#include "Thesaurus.h"
-#include "support/debug.h"
-#include "support/gettext.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/os.h"
-#include "support/unicode.h"
-#include "frontends/alert.h"
+#include <cstdio>
+#include MYTHES_H_LOCATION
#include <algorithm>
#include <cstring>
+#include <fstream>
using namespace std;
using namespace lyx::support;
namespace lyx {
-#ifndef HAVE_LIBMYTHES
-#ifdef HAVE_LIBAIKSAURUS
-
-
-Thesaurus::Thesaurus()
- : thes_(new Aiksaurus)
-{}
-
+namespace {
-Thesaurus::~Thesaurus()
-{
- delete thes_;
-}
+typedef std::map<docstring, MyThes *> Thesauri;
+} // namespace
-Thesaurus::Meanings Thesaurus::lookup(docstring const & t, docstring const &)
+struct Thesaurus::Private
{
- Meanings meanings;
-
- // 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;
-
- string const text = to_ascii(t);
-
- docstring error = from_ascii(thes_->error());
- if (!error.empty()) {
- static bool sent_error = false;
- if (!sent_error) {
- frontend::Alert::error(_("Thesaurus failure"),
- bformat(_("Aiksaurus returned the following error:\n\n%1$s."),
- error));
- sent_error = true;
+ ~Private()
+ {
+ for (Thesauri::iterator it = thes_.begin();
+ it != thes_.end(); ++it) {
+ delete it->second;
}
- return meanings;
}
- if (!thes_->find(text.c_str()))
- return meanings;
-
- // 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));
+ ///
+ 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;
}
-
- ret = thes_->next(cur_meaning);
+ return false;
}
- for (Meanings::iterator it = meanings.begin();
- it != meanings.end(); ++it)
- sort(it->second.begin(), it->second.end());
-
- return meanings;
-}
+ ///
+ typedef std::pair<std::string, std::string> 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);
+ /// the thesauri
+ Thesauri thes_;
-bool Thesaurus::thesaurusAvailable(docstring const & lang) const
-{
- // we support English only
- return prefixIs(lang, from_ascii("en_"));
-}
+ /// the location below system/user directory
+ /// there the data+idx files lookup will happen
+ const string dataDirectory(void) { return "thes"; }
-#endif // HAVE_LIBAIKSAURUS
-#endif // !HAVE_LIBMYTHES
+};
-#ifdef HAVE_LIBMYTHES
-
-namespace {
-
-string const to_iconv_encoding(docstring const & s, string const & encoding)
+pair<string,string> Thesaurus::Private::getThesaurus(string const & path, docstring const & lang)
{
- std::vector<char> const encoded =
- ucs4_to_eightbit(s.data(), s.length(), encoding);
- return string(encoded.begin(), encoded.end());
+ 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;
+ }
+ }
+ 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);
}
-docstring const from_iconv_encoding(string const & s, string const & encoding)
+pair<string,string> Thesaurus::Private::getThesaurus(docstring const & lang)
{
- std::vector<char_type> const ucs4 =
- eightbit_to_ucs4(s.data(), s.length(), encoding);
- return docstring(ucs4.begin(), ucs4.end());
-}
-
-} // namespace anon
-
-
-Thesaurus::Thesaurus()
-{}
+ string const thes_path = external_path(lyxrc.thesaurusdir_path);
+ pair<string,string> result ;
+ if (thesaurusAvailable(lang))
+ return make_pair(string(), string());
-Thesaurus::~Thesaurus()
-{
- for (Thesauri::iterator it = thes_.begin();
- it != thes_.end(); ++it) {
- delete it->second;
+ 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;
}
-bool Thesaurus::addThesaurus(docstring const & lang)
+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;
- }
- }
+ ThesFiles files = getThesaurus(lang);
+ string const idx = files.first;
+ string const data = files.second;
if (idx.empty() || data.empty())
return false;
bool Thesaurus::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 d->thesaurusAvailable(lang);
+}
+
- return false;
+bool Thesaurus::thesaurusInstalled(docstring const & lang) const
+{
+ if (thesaurusAvailable(lang))
+ return true;
+ pair<string, string> files = d->getThesaurus(lang);
+ return (!files.first.empty() && !files.second.empty());
}
-Thesaurus::Meanings Thesaurus::lookup(docstring const & t, docstring const & lang)
+Thesaurus::Meanings Thesaurus::lookup(WordLangTuple const & wl)
{
Meanings meanings;
- MyThes * mythes = 0;
+ MyThes * mythes = nullptr;
- if (!addThesaurus(lang))
+ 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 = thes_.begin();
- it != thes_.end(); ++it) {
- if (it->first == lang) {
+ for (Thesauri::const_iterator it = d->thes_.begin();
+ it != d->thes_.end(); ++it) {
+ if (it->first == lang_code) {
mythes = it->second;
break;
}
return meanings;
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;
+ 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<docstring> 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
mythes->CleanUpAfterLookup(&pmean, count);
return meanings;
}
-#else
-#ifndef HAVE_LIBAIKSAURUS
+
Thesaurus::Thesaurus()
+ : d(new Thesaurus::Private)
{
}
Thesaurus::~Thesaurus()
{
+ delete d;
}
-
-Thesaurus::Meanings Thesaurus::lookup(docstring const &, docstring const &)
-{
- return Meanings();
-}
-
-#endif
-#endif // HAVE_LIBMYTHES
-
// Global instance
Thesaurus thesaurus;