X-Git-Url: https://git.lyx.org/gitweb/?a=blobdiff_plain;f=src%2FThesaurus.cpp;h=1c1adc03b0e22cb91154ac21e396957b05997bb3;hb=210ea66d2d1113cddcc1b0f3547c1063b7dc4d47;hp=c9e9e8f3b1293ec4fc427faf3ddbaad636aac63e;hpb=d2d31b9130a3a0eafef6378864baaddde90999f7;p=lyx.git diff --git a/src/Thesaurus.cpp b/src/Thesaurus.cpp index c9e9e8f3b1..1c1adc03b0 100644 --- a/src/Thesaurus.cpp +++ b/src/Thesaurus.cpp @@ -4,7 +4,7 @@ * Licence details can be found in the file COPYING. * * \author John Levon - * \author Jürgen Spitzmüller + * \author Jürgen Spitzmüller * * Full author contact details are available in file CREDITS. */ @@ -13,152 +13,268 @@ #include "Thesaurus.h" -#include "debug.h" +#include "LyXRC.h" + +#include "SpellChecker.h" +#include "WordLangTuple.h" + +#include "support/FileNameList.h" +#include "support/Package.h" +#include "support/debug.h" +#include "support/docstring.h" +#include "support/docstring_list.h" +#include "support/filetools.h" #include "support/lstrings.h" -#include "support/unicode.h" +#include "support/os.h" + +#include +#include MYTHES_H_LOCATION #include +#include +#include +using namespace std; +using namespace lyx::support; +using namespace lyx::support::os; namespace lyx { -using std::sort; -using std::string; -using std::endl; - +namespace { -#ifndef HAVE_LIBMYTHES -#ifdef HAVE_LIBAIKSAURUS -Thesaurus::Thesaurus() - : thes_(new Aiksaurus) -{} +typedef std::map Thesauri; +} // namespace -Thesaurus::~Thesaurus() +struct Thesaurus::Private { - delete thes_; -} - + ~Private() + { + for (Thesauri::iterator it = thes_.begin(); + it != thes_.end(); ++it) { + delete it->second; + } + } + /// + bool thesaurusAvailable(docstring const & lang) const + { + for (Thesauri::const_iterator it = thes_.begin(); + it != thes_.end(); ++it) { + if (it->first == lang) + if (it->second) + return true; + } + return false; + } -Thesaurus::Meanings Thesaurus::lookup(docstring const & t) -{ - Meanings meanings; + /// + typedef std::pair ThesFiles; + /// + ThesFiles getThesaurus(string const & path, docstring const & lang); + ThesFiles getThesaurus(docstring const & lang); + /// add a thesaurus to the list + bool addThesaurus(docstring const & lang); - // aiksaurus is for english text only, therefore it does not work - // with non-ascii strings. - // The interface of the Thesaurus class uses docstring because a - // non-english thesaurus is possible in theory. - if (!support::isAscii(t)) - // to_ascii() would assert - return meanings; + /// the thesauri + Thesauri thes_; - string const text = to_ascii(t); - if (!thes_->find(text.c_str())) - return meanings; + /// the location below system/user directory + /// there the data+idx files lookup will happen + const string dataDirectory(void) { return "thes"; } - // weird api, but ... +}; - int prev_meaning = -1; - int cur_meaning; - docstring meaning; - // correct, returns "" at the end - string ret = thes_->next(cur_meaning); - - while (!ret.empty()) { - if (cur_meaning != prev_meaning) { - meaning = from_ascii(ret); - ret = thes_->next(cur_meaning); - prev_meaning = cur_meaning; - } else { - if (ret != text) - meanings[meaning].push_back(from_ascii(ret)); +pair Thesaurus::Private::getThesaurus(string const & path, docstring const & lang) +{ + FileName base(path); + if (!base.isDirectory()) { + return make_pair(string(), string()); + } + FileNameList const idx_files = base.dirList("idx"); + FileNameList const data_files = base.dirList("dat"); + string idx; + string data; + string