casa-mp-base

# -*- coding: utf-8; mode: tcl; tab-width: 4; indent-tabs-mode: nil; c-basic-offset: 4 -*- vim:fenc=utf-8:ft=tcl:et:sw=4:ts=4:sts=4

​

PortSystem          1.0

​

name                jdepp

version             2015-02-08

categories          textproc japanese

maintainers         nomaintainer

​

description         C++ implementation of Japanese Dependency Parsers

​

long_description    J.DepP is a C++ implementation of Japanese dependency \

                    parsing algorithms. The parser takes a raw sentence as input \

                    and performs word segmentation, POS tagging (thanks to MeCab), \

                    bunsetsu chunking and dependency parsing. J.DepP is meant \

                    for those who want to parse massive texts (e.g., entire blog \

                    feeds) efficiently with state-of-the-art parsing accuracy.

​

homepage            http://www.tkl.iis.u-tokyo.ac.jp/~ynaga/jdepp/

platforms           darwin

license             {GPL-2 LGPL-2.1}

​

master_sites        ${homepage}

checksums           ${distname}${extract.suffix} \

                    rmd160  845584ae822b4ddb0bb439765be7714a97be0ed8 \

                    sha256  55a1a99b6645c4e01126f931fd72bf308dde96df8319f09e1c7cb9cd1ffdae7f

​

depends_lib         port:mecab-base \

                    port:mecab-jumandic-utf8

​

patchfiles          patch-configure.diff

​

post-patch {

    reinplace "s|KNBC_CORPUS_DIR=\${PWD}/KNBC_v1.0_090925|KNBC_CORPUS_DIR=${prefix}/share/KNBC|" \

        ${worksrcpath}/configure

}

​

configure.args      --enable-standalone

​

variant knbc conflicts kyoto kyoto_partial description {Train a parser wht KNBC (Kyoto-University and NTT Blog Corpus)} {

    depends_build-append  port:KNBC

​

    configure.args-append --with-corpus=knbc

​

    build.target          model

    build.args-append     CORPUS_DIR=${prefix}/share/KNBC

}

​

variant kyoto_partial conflicts kyoto knbc description {Train a parser with Kyoto-University Text Corpus annotation without Mainichi news articles} {

    set kyoto_corpus      KyotoCorpus4.0

    set corpus_distfile   ${kyoto_corpus}${extract.suffix}

    distfiles-append      ${corpus_distfile}:corpus

    master_sites-append   http://nlp.ist.i.kyoto-u.ac.jp/nl-resource/corpus/:corpus

    checksums-append      ${corpus_distfile} \

                          rmd160  9c8212d939b7112a79aff30ab0880f94c3ee9b48 \

                          sha256  6fe4f7bac19e192f7545c4de0c765f690a524c7611191e2af9be031d4923871f

​

    configure.args-append --with-corpus=kyoto-partial \

                          --enable-autopos-train=no

​

    build.target          model

    build.args-append     CORPUS_DIR=${workpath}/${kyoto_corpus}

}

​

# To use the complete kyoto corpus with +kyoto variant, users must put their data into the following directory.

set kyoto_dir       /tmp/KyotoCorpus4.0

​

variant kyoto conflicts kyoto_partial knbc description {Train a parser with Kyoto-University Text Corpus} {

    if {[file exists ${kyoto_dir}]} {

        configure.args-append --with-corpus=kyoto 

        build.target          model

        build.args-append     CORPUS_DIR=${kyoto_dir}

    } else {

        pre-fetch {

            ui_error "To train a parser with Kyoto Corpus, please put the files in ${kyoto_dir}"

            return -code error "Kyoto Corpus is not available at ${kyoto_dir}"

        }

    }

}

​

variant ipadic conflicts naistjdic unidic description {Build a parser with auto POSs given by MeCab/ipadic} {

    depends_lib-delete    port:mecab-jumandic-utf8

    depends_lib-append    port:mecab-ipadic-utf8

​

    configure.args-append --with-mecab-dict=IPA

}

​

variant naistjdic conflicts ipadic unidic description {Build a parser with auto POSs given by MeCab/NAIST-jdic} {

    depends_lib-delete    port:mecab-jumandic-utf8

    depends_lib-append    port:mecab-naist-jdic-utf8

​