X-Git-Url: https://git.lyx.org/gitweb/?a=blobdiff_plain;f=src%2FThesaurus.cpp;h=b4655ac27f93e6290fc320ba928c37404c515a45;hb=2e72a0af48023956ae6ce3285b47168c2fed5099;hp=42daa0c8ebe018b3dc6b6c1c0ffb24512b5dc429;hpb=59b58c5c0a6df26de57743eb8d8319c963dde4c0;p=lyx.git

diff --git a/src/Thesaurus.cpp b/src/Thesaurus.cpp
index 42daa0c8eb..b4655ac27f 100644
--- a/src/Thesaurus.cpp
+++ b/src/Thesaurus.cpp
@@ -13,28 +13,32 @@
 
 #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_list.h"
 #include "support/filetools.h"
+#include "support/gettext.h"
 #include "support/lstrings.h"
 #include "support/os.h"
-#include "support/unicode.h"
-
-#include "frontends/alert.h"
 
-#ifdef HAVE_LIBMYTHES
+#ifdef USE_EXTERNAL_MYTHES
 #include MYTHES_H_LOCATION
 #else
-#ifdef HAVE_LIBAIKSAURUS
-#include AIKSAURUS_H_LOCATION
-#endif // HAVE_LIBAIKSAURUS
-#endif // !HAVE_LIBMYTHES
+#include <cstdio>
+#include "support/mythes/mythes.hxx"
+#endif
+
+#include "frontends/alert.h"
 
 #include <algorithm>
 #include <cstring>
+#include <fstream>
 
 using namespace std;
 using namespace lyx::support;
@@ -42,30 +46,12 @@ using namespace lyx::support::os;
 
 namespace lyx {
 
-#ifdef HAVE_LIBMYTHES
-
 namespace {
 
-string const to_iconv_encoding(docstring const & s, string const & encoding)
-{
-	std::vector<char> const encoded =
-		ucs4_to_eightbit(s.data(), s.length(), encoding);
-	return string(encoded.begin(), encoded.end());
-}
-
-
-docstring const from_iconv_encoding(string const & s, string const & encoding)
-{
-	std::vector<char_type> const ucs4 =
-		eightbit_to_ucs4(s.data(), s.length(), encoding);
-	return docstring(ucs4.begin(), ucs4.end());
-}
-
 typedef std::map<docstring, MyThes *> Thesauri;
 
 } // namespace anon
 
-
 struct Thesaurus::Private
 {
 	~Private()
@@ -87,47 +73,104 @@ struct Thesaurus::Private
 		return false;
 	}
 
+	///
+	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::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;
+	/// the location below system/user directory
+	/// there the data+idx files lookup will happen
+	const string dataDirectory(void) { return "thes"; }
+
+};
 
-	if (thesaurusAvailable(lang))
-		return true;
 
-	FileNameList const idx_files = FileName(thes_path).dirList("idx");
-	FileNameList const data_files = FileName(thes_path).dirList("dat");
+pair<string,string> 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;
-
-	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();
+	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))) {
+			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;
-			}
 		}
-
-	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();
+	}
+	if (idx.empty())
+		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);
+}
+
+
+pair<string,string> Thesaurus::Private::getThesaurus(docstring const & lang)
+{
+	string const thes_path = external_path(lyxrc.thesaurusdir_path);
+	pair<string,string> result ;
+
+	if (thesaurusAvailable(lang))
+		return make_pair(string(), string());
+
+	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::Private::addThesaurus(docstring const & lang)
+{
+	if (thesaurusAvailable(lang))
+		return true;
+
+	ThesFiles files = getThesaurus(lang);
+	string const idx = files.first;
+	string const data = files.second;
 
 	if (idx.empty() || data.empty())
 		return false;
@@ -145,17 +188,29 @@ bool Thesaurus::thesaurusAvailable(docstring const & lang) const
 }
 
 
-Thesaurus::Meanings Thesaurus::lookup(docstring const & t, docstring const & lang)
+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(WordLangTuple const & wl)
 {
 	Meanings meanings;
 	MyThes * mythes = 0;
 
-	if (!d->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 = d->thes_.begin();
 	     it != d->thes_.end(); ++it) {
-		if (it->first == lang) {
+		if (it->first == lang_code) {
 			mythes = it->second;
 			break;
 		}
@@ -170,24 +225,38 @@ Thesaurus::Meanings Thesaurus::lookup(docstring const & t, docstring const & lan
 	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, "- ");
 		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);
@@ -199,101 +268,9 @@ Thesaurus::Meanings Thesaurus::lookup(docstring const & t, docstring const & lan
 	return meanings;
 }
 
-#else // HAVE_LIBMYTHES
-#ifdef HAVE_LIBAIKSAURUS
-
-struct Thesaurus::Private
-{
-	Private(): thes_(new Aiksaurus) {}
-	Aiksaurus * thes_;
-};
-
-Thesaurus::Meanings Thesaurus::lookup(docstring const & t, docstring const &)
-{
-	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(d->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;
-		}
-		return meanings;
-	}
-	if (!d->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 = d->thes_->next(cur_meaning);
-
-	while (!ret.empty()) {
-		if (cur_meaning != prev_meaning) {
-			meaning = from_ascii(ret);
-			ret = d->thes_->next(cur_meaning);
-			prev_meaning = cur_meaning;
-		} else {
-			if (ret != text)
-				meanings[meaning].push_back(from_ascii(ret));
-		}
-
-		ret = d->thes_->next(cur_meaning);
-	}
-
-	for (Meanings::iterator it = meanings.begin();
-	     it != meanings.end(); ++it)
-		sort(it->second.begin(), it->second.end());
-
-	return meanings;
-}
-
-
-bool Thesaurus::thesaurusAvailable(docstring const & lang) const
-{
-	// we support English only
-	return prefixIs(lang, from_ascii("en_"));
-}
-
-#else // HAVE_LIBAIKSAURUS
-
-struct Thesaurus::Private
-{
-};
-
-
-Thesaurus::Meanings Thesaurus::lookup(docstring const &, docstring const &)
-{
-	return Meanings();
-}
-
-
-bool Thesaurus::thesaurusAvailable(docstring const &) const
-{
-	return false;
-}
-
-#endif // HAVE_LIBAIKSAURUS
-#endif // HAVE_LIBMYTHES
 
-Thesaurus::Thesaurus() : d(new Thesaurus::Private)
+Thesaurus::Thesaurus()
+	: d(new Thesaurus::Private)
 {
 }