basename; + + LYXERR(Debug::FILES, "thesaurus path: " << path); + for (FileNameList::const_iterator it = idx_files.begin(); it != idx_files.end(); ++it) { + basename = it->onlyFileNameWithoutExt(); + if (contains(basename, to_ascii(lang))) { + // do not use more specific dicts. + if (contains(basename, '_') && !contains(lang, '_')) + continue; + if (contains(basename, '-') && !contains(lang, '-')) + continue; + ifstream ifs(it->absFileName().c_str()); + if (ifs) { + // check for appropriate version of index file + string encoding; // first line is encoding + int items = 0; // second line is no. of items + getline(ifs,encoding); + ifs >> items; + if (ifs.fail()) { + LYXERR(Debug::FILES, "ignore irregular thesaurus idx file: " << it->absFileName()); + continue; + } + if (encoding.length() == 0 || encoding.find_first_of(',') != string::npos) { + LYXERR(Debug::FILES, "ignore version1 thesaurus idx file: " << it->absFileName()); + continue; + } + } + idx = it->absFileName(); + LYXERR(Debug::FILES, "selected thesaurus idx file: " << idx); + break; } - - ret = thes_->next(cur_meaning); } - - for (Meanings::iterator it = meanings.begin(); - it != meanings.end(); ++it) - sort(it->second.begin(), it->second.end()); - - return meanings; + if (idx.empty()) { + // try with a more general dictionary + docstring shortcode; + if (contains(lang, '_')) { + split(lang, shortcode, '_'); + LYXERR(Debug::FILES, "Did not find thesaurus for LANG code " + << lang << ". Trying with " << shortcode); + return getThesaurus(path, shortcode); + } + else if (contains(lang, '-')) { + split(lang, shortcode, '-'); + LYXERR(Debug::FILES, "Did not find thesaurus for LANG code " + << lang << ". Trying with " << shortcode); + return getThesaurus(path, shortcode); + } + return make_pair(string(), string()); + } + for (support::FileNameList::const_iterator it = data_files.begin(); it != data_files.end(); ++it) { + if (contains(it->onlyFileName(), basename)) { + data = it->absFileName(); + LYXERR(Debug::FILES, "selected thesaurus data file: " << data); + break; + } + } + return make_pair(idx, data); } -#endif // HAVE_LIBAIKSAURUS -#endif // !HAVE_LIBMYTHES +pair Thesaurus::Private::getThesaurus(docstring const & lang) +{ + string const thes_path = external_path(lyxrc.thesaurusdir_path); + pair result ; -#ifdef HAVE_LIBMYTHES - -namespace { + if (thesaurusAvailable(lang)) + return make_pair(string(), string()); -string const to_iconv_encoding(docstring const & s, string const & encoding) -{ - std::vector const encoded = - ucs4_to_eightbit(s.data(), s.length(), encoding); - return string(encoded.begin(), encoded.end()); + if (!thes_path.empty()) + result = getThesaurus(thes_path, lang); + if (result.first.empty() || result.second.empty()) { + string const sys_path = external_path(addName(lyx::support::package().system_support().absFileName(),dataDirectory())) ; + result = getThesaurus(sys_path, lang); + } + if (result.first.empty() || result.second.empty()) { + string const user_path = external_path(addName(lyx::support::package().user_support().absFileName(),dataDirectory())) ; + result = getThesaurus(user_path, lang); + } + return result; } -docstring const from_iconv_encoding(string const & s, string const & encoding) +bool Thesaurus::Private::addThesaurus(docstring const & lang) { - std::vector const ucs4 = - eightbit_to_ucs4(s.data(), s.length(), encoding); - return docstring(ucs4.begin(), ucs4.end()); -} + if (thesaurusAvailable(lang)) + return