X-Git-Url: https://git.lyx.org/gitweb/?a=blobdiff_plain;f=src%2FThesaurus.cpp;h=1c1adc03b0e22cb91154ac21e396957b05997bb3;hb=210ea66d2d1113cddcc1b0f3547c1063b7dc4d47;hp=dea7c05e3d5d3a4db3820404a1263b248326852d;hpb=62ca7f3ae55ad2e0c395cb554d71afab87de1ee3;p=lyx.git

diff --git a/src/Thesaurus.cpp b/src/Thesaurus.cpp
index dea7c05e3d..1c1adc03b0 100644
--- a/src/Thesaurus.cpp
+++ b/src/Thesaurus.cpp
@@ -15,19 +15,24 @@
 
 #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/gettext.h"
 #include "support/lstrings.h"
 #include "support/os.h"
 
-#include "support/mythes/mythes.hxx"
-
-#include "frontends/alert.h"
+#include <cstdio>
+#include MYTHES_H_LOCATION
 
 #include <algorithm>
 #include <cstring>
+#include <fstream>
 
 using namespace std;
 using namespace lyx::support;
@@ -39,8 +44,7 @@ namespace {
 
 typedef std::map<docstring, MyThes *> Thesauri;
 
-} // namespace anon
-
+} // namespace
 
 struct Thesaurus::Private
 {
@@ -66,54 +70,113 @@ struct Thesaurus::Private
 	///
 	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_;
+
+	/// the location below system/user directory
+	/// there the data+idx files lookup will happen
+	const string dataDirectory(void) { return "thes"; }
+
 };
 
 
-pair<string, string> Thesaurus::Private::getThesaurus(docstring const & lang)
+pair<string,string> Thesaurus::Private::getThesaurus(string const & path, docstring const & lang)
 {
-	string const thes_path = external_path(lyxrc.thesaurusdir_path);
-	LYXERR(Debug::FILES, "thesaurus path: " << thes_path);
-	if (thes_path.empty())
+	FileName base(path);
+	if (!base.isDirectory()) {
 		return make_pair(string(), string());
-
-	if (thesaurusAvailable(lang))
-		return make_pair(string(), string());
-
-	FileNameList const idx_files = FileName(thes_path).dirList("idx");
-	FileNameList const data_files = FileName(thes_path).dirList("dat");
+	}
+	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))) {
+	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;
-			}
 		}
-
-	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))) {
+	}
+	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);
 }
 
 
+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))
@@ -141,22 +204,27 @@ bool Thesaurus::thesaurusAvailable(docstring const & lang) const
 
 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 (!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;
 		}
@@ -166,29 +234,44 @@ Thesaurus::Meanings Thesaurus::lookup(docstring const & t, docstring const & lan
 		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);
@@ -201,7 +284,8 @@ Thesaurus::Meanings Thesaurus::lookup(docstring const & t, docstring const & lan
 }
 
 
-Thesaurus::Thesaurus() : d(new Thesaurus::Private)
+Thesaurus::Thesaurus()
+	: d(new Thesaurus::Private)
 {
 }