$"))
 195                (for-each (λ (file) (install-file file include))
 196                          (find-files "src" "\\.h$"))
 197                #t))))))
 198     (home-page "https://troydhanson.github.io/uthash/")
 199     (synopsis
 200      "Hash tables, lists, and other data structures implemented as C macros")
 201     (description
 202      "uthash implements a hash table and a few other basic data structures
 203 as C preprocessor macros.  It aims to be minimalistic and efficient: it's
 204 around 1,000 lines of code which, being macros, inline automatically.
 205
 206 Unlike function calls with fixed prototypes, macros operate on untyped
 207 arguments.  Thus, they are able to work with any type of structure and key.
 208 Any C structure can be stored in a hash table by adding @code{UT_hash_handle}
 209 to the structure and choosing one or more fields to act as the key.")
 210     (license license:bsd-2)))
 211
 212 (define-public sdsl-lite
 213   (package
 214     (name "sdsl-lite")
 215     (version "2.1.1")
 216     (source (origin
 217               (method url-fetch)
 218               (uri (string-append "https://github.com/simongog/sdsl-lite/"
 219                                   "releases/download/v" version "/"
 220                                   "sdsl-lite-" version
 221                                   ".tar.gz.offline.install.gz"))
 222               (sha256
 223                (base32
 224                 "1v86ivv3mmdy802i9xkjpxb4cggj3s27wb19ja4sw1klnivjj69g"))
 225               (modules '((guix build utils)))
 226               (snippet
 227                '(begin
 228                   (delete-file-recursively "external") #t))
 229               (patches
 230                 (list (origin
 231                         (method url-fetch)
 232                         (uri "https://salsa.debian.org/science-team/libsdsl/raw/debian/2.1.1+dfsg-2/debian/patches/0001-Patch-cmake-files.patch")
 233                         (file-name "sdsl-lite-dont-use-bundled-libraries.patch")
 234                         (sha256
 235                          (base32
 236                           "0m542xpys54bni29zibgrfpgpd0zgyny4h131virxsanixsbz52z")))))))
 237     (build-system cmake-build-system)
 238     (arguments
 239      '(#:phases
 240        (modify-phases %standard-phases
 241          (add-after 'install 'install-static-library
 242            (lambda* (#:key outputs #:allow-other-keys)
 243              (let ((out (assoc-ref outputs "out")))
 244                (copy-file "lib/libsdsl_static.a"
 245                           (string-append out "/lib/libsdsl.a")))
 246              #t)))))
 247     (native-inputs
 248      `(("libdivsufsort" ,libdivsufsort)))
 249     (home-page "https://github.com/simongog/sdsl-lite")
 250     (synopsis "Succinct data structure library")
 251     (description "The Succinct Data Structure Library (SDSL) is a powerful and
 252 flexible C++11 library implementing succinct data structures.  In total, the
 253 library contains the highlights of 40 research publications.  Succinct data
 254 structures can represent an object (such as a bitvector or a tree) in space
 255 close to the information-theoretic lower bound of the object while supporting
 256 operations of the original object efficiently.  The theoretical time
 257 complexity of an operation performed on the classical data structure and the
 258 equivalent succinct data structure are (most of the time) identical.")
 259     (license license:gpl3+)))
 260
 261 (define-public libdivsufsort
 262   (package
 263     (name "libdivsufsort")
 264     (version "2.0.1")
 265     (source (origin
 266               (method git-fetch)
 267               (uri (git-reference
 268                     (url "https://github.com/y-256/libdivsufsort")
 269                     (commit version)))
 270               (file-name (git-file-name name version))
 271               (sha256
 272                (base32
 273                 "0fgdz9fzihlvjjrxy01md1bv9vh12rkgkwbm90b1hj5xpbaqp7z2"))))
 274     (build-system cmake-build-system)
 275     (arguments
 276      '(#:tests? #f                      ; there are no tests
 277        #:configure-flags
 278        ;; Needed for rapmap and sailfish.
 279        '("-DBUILD_DIVSUFSORT64=ON")))
 280     (home-page "https://github.com/y-256/libdivsufsort")
 281     (synopsis "Lightweight suffix-sorting library")
 282     (description "libdivsufsort is a software library that implements a
 283 lightweight suffix array construction algorithm.  This library provides a
 284 simple and an efficient C API to construct a suffix array and a
 285 Burrows-Wheeler transformed string from a given string over a constant-size
 286 alphabet.  The algorithm runs in O(n log n) worst-case time using only 5n+O(1)
 287 bytes of memory space, where n is the length of the string.")
 288     (license license:expat)))
 289
 290 (define-public robin-map
 291   (package
 292     (name "robin-map")
 293     (version "0.6.3")
 294     (source (origin
 295               (method git-fetch)
 296               (uri (git-reference
 297                     (url "https://github.com/Tessil/robin-map")
 298                     (commit (string-append "v" version))))
 299               (file-name (git-file-name name version))
 300               (sha256
 301                (base32
 302                 "1li70vwsksva9c4yly90hjafgqfixi1g6d52qq9p6r60vqc4pkjj"))))
 303     (build-system cmake-build-system)
 304     (native-inputs
 305      `(("boost" ,boost)))  ; needed for tests
 306     (arguments
 307      `(#:phases
 308        (modify-phases %standard-phases
 309          (replace 'check
 310            (lambda _
 311              (mkdir "tests")
 312              (with-directory-excursion "tests"
 313                (invoke "cmake" "../../source/tests")
 314                (invoke "cmake" "--build" ".")
 315                (invoke "./tsl_robin_map_tests")))))))
 316     (home-page "https://github.com/Tessil/robin-map")
 317     (synopsis "C++ implementation of a fast hash map and hash set")
 318     (description "The robin-map library is a C++ implementation of a fast hash
 319 map and hash set using open-addressing and linear robin hood hashing with
 320 backward shift deletion to resolve collisions.
 321
 322 Four classes are provided: tsl::robin_map, tsl::robin_set, tsl::robin_pg_map
 323 and tsl::robin_pg_set. The first two are faster and use a power of two growth
 324 policy, the last two use a prime growth policy instead and are able to cope
 325 better with a poor hash function.")
 326     (license license:expat)))