Commit graph

677 commits

Author SHA1 Message Date
jez_higgins
655a906071 gcc4 fixes 2006-01-05 16:30:31 +00:00
jez_higgins
16a95885ca lots of XSLT compilation and matching tests 2006-01-05 16:28:19 +00:00
jez_higgins
648a89dd9d everything compiled, but to the wrong thing. now compiles to the correct thing, at least as far as my current tests go 2006-01-05 16:25:30 +00:00
jez_higgins
93ebae8f14 added NotAttributeNodeTest 2006-01-05 16:24:11 +00:00
jez_higgins
f07e022936 now compiles compound XSLT patterns of the form test1|test2|test3 ... |testN. that's them all done. I love Spirit 2006-01-05 10:50:01 +00:00
jez_higgins
211ae3d2b5 getAxis now calls itself when it encounters an AxisSpecifier. both switches have been combined as a result, and more XSLT patterns compile without further work 2006-01-04 16:48:23 +00:00
jez_higgins
aa4be03ea0 MinGW patches from Cromwell Enage, spanage@yahoo.com 2006-01-04 10:29:24 +00:00
jez_higgins
54d4ce66f8 for some reason BOOL is now an ambiguous symbol 2005-12-22 00:10:03 +00:00
jez_higgins
1ed2ea7ee8 compiling more XSLT patterns 2005-12-22 00:02:48 +00:00
jez_higgins
409e86f42e added AttributeNodeTest - needed for XSLT patterns 2005-12-21 23:59:43 +00:00
jez_higgins
fab22a313b compiling more matches 2005-12-19 15:55:25 +00:00
jez_higgins
7acb2d622f started evaluating matches 2005-12-19 15:55:06 +00:00
jez_higgins
57a07ca171 added assertFalse 2005-12-19 15:52:26 +00:00
jez_higgins
8b7f3196d8 Started try to compile XSLT patterns 2005-12-18 22:22:13 +00:00
jez_higgins
1c9175c98b don't need to keep '::' in ChildOrAttributeAxisSpecifier 2005-12-18 20:17:33 +00:00
jez_higgins
9485917c42 extend grammar for XSLT matches - added some parse tests 2005-12-16 19:09:57 +00:00
jez_higgins
be0dcae896 extend grammar for XSLT matches 2005-12-16 19:09:18 +00:00
jez_higgins
e348364a82 make gcc not complain about initialisation order 2005-12-16 19:08:46 +00:00
jez_higgins
ec7d5de834 added missed case for node() 2005-12-16 10:17:39 +00:00
jez_higgins
2238256e67 added extra patterns needed for xsl:template match attributes 2005-12-15 22:11:16 +00:00
jez_higgins
d89f09dd9a Now compiles for std::string and std::wstring, and probably for anything
else that has a size() member function and an operator[]
Will need to be reworked at some point in the future, but should do for now
2005-12-15 19:48:19 +00:00
jez_higgins
242e43b195 uncovered a missed case 2005-12-15 12:27:48 +00:00
jez_higgins
071e62b2dd Factored out the XML escaping stuff 2005-12-14 14:49:57 +00:00
jez_higgins
f8f3af0a76 From: Isak Johnsson, isak@hypergene.com
gcc cannot be used as the linker for C++ on Mac OS X
see http://www.cocoabuilder.com/archive/message/xcode/2005/11/21/1356
use g++ instead of gcc
2005-12-13 14:08:05 +00:00
jez_higgins
8ecc72511f getAttribute now returns a const string_type& 2005-12-09 16:01:31 +00:00
jez_higgins
e6099cc42f getData now returns a const string_type& 2005-12-09 15:57:21 +00:00
jez_higgins
7a32af4507 getTarget now returns a const string_type& 2005-12-09 15:55:29 +00:00
jez_higgins
affb66609e getValue now returns a const string_type& 2005-12-09 15:54:23 +00:00
jez_higgins
427d8b64f2 getName now returns a const string_type& 2005-12-09 15:53:05 +00:00
jez_higgins
eebbdbf121 getNodeValue, getNamespaceURI, getPrefix and getValue all now return const string_type& instead of string_type 2005-12-09 15:49:52 +00:00
jez_higgins
99d08a5fd4 corrected prefix handling 2005-12-09 15:20:08 +00:00
jez_higgins
3bc05eafdf getNodeName now returns const stringT& 2005-12-09 14:17:53 +00:00
jez_higgins
d21c753efd getNodeName now returns const stringT& 2005-12-09 14:09:32 +00:00
jez_higgins
756e0f6876 getTagName now returns const stringT& 2005-12-09 14:09:08 +00:00
jez_higgins
ba05fced6e types: removed private destructor to quiet a GCC warning. types isn't actually instantiated anywhere 2005-12-09 12:04:28 +00:00
jez_higgins
9fc4a8b000 Element & attribute names and namespace URIs are now held in a
string pool attached the owning document.
The string pool is just a std::list, because it gives us stable pointers
into the contents.  Easy!
2005-12-09 11:49:03 +00:00
jez_higgins
03661daaf8 reverted to runtime_error to bad_cast because that's what it is 2005-12-09 11:46:53 +00:00
jez_higgins
116385f350 Borland workarounds submitted by Eric Britz 2005-12-07 16:09:37 +00:00
jez_higgins
bc15f5eedf builds with Cygwin 2005-12-07 11:23:44 +00:00
jez_higgins
2364434694 Rather alarmingly, this won't build is language extensions are turned off, even though everything appears to be correct 2005-12-07 10:52:48 +00:00
jez_higgins
35acd2eea7 Added third template parameter to wrappers, together with a bit of
meta-programming magic so that everything builds through with MSXML.

See http://www.jezuk.co.uk/cgi-bin/view/SAX/news?id=2713 for details of
the problem.

This may break VC6, but I haven't checked
2005-12-07 10:51:00 +00:00
jez_higgins
a36a7adde7 identical to utf16utf8codecvt but better named 2005-12-07 10:48:15 +00:00
jez_higgins
2563eee75c added DualMode 2005-12-06 11:20:42 +00:00
jez_higgins
80617d400a increased buffer sizes 2005-12-02 17:04:56 +00:00
jez_higgins
7bcc01a51e Initial work on a dual tree/streaming mode DOM Parser
This work is based on something worked up by Terris Linenbach.  The callback
interface will be extended, but I'm kind of feeling my way with this.
Comments welcome.
2005-12-01 16:13:42 +00:00
jez_higgins
9e418db8e2 include file fixes 2005-11-28 12:40:20 +00:00
jez_higgins
77e654d41b Cygwin fixes 2005-11-28 12:39:13 +00:00
jez_higgins
96c6da64c4 implement Node::purgeChild 2005-11-25 22:39:16 +00:00
jez_higgins
2b051fc1ff gcc 4.0.2 fixes
Use ARABICA_NO_WCHAR_T to exclude the bits that don't build under Cygwin
2005-11-25 15:54:08 +00:00
jez_higgins
b312078ac5 Corrected so they actually build the right things :) 2005-11-25 15:53:28 +00:00