X-Git-Url: https://git.lyx.org/gitweb/?a=blobdiff_plain;f=src%2Fchset.C;h=ed8ee462b502203144b737d0757b80bbdab36333;hb=e28331ed63062dea10d0a21b9ec12034b4b17b9a;hp=60c935e41fc4b39ea1d431338f78299369ed1198;hpb=a320e34382a466574fe174da867a7af21d1fe87f;p=lyx.git

diff --git a/src/chset.C b/src/chset.C
index 60c935e41f..ed8ee462b5 100644
--- a/src/chset.C
+++ b/src/chset.C
@@ -1,50 +1,78 @@
+/**
+ * \file chset.C
+ * This file is part of LyX, the document processor.
+ * Licence details can be found in the file COPYING.
+ *
+ * \author Lars Gullik Bjønnes
+ * \author Jean-Marc Lasgouttes
+ *
+ * Full author contact details are available in file CREDITS.
+ */
+
 #include <config.h>
 
-#ifdef __GNUG__
-#pragma implementation
-#endif
+#include "chset.h"
+
+#include "debug.h"
+
+#include "support/convert.h"
+#include "support/filetools.h"
+
+#include <boost/regex.hpp>
 
 #include <fstream>
+
+
+namespace lyx {
+
+using support::libFileSearch;
+
+using boost::regex;
+using boost::smatch;
+
+using std::endl;
+using std::getline;
+using std::make_pair;
+
 using std::ifstream;
+using std::pair;
+using std::string;
 
-#include "chset.h"
-#include "support/filetools.h"
-#include "support/LRegex.h"
-#include "support/LSubstring.h"
-#include "debug.h"
 
 bool CharacterSet::loadFile(string const & fname)
 {
 	map_.clear();
-	name_.clear();
-	
-	if (fname.empty() || fname == "ascii") 
-		return true;	// ascii 7-bit
-	
+	name_.erase();
+
+	// ascii 7-bit
+	if (fname.empty() || fname == "ascii")
+		return true;
+
 	// open definition file
 	lyxerr[Debug::KBMAP]
 		<< "Reading character set file " << fname << ".cdef" << endl;
-	string filename = LibFileSearch("kbd", fname.c_str(), "cdef");
-	ifstream ifs(filename.c_str());
+	support::FileName const filename = libFileSearch("kbd", fname, "cdef");
+	ifstream ifs(filename.toFilesystemEncoding().c_str());
 	if (!ifs) {
 		lyxerr << "Unable to open character set file" << endl;
 		return true;		// no definition, use 7-bit ascii
 	}
 	name_ = fname;
-	
-	string str;
-	int n;
+
 	string line;
-	LRegex reg("^([12][0-9][0-9])[ \t]+\"([^ ]+)\".*");
-	while(getline(ifs, line)) {
-		if (reg.exact_match(line)) {
-			LRegex::SubMatches const & sub = reg.exec(line);
-			n = atoi(line.substr(sub[1].first,
-					     sub[1].second).c_str());
-			str = LSubstring(line, sub[2].first, sub[2].second);
+	// Ok, I'll be the first to admit that this is probably not
+	// the fastest way to parse the cdef files, but I though it
+	// was a bit neat. Anyway it is wrong to use the lyxlex parse
+	// without the use of a keyword table.
+	regex reg("^([12][0-9][0-9])[ \t]+\"([^ ]+)\".*");
+	while (getline(ifs, line)) {
+		smatch sub;
+		if (regex_match(line, sub, reg)) {
+			int const n = convert<int>(sub.str(1));
+			string const str = sub.str(2);
 			if (lyxerr.debugging(Debug::KBMAP))
 				lyxerr << "Chardef: " << n
-				       << " to [" << str << "]" << endl;
+				       << " to [" << str << ']' << endl;
 			map_[str] = n;
 		}
 	}
@@ -53,11 +81,20 @@ bool CharacterSet::loadFile(string const & fname)
 }
 
 
-bool CharacterSet::encodeString(string & str) const
+pair<bool, int> const CharacterSet::encodeString(string const & str) const
 {
+	lyxerr[Debug::KBMAP] << "Checking if we know [" << str << ']' << endl;
+	bool ret = false;
+	int val = 0;
 	Cdef::const_iterator cit = map_.find(str);
-	if (cit != map_.end()) return true;
-	return false;
+	if (cit != map_.end()) {
+		ret =  true;
+		val = cit->second;
+	}
+	lyxerr[Debug::KBMAP] << "   "
+			     << (ret ? "yes we" : "no we don't")
+			     <<  " know [" << str << ']' << endl;
+	return make_pair(ret, val);
 }
 
 
@@ -65,3 +102,6 @@ string const & CharacterSet::getName() const
 {
 	return name_;
 }
+
+
+} // namespace lyx