jez
2e1b775e71
changed the handling of RelativeLocationPath token when building a step list. There might be things following the path, and must take them into account. Test position89 exposed a problem prompting this change.
2008-05-28 15:08:35 +00:00
jez
53fbf26ff0
Correctly implemented Namespace Nodes.
...
The XPath data model requires that namespace nodes are associated with an element, and sort ahead of attribute nodes in document order. Until now, Arabica's namespace node had no parent, or owner document and so was failing these requirements. The Xalan position111 test case highlights this requirement.
2008-05-28 08:53:21 +00:00
jez
5d7f3364ba
Visual Studio, how I curse your useless warning C4800
2008-05-28 08:45:25 +00:00
jez
f89a2be3cc
fix in MultiplicativeExpr rule for Boost 1.35
2008-05-04 18:10:03 +00:00
jez
261f589f47
2008-05-04 14:41:23 +00:00
jez
84d7d9390a
spelled underlying_impl wrong all these years and never realised, start work on sorting namespace nodes ahead of attribute nodes
2008-04-23 14:41:00 +00:00
jez
c6564e798c
added missing typename
2008-04-23 14:39:11 +00:00
jez
f818f192af
fixed match rewriting when there are leading predicates - http://www.jezuk.co.uk/jez/2008April#3711
2008-04-18 09:11:51 +00:00
jez
a049a72c08
fixed sorting bug when reverse sorting a nodeset which includes the root node
2008-04-17 20:32:32 +00:00
jez
0850f50269
tweaking document order across documents - not for any better reason than making a test case pass
2007-12-25 21:56:28 +00:00
jez
2e6c372933
hacked up a ncname test, although it's not working at the moment and I'm tired and can't think
2007-12-22 22:21:26 +00:00
jez
235f81718d
Added Expression_scanner and scan (a visitor pattern basically) to allow XPathExpression_impls to be examined. This is used in the match rewriting to search for position() and last() function calls.
...
To ease implementing scan, BinaryExpression and UnaryExpression now inherit XPathExpression_impl. Other classes now derived only from Binary|UnaryExpression rather than from XPathExpression_impl as well.
2007-12-21 15:56:04 +00:00
jez
a8ce93f89b
First proper go at rewriting positional matches.
2007-12-21 14:13:36 +00:00
jez
cbc9cf3797
it's late
2007-12-20 23:24:16 +00:00
jez
f87c01fc53
2007-12-20 16:23:56 +00:00
jez
cdac22542e
added type() to XPath expressions. I need this to allow XSLT match patterns to be rewritten properly - see http://www.jezuk.co.uk/cgi-bin/view/arabica/log?id=3546 - but it will also be useful for compile time checking (should I get sufficiently keen :)
2007-12-18 23:03:16 +00:00
jez
99a5d5cb85
fixed local-name for namespace nodes
2007-11-10 22:23:36 +00:00
jez
a52aafebd5
In XPath node() matches any node of any type. In an XSLT match pattern, node() matches everything except attributes and the document root node. Tweaked that match pattern grammar to introduce a new token in the AST.
2007-10-26 23:24:58 +00:00
jez
229a859c91
whitespace change
2007-10-25 20:58:27 +00:00
jez
af3ebfde60
s/XPathExpressionPtr/XPathExpression/g
2007-10-25 20:42:00 +00:00
jez
a7d050841d
s/XPathExpression/XPathExpression_impl/
2007-10-23 21:37:24 +00:00
jez
fac7cf88de
XPathExpressionPtr no longer derives from boost::shared_ptr, it contains it instead
2007-10-22 20:25:35 +00:00
jez
f6b2ed4966
OK, I think that's it. XPathValue is now your man. Existing code using XPathValuePtr should still work, but new stuff should use XPathValue
2007-10-22 17:42:50 +00:00
jez
6e2f594dc9
more reworking on XPathValuePtr -> XPathValue. XPathValue contains rather than derives from boost::shared_ptr. Added operator== so can still compare to 0 to see if the value is null or not.
2007-10-22 14:10:49 +00:00
jez
9b32ed13e4
Some time ago, it was gently suggested to me that XPathValuePtr and XPathExpressionPtr both exposed an implementation detail, because they derive fromboost::shared_ptr, and provided an interface that was inconsisted with the DOM classes, because you accessed the member functions via -> rather than .
...
At the time, I was just pleased to have got the XPath stuff done and wasn't really fussed, so I left it. Since then though, it's niggled and niggled away at the back of my mind and now I've decided to do something about it.
XPathValuePtr will become XPathValue, with the member functions accessed through the . operator. The XPathValuePtr name and -> member access will be retained for the meantime, so that existing code won't be broken. XPathExpressionPtr will be similarly changed.
This commit is the first bit of that work, now I've satisfied myself it's going to be pretty easy so long as I pay proper attention.
2007-10-19 21:59:24 +00:00
jez
3f156c9fcf
reworked MatchExpr so that it doesn't expose its inner XPathExpression
2007-10-14 20:06:27 +00:00
jez
3d5c7995cb
added value_type typedef as per Evan Burkitt's suggestion
2007-10-05 14:56:26 +00:00
jez
c3bffa005a
renamed include/Utils to include/Arabica
2007-09-10 17:52:04 +00:00
jez
17248585af
more namespace and file moving shenanigans
2007-09-10 17:24:17 +00:00
jez
63ed794d58
finished adding string adaptor param to DOM. everything builds and runs now
2007-09-08 22:31:24 +00:00
jez
7b43037a37
Added string_adaptor parameter to DOM classes. DOM and XPath tests compile and run, but still need to add defaults so more "normal" bits of code will go.
2007-09-07 23:52:30 +00:00
jez
7d46f6f4e2
Moved DOM:: into Arabica::DOM::
2007-09-05 11:47:13 +00:00
jez
7cf3897061
renamed all .h to .hpp
2007-09-04 22:55:47 +00:00
jez
dc6a3ad454
fixed #includes
2007-08-22 12:22:09 +00:00
jez
645247287d
see http://www.jezuk.co.uk/cgi-bin/view/SAX/news?id=3285
2007-08-07 14:43:44 +00:00
jez
b95e3afeb2
merged with mangle-dev branch
2007-07-19 17:01:31 +00:00
jez
543f7357df
started moving headers out into an include directory
2006-08-07 21:08:11 +00:00