true; -} // namespace anon + ThesFiles files = getThesaurus(lang); + string const idx = files.first; + string const data = files.second; + if (idx.empty() || data.empty()) + return false; -Thesaurus::Thesaurus() -{ - string const idx("/home/juergen/updates/MyThes-1.0/th_de_DE_v2.idx"); - string const data("/home/juergen/updates/MyThes-1.0/th_de_DE_v2.dat"); char const * af = idx.c_str(); char const * df = data.c_str(); - thes_ = new MyThes(af, df); + thes_[lang] = new MyThes(af, df); + return true; } -Thesaurus::~Thesaurus() +bool Thesaurus::thesaurusAvailable(docstring const & lang) const +{ + return d->thesaurusAvailable(lang); +} + + +bool Thesaurus::thesaurusInstalled(docstring const & lang) const { - if (thes_) - delete thes_; + if (thesaurusAvailable(lang)) + return true; + pair files = d->getThesaurus(lang); + return (!files.first.empty() && !files.second.empty()); } -Thesaurus::Meanings Thesaurus::lookup(docstring const & t) +Thesaurus::Meanings Thesaurus::lookup(WordLangTuple const & wl) { Meanings meanings; + MyThes * mythes = nullptr; + + docstring const lang_code = from_ascii(wl.lang()->code()); + docstring const & t = wl.word(); + + if (!d->addThesaurus(lang_code)) + return meanings; + + for (Thesauri::const_iterator it = d->thes_.begin(); + it != d->thes_.end(); ++it) { + if (it->first == lang_code) { + mythes = it->second; + break; + } + } + + if (!mythes) + return meanings; + + string const encoding = mythes->get_th_encoding(); - string const encoding = thes_->get_th_encoding(); - mentry * pmean; string const text = to_iconv_encoding(support::lowercase(t), encoding); int len = strlen(text.c_str()); - int count = thes_->Lookup(text.c_str(), len, &pmean); - if (!count) - return meanings; + int count = mythes->Lookup(text.c_str(), len, &pmean); + if (!count) { + SpellChecker * speller = theSpellChecker(); + if (!speller) + return meanings; + docstring_list suggestions; + speller->stem(wl, suggestions); + for (size_t i = 0; i != suggestions.size(); ++i) { + string const wordform = to_iconv_encoding(support::lowercase(suggestions[i]), encoding); + len = strlen(wordform.c_str()); + count = mythes->Lookup(wordform.c_str(), len, &pmean); + if (count) + break; + } + if (!count) + return meanings; + } // don't change value of pmean or count // they are needed for the CleanUpAfterLookup routine mentry * pm = pmean; docstring meaning; - docstring ret; for (int i = 0; i < count; i++) { + vector ret; meaning = from_iconv_encoding(string(pm->defn), encoding); // remove silly item if (support::prefixIs(meaning, '-')) meaning = support::ltrim(meaning, "- "); + ret.reserve(pm->count); for (int j = 0; j < pm->count; j++) { - ret = from_iconv_encoding(string(pm->psyns[j]), encoding); + ret.push_back(from_iconv_encoding(string(pm->psyns[j]), encoding)); } - meanings[meaning].push_back(ret); - pm++; + meanings[meaning] = ret; + ++pm; } - // now clean up all allocated memory - thes_->CleanUpAfterLookup(&pmean, count); + // now clean up all allocated memory + mythes->CleanUpAfterLookup(&pmean, count); for (Meanings::iterator it = meanings.begin(); it != meanings.end(); ++it) @@ -167,25 +283,18 @@ Thesaurus::Meanings Thesaurus::lookup(docstring const & t) return meanings; } -#else Thesaurus::Thesaurus() + : d(new Thesaurus::Private) { } Thesaurus::~Thesaurus() { + delete d; } - -Thesaurus::Meanings Thesaurus::lookup(docstring const &) -{ - return Meanings(); -} - -#endif // HAVE_LIBMYTHES - // Global instance Thesaurus thesaurus;