Commits

Author Commit Message Labels Comments Date
Nick Wellnhofer
Fix potential memory leaks when freeing parse context
Branches
memory-leaks
Nick Wellnhofer
Fix memory leaks in $parser->finish_push
Branches
memory-leaks
Nick Wellnhofer
Add TODO comment about memory leak
Branches
memory-leaks
Nick Wellnhofer
Fix memory leak when push parsing is aborted
Branches
memory-leaks
Nick Wellnhofer
Fix memory leak in _parse_sax_file
Branches
memory-leaks
Nick Wellnhofer
Fix memory leak in $node->setBaseURI
Branches
memory-leaks
Nick Wellnhofer
Fix memory leak in $doc->URI
Branches
memory-leaks
Nick Wellnhofer
Fix memory leak in _parse_string Messing with ctxt->directory causes a memory leak. It's generally a bad idea to access parser internals directly. I'm not sure what this was used for but it doesn't seem to break anything.
Branches
memory-leaks
Nick Wellnhofer
Skip t/40reader_mem_error.t with libxml2 < 2.7.4 The failure is probably due to a known double-free bug in libxml2. https://rt.cpan.org/Ticket/Display.html?id=84564 https://bugzilla.gnome.org/show_bug.cgi?id=447899
Branches
rt84564-v2
Nick Wellnhofer
Close branch rt84564.
Branches
rt84564
Nick Wellnhofer
Skip t/40reader_mem_error.t with libxml2 < 2.7.4 The failure is probably due to a known double-free bug in libxml2. https://rt.cpan.org/Ticket/Display.html?id=84564 https://bugzilla.gnome.org/show_bug.cgi?id=447899
Branches
rt84564
Shlomi Fish
Commit merged
Nick Wellnhofer
Forgot to add test file
Nick Wellnhofer
Make parse_fh more robust wrt encoding layers Throw an exception instead of crashing when a file handle which uses an encoding layer is passed to parse_fh and reads more bytes than requested. Fixes RT #78448: https://rt.cpan.org/Ticket/Display.html?id=78448 A better fix would require to rewrite the LibXML_read_perl interface to allow for variable size results.
Nick Wellnhofer
Simplify Perl/C string conversion libxml2 only uses UTF-8 as internal encoding. Rewrite the string conversion functions to always assume UTF-8 and handle the few cases where charset conversion is needed with new helper functions: * LibXML_encode_string * LibXML_decode_string (not yet implemented) Additionally, the conversion from Perl to C strings now handles Latin1 strings correctly. This fixes RT #53532: https://rt.cpan.org/T…
Branches
string-conversion
Nick Wellnhofer
Fix some text methods to work with UTF-8 insertData, deleteData, replaceData used xmlStrsub and xmlStrlen instead of xmlUTF8Strsub and xmlUTFStrlen.
Nick Wellnhofer
Fix substringData method on text nodes The length argument passed to xmlStrsub was bogus. Furthermore, xmlUTF8Strsub should be used. Fixes RT #88730 https://rt.cpan.org/Ticket/Display.html?id=88730
Shlomi Fish
Fix RT #91800.
Shlomi Fish
Fix Changes
Shlomi Fish
Fix the description of the bug in the Changes. Thanks to Father C again.
Shlomi Fish
Tagging the XML-LibXML release as 2.0110
Shlomi Fish
Bump version
Tags
XML-LibXML-2.0110
Shlomi Fish
=head1 LICENSE to the docs.
Shlomi Fish
MIN_PERL_VERSION.
Shlomi Fish
use strict and use warnings
Shlomi Fish
Tagging the XML-LibXML release as 2.0109
Shlomi Fish
Version
Tags
XML-LibXML-2.0109
Shlomi Fish
Bump ver to 2.0109
Shlomi Fish
Fix markstack handling in BOOT The BOOT section of an XS module is itself an XSUB body, which gets an implicit dXSARGS when compiled. dXSARGS implied POPMARK, which pops the markstack, which keeps track of how many arguments are in the cur- rent list, and where they begin. If we call the XSUB body generated from another BOOT section, we end up popping two marks, one of which does not belong to us, corrupting the perl mark sta…
Shlomi Fish
Tagging the XML-LibXML release as 2.0108
  1. Prev
  2. Next