From ed8b1bcdad7c430af1eef5fbe78b6ec4eb3eb60e Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?P=C5=99emysl=20Eric=20Janouch?= Date: Thu, 7 Oct 2021 03:37:12 +0200 Subject: [PATCH] Add sample dictionary downloaders/builders --- CMakeLists.txt | 16 +++++++++++++++- README.adoc | 5 +++-- dicts/gnu-fdl-en-cz.sh | 21 +++++++++++++++++++++ dicts/slovnik-cizich-slov.sh | 8 ++++++++ 4 files changed, 47 insertions(+), 3 deletions(-) create mode 100755 dicts/gnu-fdl-en-cz.sh create mode 100755 dicts/slovnik-cizich-slov.sh diff --git a/CMakeLists.txt b/CMakeLists.txt index bd42553..9f07bfb 100644 --- a/CMakeLists.txt +++ b/CMakeLists.txt @@ -172,6 +172,21 @@ endforeach () add_custom_target (tools DEPENDS ${tools}) +# Example dictionaries +file (GLOB dicts_scripts "${PROJECT_SOURCE_DIR}/dicts/*.sh") +set (dicts_targets) +foreach (dict_script ${dicts_scripts}) + get_filename_component (dict_name "${dict_script}" NAME_WE) + list (APPEND dicts_targets "dicts-${dict_name}") + add_custom_target (dicts-${dict_name} + COMMAND sh -c "PATH=.:$PATH \"$0\"" "${dict_script}" + DEPENDS tabfile + COMMENT "Generating sample dictionary ${dict_name}" + VERBATIM) +endforeach () + +add_custom_target (dicts DEPENDS ${dicts_targets}) + # The files to be installed include (GNUInstallDirs) install (TARGETS ${PROJECT_NAME} DESTINATION ${CMAKE_INSTALL_BINDIR}) @@ -212,4 +227,3 @@ set (CPACK_SOURCE_IGNORE_FILES "/\\\\.git;/build;/CMakeLists.txt.user") set (CPACK_SOURCE_PACKAGE_FILE_NAME "${PROJECT_NAME}-${PROJECT_VERSION}") include (CPack) - diff --git a/README.adoc b/README.adoc index 4f98061..724c7e8 100644 --- a/README.adoc +++ b/README.adoc @@ -101,13 +101,14 @@ Dictionaries Unfortunately this application only really works with specific dictionaries. Word definitions have to be in plain text, separated by newlines. +The `make dicts` command will build some examples from freely available sources. + You may use the included transform tool to transform existing dictionaries that are almost useful as they are, e.g. after stripping XML tags. You might want to fix up the `sametypesequence` of the resulting '.ifo' file afterwards, and run dictzip on the resulting '.dict' file. -https://mega.co.nz/#!axtD0QRK!sbtBgizksyfkPqKvKEgr8GQ11rsWhtqyRgUUV0B7pwg[ -CZ <--> { EN, DE, PL, RU } dictionaries] +https://mega.co.nz/#!axtD0QRK!sbtBgizksyfkPqKvKEgr8GQ11rsWhtqyRgUUV0B7pwg[CZ <--> EN/DE/PL/RU dictionaries] Contributing and Support ------------------------ diff --git a/dicts/gnu-fdl-en-cz.sh b/dicts/gnu-fdl-en-cz.sh new file mode 100755 index 0000000..8feeea9 --- /dev/null +++ b/dicts/gnu-fdl-en-cz.sh @@ -0,0 +1,21 @@ +#!/bin/sh -e +# GNU/FDL English-Czech dictionary, see https://www.svobodneslovniky.cz/ +curl -Lo- https://www.svobodneslovniky.cz/data/en-cs.txt.gz | \ +zcat | grep -v ^# | sed 's/\\//g' | perl -CSD -F\\t -le ' + sub e { shift =~ s/\\/\\\\/gr =~ s/\n/\\n/gr =~ s/\t/\\t/gr } + sub w { + open(my $f, "|-", "tabfile gnu-fdl-$_[0]") or die $!; + print $f e($k) . "\t" . e(join("\n", @$v)) + while ($k, $v) = each %{$_[1]}; + close($f); + } + my ($en, $cz, $notes, $special, $translator) = @F; + if ($cz) { + $notes =~ s/\w+:\s?//g; # remove word classes + $notes =~ s/(\w+\.)(?!])/($1)/; # quote "pl." + push(@{$encz{$en}}, $notes ? "$cz " . $notes : $cz); + push(@{$czen{$cz}}, $notes ? "$en " . $notes : $en); + } END { + w("en-cz", \%encz); + w("cz-en", \%czen); + }' diff --git a/dicts/slovnik-cizich-slov.sh b/dicts/slovnik-cizich-slov.sh new file mode 100755 index 0000000..cb99ea1 --- /dev/null +++ b/dicts/slovnik-cizich-slov.sh @@ -0,0 +1,8 @@ +#!/bin/sh -e +# Slovník cizích slov, see https://slovnik-cizich-slov.abz.cz/web.php/o-slovniku +# TODO: Skipping the optional pronunciation field, tabfile can't handle it yet, +# but could be made to accept a lowercase sametypesequence +curl -Lo- https://slovnik-cizich-slov.abz.cz/export.php | \ +iconv -f latin2 -t UTF-8 | perl -CSD -F\\\| -le ' + print "$_\t" . $F[2] =~ s/\\/\\\\/gr =~ s/; /\\n/gr for split(", ", $F[0]) +' | sort -u | tabfile slovnik-cizich-slov