Changed quoted strings handling. Also not missing any doxygen commands now.

git-svn-id: https://swig.svn.sourceforge.net/svnroot/swig/branches/gsoc2012-doxygen@13563 626c5289-ae23-0410-ae9c-e8d60b6d4f22

Examples/test-suite/python/doxygen_translate_all_tags_runme.py

@@ -21,7 +21,7 @@ check(doxygen_translate_all_tags.function.__doc__, ''
 ''
 '  SomeLatexIndex '
 ''
-'  someGroupSome title '
+'  someGroup"Some title" '
 ''
 '  theAnchor'
 ''
@@ -84,8 +84,8 @@ check(doxygen_translate_all_tags.function.__doc__, ''
 '  digraph example { '
 '  node [shape=record, fontname=Helvetica, fontsize=10]; '
 ''
-'  b [ label="class B" URL="\ref B"]; '
+'  b [ label="class B" URL=" B"]; '
-'  c [ label="class C" URL="\ref C"]; '
+'  c [ label="class C" URL=" C"]; '
 '  b -> c [ arrowhead="open", style="dashed" ]; '
 '  } '
 ''
@@ -171,9 +171,9 @@ check(doxygen_translate_all_tags.function.__doc__, ''
 '  someThing'
 ''
 '  Sender,Receiver; '
-'  Sender->Receiver [label="Command()", URL="\ref '
+'  Sender->Receiver [label="Command()", URL=" '
 '  Receiver::Command()"]; '
-'  Sender<-Receiver [label="Ack()", URL="\ref '
+'  Sender<-Receiver [label="Ack()", URL=" '
 '  Ack()", ID="1"]; '
 ''
 '  mscFile.mscThe caption'
@@ -312,6 +312,7 @@ check(doxygen_translate_all_tags.function.__doc__, ''
 ''
 '  This will only appear in XML '
 ''
+'  todo"Todo"Todo List'
 ''
 '  Here goes test of symbols: '
 '  $@\&~<>#%".::'

Source/DoxygenTranslator/src/DoxygenParser.cpp

@@ -106,7 +106,23 @@ int DoxygenParser::commandBelongs(std::string theCommand) {
 std::string DoxygenParser::getNextWord(TokenList & tokList) {
 	Token nextToken = tokList.peek();
 	if (nextToken.tokenType == PLAINSTRING) {
+		// handle quoted strings as words
+		if (nextToken.tokenString[0] == '"'
+				&& nextToken.tokenString[nextToken.tokenString.size() - 1] != '"') {
+			string word = nextToken.tokenString + " ";
 			nextToken = tokList.next();
+			while (true) {
+				string nextWord = getNextWord(tokList);
+				if (!nextWord.size()) // maybe report unterminated string error
+					return word;
+				word += nextWord;
+				if (word[word.size() - 1] == '"') // strip quotes
+					return word.substr(1, word.size() - 2);
+				word += " ";
+			}
+		}
+
+		tokList.next();
 		return nextToken.tokenString;
 	}
 	return "";
@@ -120,7 +136,6 @@ std::list < Token >::iterator DoxygenParser::getOneLine(TokenList & tokList) {
 			//endOfLine++;
 			return endOfLine;
 		}
-		//cout << (* endOfLine).toString();
 		endOfLine++;
 	}
 
@@ -390,6 +405,7 @@ int DoxygenParser::addCommandUnique(std::string theCommand, TokenList & tokList,
 		aNewList.push_front(DoxygenEntity("plainstd::string", title));
 		aNewList.push_front(DoxygenEntity("plainstd::string", heading));
 		aNewList.push_front(DoxygenEntity("plainstd::string", key));
+		doxyList.push_back(DoxygenEntity(theCommand, aNewList));
 		return 1;
 	}
 	// \ingroup (<groupname> [<groupname> <groupname>])

Source/DoxygenTranslator/src/TokenList.cpp

@@ -28,20 +28,25 @@ TokenList TokenList::tokenizeDoxygenComment(const std::string &doxygenComment, c
 
   bool isPlainString = false;
   string::size_type pos, lastPos = 0;
+  char prevChar = doxygenComment[lastPos];
   string currentWord;
   while (true) {
     isPlainString = false;
-    pos = doxygenComment.find_first_of("\t\n ", lastPos);
+    pos = doxygenComment.find_first_of("\\@\t\n ", lastPos);
     if (pos == string::npos)
       pos = doxygenComment.size();
-    if (pos > lastPos) {
+
     currentWord = doxygenComment.substr(lastPos, pos-lastPos);
-      if (currentWord[0] == '\\' || currentWord[0] == '@') {
+    if (prevChar == '\n')
+      tokList.m_tokenList.push_back(Token(END_LINE, "\n"));
+    else if (prevChar == '\\' || prevChar == '@') {
       // it's a doxygen command
-        currentWord = currentWord.substr(1, currentWord.length() - 1);
+      // hack to get commands like \\ or \@ or @\ or @@
+      if (doxygenComment[pos] == '@' || doxygenComment[pos] == '\\')
+        currentWord += doxygenComment[pos];
       tokList.m_tokenList.push_back(Token(COMMAND, currentWord));
     }
-      else if (currentWord[0] == '!' || currentWord[0] == '*' || currentWord[0] == '/') {
+    else if (currentWord.size() && (currentWord[0] == '!' || currentWord[0] == '*' || currentWord[0] == '/')) {
       // check if it's one of the '!!!', '***', '///' of any length
       char c = currentWord[0];
       isPlainString = false;
@@ -54,31 +59,13 @@ TokenList TokenList::tokenizeDoxygenComment(const std::string &doxygenComment, c
     else
       isPlainString = true;
 
-      if (isPlainString) {
+    if (isPlainString && currentWord.size())
-        // handle quoted words
-        if (currentWord[0] == '"') {
-          pos = doxygenComment.find_first_of("\"\n", lastPos + 1);
-          if (pos == string::npos)
-            pos = doxygenComment.size();
-          if (pos == doxygenComment.size() || doxygenComment[pos] != '"') {
-            // hack to print error message before the list is constructed
-            tokList.m_tokenListIter = tokList.m_tokenList.end();
-            tokList.printListError("Unterminated string");
-          }
-          // also strip quotes
-          currentWord = doxygenComment.substr(lastPos + 1, pos-lastPos-1);
-        }
-        if (currentWord.size())
       tokList.m_tokenList.push_back(Token(PLAINSTRING, currentWord));
-      }
-    }
 
+    prevChar = doxygenComment[pos];
     lastPos = pos + 1;
     if (lastPos >= doxygenComment.size())
       break;
-
-    if (doxygenComment[pos] == '\n')
-      tokList.m_tokenList.push_back(Token(END_LINE, "\n"));
   }
   tokList.m_tokenListIter = tokList.m_tokenList.begin();
   return tokList;