[xml] Debugging PE identification regression

[Gary Pennington <Gary Pennington uk sun com>]

Hi,

Found the bug in PE identification. Or to put it another way, that
change is a good enough fix for me. I'm not sure that it's definitive,
but as far as I can tell it is.

diff

calvin:LIB_XML_BUILD/libxml2-2.3.3:$ diff -c
/packages/LIB_XML_BUILD/libxml2-2.3.3/parser.c parser.c
*** /packages/LIB_XML_BUILD/libxml2-2.3.3/parser.c      Mon Feb 26
03:35:06 2001
--- parser.c    Wed Mar  7 18:32:27 2001
***************
*** 1641,1647 ****
            return(ret);
        }
      }
!     xmlParseNameComplex(ctxt);
  }

  xmlChar *
--- 1641,1647 ----
            return(ret);
        }
      }
!     return (xmlParseNameComplex(ctxt));
  }

  xmlChar *

It was that one line that was causing all the problems.

I don't think that your change that you submitted earlier in the week
will make any difference to this problem, since the change above is
against plain old vanilla 2.3.3 as you shipped it. I don't know if that
means you want to pack out that change or if you think that's needed to
fix some other possible category of problems.

Gary