* configure.in: fixed configure for MPE/iX from Markus Henke
* xmlmemory.c: fixed initialization problems
* xpath.c: another set of patches from Richard Jinks this
fixes "make XPathtests" on linux
Daniel
* trionan.c trionan.h xpath.c: more patches from Richard Jinks
* test/XPath/expr/compare test/XPath/expr/equality
test/XPath/expr/floats test/XPath/expr/functions
test/XPath/expr/strings result/XPath/expr/compare
result/XPath/expr/equality result/XPath/expr/floats
result/XPath/expr/functions result/XPath/expr/strings: Updated
tests though they show a divergence on Linux
Daniel
* AUTHORS HACKING: Added Igor Zlatkovic as official maintainer
* python/Makefile.am python/tests/Makefile.am: Albert Chin pointed
that $(datadir) should be used for docs
Daniel
* xpath.c: fixing #75619, related to a problem when trying
to evaluate condition when the current node set resulting
from that sub-step evaluation is empty. Also fixes 2 potential
problem with previous-sibling and next-siblings axis.
Daniel
* xpath.c: pretty insane thing, the xmlXPathFormatNumber()
was not serializing 1 as "1" if LC_ALL=sv_SE :-( and in the
context of ScrollKeeper, made sure that if the number is
an integer, the serialization follows the description at
http://www.w3.org/TR/xpath#section-String-Functions
Daniel
* configure.in: preparing 2.4.18
* doc/*: updated and rebuilt the web site
* *.c libxml.h: implement the new IN_LIBXML scheme discussed with
the Windows and Cygwin maintainers.
* parser.c: humm, changed the way the SAX parser work when
xmlSubstituteEntitiesDefault(1) is set, it will then
do the entity registration and loading by itself in case the
user provided SAX getEntity() returns NULL.
* testSAX.c: added --noent to test the behaviour.
Daniel
* xpath.c: small fix to avoid potential problem due to
ordering of freeing data
* python/Makefile.am: people were complaining about
the generated file in python dir not being built
Daniel
* xpath.c: speedup some node selection operations, this can
have a significant impact on DocBook Norm's stylesheets
* nanohttp.c: someone reported that SOCKLEN_T may not be defined
make sure it's always the case
* debugXML.c: distinguish CDATA and comments in ls operations
Daniel
* configure.in: preparing 2.4.17 release
* doc/*: updated and rebuilt the docs
* xpath.c: fixed a comment
* python/libxml.c: fixed a possible reentrancy problem
Daniel
* xpath.c: trying to avoid bug #72150 which was apparently
caused by a gcc bug (or a processor problem) as detailed
at http://veillard.com/gcc.bug
Daniel
* xpath.c: fixing #61290 "namespace nodes have no parent"
long standing divergence from the XPath REC. NodeSets
simply hold a copy of namespace nodes and those node ->next
points to the parent (which may not be the node carrying the
definition).
* include/libxml/xpath.h: flagged but didn't added a possible
speedup
* DOCBparser.c HTMLparser.c: removed some warnings from push
parser due to new state being added.
* tree.c: new fix from Boris Erdmann
* configure.in c14n.c include/libxml/c14n.h testC14N.c: added
the XML Canonalization support from Aleksey Sanin
Daniel
* configure.in include/libxml/xmlwin32version.h: preparing 2.4.16
* doc/* python/libxml2class.txt: updated and rebuilt the docs,
rebuilt the API and web site
* xpath.c: fixed#71978 portability bugs
Daniel
* xpath.c: minor optimization
* python/generator.py python/libxml.c python/libxml.py
python/libxml_wrap.h: more work on the python bindings,
they now support XPath and there is no evident leak
Daniel
* SAX.c: cleanup patch from Anthony Jones
* doc/Makefile.am: fix the headers to avoid in make scan
* parserInternals.c xpath.c include/libxml/*.h: cleanup of the
includes, * vs Ptr and general cleanup
* parsedecl.py: first version of a script to extract the
module interfaces, the goal will be to provide .decl or XML
specification of the interfaces to build wrappers.
Daniel
* HTMLtree.c catalog.c debugXML.c entities.c nanoftp.c
parser.c valid.c xmlmemory.c xpath.c xpointer.c: started
integrating the non-controversial parts of Gary Pennington
multithread patches
* catalog.c: corrected a small bug introduced
Daniel
* configure.in: preparing 2.4.6 release
* doc/xml.html doc/html/*: updated and rebuilt the docs
* include/libxml/*.h *.c: fixed a number of teh/the widht/width typos
Daniel
* uri.c: applied fix from Mathias Hasselmann about a bug in URI
parsing.
* xpath.c: fix bug #61291 the default XML namespace node is
missing from the namespace axis.
* tree.c: refuse to create namespaces nodes with prefix "xml"
Daniel
* xpath.c: hum, restrict the integer usage gcc bug workaround
to only gcc compilers so that other architecture don't get
penalized by this limitation.
* include/libxml/xpath.h: small typo fix from Heiko W. Rupp
Daniel
* Makefile.am catalog.c xmlcatalog.c include/libxml/catalog.h:
more work on the XML catalog support.
* parser.c include/libxml/parser.h: small cleanup seems using
list as a public parameter name can give portability troubles
* trionan.c trionan.h xpath.c include/libxml/trionan.h
include/libxml/xpath.h include/libxml/Makefile.am: removed
trionan from the libxml API, added xmlXPathIsInf and xmlXPathIsNaN
wrappers
Daniel
* xpath.c: serious changes on Result Value Trees and NodeSets
w.r.t. deallocation and collect operations. Probably not
100% clean (merge of allocated trees smells like a problem).
Seems sufficient to close#58943
Daniel
* xpath.c: fixed a bug when walking the descendants and
the current node has no children
* debugXML.c: show up when a text node is supposed to not be escaped
Daniel
* xpath.c: fixed a bug in xmlXPathNodeTrailingSorted (for now it
worked like the set:leading() function)
* include/libxml/xpathInternals.h: added xmlXPathNodeSetContains
* xpath.c include/libxml/xpath.h: fixed a serious memory problen
when walking the namespace axis showing up in
libxst/tests/general/bug-12
* xmlmemory.c: added the possibility to trace a given block
defined by its address
Daniel
* parser.c parserInternals.c: fixed the xmlLineNumbersDefault()
errors, lesson don't add new functions at 1am before a release
* xpath.c: integrated fix from Bjorn to avoid divide by zero
from XPath initialization when possible.
Daniel
Cleanup, cleanup ..
* configure.in Makefile.am: removed libxml softlink for good
* include/libxml/*.h *.c doc/Makefile.am: cleanup to get
100% coverage by gtk-doc
Daniel
* include/libxml/xpath{,Internals}.h xpath.c: added a more
convenient extension API for value and context managing
Now handles external objects through xmlXPathPopExternal,
xmlXPathWrapExternal and xmlXPathReturnExternal.
Added functions for sets operations (intersection, etc.)
* SAX.c parser.c testXPath.c xpath.c: trying to fix#56948, this
led to an XPath fix, improvements of SAX initialization, and
an added option --nocdata to testXPath
Daniel
lot of optimization work, results in significant improvements
when handling really complex XPath queries. Add a small optimizer
for unions, improve [n] and [last()], avoid some costly ops.
Daniel
