Bug: Bad parsing after buffer

[ted]

I found this weird bug that occurs only sometimes and finally tracked it down.

Here's some code, which is pretty standard load file into memory and parse into tinyXml.

  char* buf = LoadFile( filename );
  TiXmlDocument xmlDoc;
  xmlDoc.Parse( buf );
  if ( xmlDoc.Error() )
  {
     // Parsing failed
     return;
  }

Sometimes I get a fail and sometimes I get a success.  And the reason I get a fail is because of this:

Let's say the file is 100 bytes.

Memory:
buf:  0x00000000   [100 bytes of the contents of the file]
addr: 0x00000064   [random data in the memory]

By chance, I parsed another file right before this one and the [random data in the memory] looked like this:  [<TagFromOldFile>...].  I traced through the Parse() function and it errored because it couldn't find the end tag, </TagFromOldFile>.

Is this a known bug that TinyXml will continue on with it's parsing, even after the top most tag has been matched with it's end tag?

Thanks,
Ted

[ted]

I found the bug.  The 100 bytes of contents of the file should be null terminated so buf's size should be 101 and buf[100] = '\0';

Thanks,
Ted