X-Git-Url: https://git.lyx.org/gitweb/?a=blobdiff_plain;f=src%2FThesaurus.cpp;h=a80ef87ad7c09bae4d9b829278cc38a255f66699;hb=fab1494f49c0a25f46921db5dc142d7d05f17d17;hp=14c82f61d3c68605a8ad56393dfc1307b21a32bc;hpb=c9e009f4dd802ae62e71b2bf40f3eaed7be41559;p=lyx.git diff --git a/src/Thesaurus.cpp b/src/Thesaurus.cpp index 14c82f61d3..a80ef87ad7 100644 --- a/src/Thesaurus.cpp +++ b/src/Thesaurus.cpp @@ -4,6 +4,7 @@ * Licence details can be found in the file COPYING. * * \author John Levon + * \author Jürgen Spitzmüller * * Full author contact details are available in file CREDITS. */ @@ -12,67 +13,176 @@ #include "Thesaurus.h" +#include "LyXRC.h" + +#include "support/FileNameList.h" +#include "support/debug.h" +#include "support/filetools.h" +#include "support/gettext.h" #include "support/lstrings.h" +#include "support/os.h" +#include "support/unicode.h" + +#include "support/mythes/mythes.hxx" + +#include "frontends/alert.h" #include +#include +using namespace std; +using namespace lyx::support; +using namespace lyx::support::os; namespace lyx { -#ifdef HAVE_LIBAIKSAURUS +namespace { + +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()); +} + -using std::sort; -using std::string; +docstring const from_iconv_encoding(string const & s, string const & encoding) +{ + std::vector const ucs4 = + eightbit_to_ucs4(s.data(), s.length(), encoding); + return docstring(ucs4.begin(), ucs4.end()); +} +typedef std::map Thesauri; -Thesaurus::Thesaurus() - : aik_(new Aiksaurus) -{} +} // namespace anon -Thesaurus::~Thesaurus() +struct Thesaurus::Private +{ + ~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; + } + + /// add a thesaurus to the list + bool addThesaurus(docstring const & lang); + + /// the thesauri + Thesauri thes_; +}; + +bool Thesaurus::Private::addThesaurus(docstring const & lang) +{ + string const thes_path = external_path(lyxrc.thesaurusdir_path); + LYXERR(Debug::FILES, "thesaurus path: " << thes_path); + if (thes_path.empty()) + return false; + + if (thesaurusAvailable(lang)) + return true; + + FileNameList const idx_files = FileName(thes_path).dirList("idx"); + FileNameList const data_files = FileName(thes_path).dirList("dat"); + string idx; + string data; + + for (FileNameList::const_iterator it = idx_files.begin(); + it != idx_files.end(); ++it) { + LYXERR(Debug::FILES, "found thesaurus idx file: " << it->onlyFileName()); + if (contains(it->onlyFileName(), to_ascii(lang))) { + idx = it->absFilename(); + LYXERR(Debug::FILES, "selected thesaurus idx file: " << idx); + break; + } + } + + for (support::FileNameList::const_iterator it = data_files.begin(); + it != data_files.end(); ++it) { + LYXERR(Debug::FILES, "found thesaurus data file: " << it->onlyFileName()); + if (contains(it->onlyFileName(), to_ascii(lang))) { + data = it->absFilename(); + LYXERR(Debug::FILES, "selected thesaurus data file: " << data); + break; + } + } + + if (idx.empty() || data.empty()) + return false; + + char const * af = idx.c_str(); + char const * df = data.c_str(); + thes_[lang] = new MyThes(af, df); + return true; +} + + +bool Thesaurus::thesaurusAvailable(docstring const & lang) const { - delete aik_; + return d->thesaurusAvailable(lang); } -Thesaurus::Meanings Thesaurus::lookup(docstring const & t) +Thesaurus::Meanings Thesaurus::lookup(docstring const & t, docstring const & lang) { Meanings meanings; + MyThes * mythes = 0; - // 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 + if (!d->addThesaurus(lang)) return meanings; - string const text = to_ascii(t); - if (!aik_->find(text.c_str())) + for (Thesauri::const_iterator it = d->thes_.begin(); + it != d->thes_.end(); ++it) { + if (it->first == lang) { + mythes = it->second; + break; + } + } + + if (!mythes) return meanings; - // weird api, but ... + string const encoding = mythes->get_th_encoding(); + + mentry * pmean; + string const text = to_iconv_encoding(support::lowercase(t), encoding); + int len = strlen(text.c_str()); + int count = mythes->Lookup(text.c_str(), len, &pmean); + if (!count) + return meanings; - int prev_meaning = -1; - int cur_meaning; + // don't change value of pmean or count + // they are needed for the CleanUpAfterLookup routine + mentry * pm = pmean; docstring meaning; - - // correct, returns "" at the end - string ret = aik_->next(cur_meaning); - - while (!ret.empty()) { - if (cur_meaning != prev_meaning) { - meaning = from_ascii(ret); - ret = aik_->next(cur_meaning); - prev_meaning = cur_meaning; - } else { - if (ret != text) - meanings[meaning].push_back(from_ascii(ret)); + docstring ret; + for (int i = 0; i < count; i++) { + meaning = from_iconv_encoding(string(pm->defn), encoding); + // remove silly item + if (support::prefixIs(meaning, '-')) + meaning = support::ltrim(meaning, "- "); + for (int j = 0; j < pm->count; j++) { + ret = from_iconv_encoding(string(pm->psyns[j]), encoding); } - - ret = aik_->next(cur_meaning); + meanings[meaning].push_back(ret); + pm++; } + // now clean up all allocated memory + mythes->CleanUpAfterLookup(&pmean, count); for (Meanings::iterator it = meanings.begin(); it != meanings.end(); ++it) @@ -81,25 +191,17 @@ Thesaurus::Meanings Thesaurus::lookup(docstring const & t) return meanings; } -#else -Thesaurus::Thesaurus() +Thesaurus::Thesaurus() : d(new Thesaurus::Private) { } Thesaurus::~Thesaurus() { + delete d; } - -Thesaurus::Meanings Thesaurus::lookup(docstring const &) -{ - return Meanings(); -} - -#endif // HAVE_LIBAIKSAURUS - // Global instance Thesaurus thesaurus;