another rev of text parsing.
diff --git a/tinyxml2.cpp b/tinyxml2.cpp
index cfe2969..ae6a747 100644
--- a/tinyxml2.cpp
+++ b/tinyxml2.cpp
@@ -7,6 +7,12 @@
using namespace tinyxml2;
+static const char LINE_FEED = (char)0x0a; // all line endings are normalized to LF
+static const char LF = LINE_FEED;
+static const char CARRIAGE_RETURN = (char)0x0d; // CR gets filtered out
+static const char CR = CARRIAGE_RETURN;
+
+
// --------- CharBuffer ----------- //
/*static*/ CharBuffer* CharBuffer::Construct( const char* in )
{
@@ -91,25 +97,54 @@
{
TIXMLASSERT( endTag && *endTag );
- char* start = SkipWhiteSpace( p );
- if ( !start )
- return 0;
+ char* start = p;
+ char* q = p; // q (target) <= p (src) in same buffer.
+ char endChar = *endTag;
+ int length = strlen( endTag );
+ char* nextTag = 0;
- char endChar = *endTag;
- p = start;
- int length = strlen( endTag );
-
+ // Inner loop of text parsing.
while ( *p ) {
- if ( *p == endChar ) {
- if ( strncmp( p, endTag, length ) == 0 ) {
- *p = 0;
- *next = p + length;
- return start;
- }
+ if ( *p == endChar && strncmp( p, endTag, length ) == 0 ) {
+ *q = 0;
+ nextTag = p + length;
+ break;
}
- ++p;
+ else if ( *p == CR ) {
+ // CR-LF pair becomes LF
+ // CR alone becomes LF
+ // LF-CR becomes LF
+ if ( *(p+1) == LF ) {
+ p += 2;
+ }
+ else {
+ ++p;
+ }
+ *q = LF;
+ }
+ else if ( *p == LF ) {
+ if ( *(p+1) == CR ) {
+ p += 2;
+ }
+ else {
+ ++p;
+ }
+ *q = LF;
+ }
+ else {
+ *q = *p;
+ ++p;
+ }
+ ++q;
}
- return 0;
+
+ // Error? If we don't have a text tag, something went wrong. (Although
+ // what the nextTag points at may be null.)
+ if ( nextTag == 0 ) {
+ return 0;
+ }
+ *next = nextTag;
+ return start;
}
@@ -129,7 +164,7 @@
void XMLComment::Print( FILE* fp, int depth )
{
XMLNode::Print( fp, depth );
- fprintf( fp, "<!-- %s -->\n", value );
+ fprintf( fp, "<!--%s-->\n", value );
}
diff --git a/tinyxml2.h b/tinyxml2.h
index 2ff58c9..23f21a5 100644
--- a/tinyxml2.h
+++ b/tinyxml2.h
@@ -68,13 +68,13 @@
}
return false;
}
+ inline static int IsUTF8Continuation( char p ) { return p & 0x80; }
/* Parses text. (Not a text node.)
- [ ] EOL normalization.
- - [x] Trim leading whitespace
- - [ ] Trim trailing whitespace.
- - [ ] Leaves inner whitespace
- - [ ] Inserts one space between lines.
+ - [X] Do not trim leading whitespace
+ - [X] Do not trim trailing whitespace.
+ - [X] Leaves inner whitespace
*/
const char* ParseText( char* in, const char* endTag, char** next );
diff --git a/tinyxml2.suo b/tinyxml2.suo
index 4fe777e..0326812 100644
--- a/tinyxml2.suo
+++ b/tinyxml2.suo
Binary files differ
diff --git a/xmltest.cpp b/xmltest.cpp
index 8df6ce0..c3f1b3b 100644
--- a/xmltest.cpp
+++ b/xmltest.cpp
@@ -7,6 +7,7 @@
int main( int argc, const char* argv )
{
+#if 0
{
static const char* test = "<!--hello world-->";
@@ -14,14 +15,14 @@
doc.Parse( test );
doc.Print( stdout );
}
- /*
+#endif
{
- static const char* test = "<hello></hello>";
+ static const char* test = "<!--hello world\n"
+ " line 2\r-->";
XMLDocument doc;
doc.Parse( test );
doc.Print( stdout );
}
- */
return 0;
}
\ No newline at end of file