Re: XML status

[Chris Burdess]

Mark Wielaard wrote:
> > I have updated the XSLT conformance page
> > (http://www.zen60362.zen.co.uk/xslt/) with the results for current
> > Classpath HEAD. There is a slight regression (8 test cases), I'm 
> > afraid
> > - I don't know why this is but am looking into it.
>
> Which failures are regressions?

Regressions:

axes_axes01
axes_axes127
copy_copy36
expression_expression02
idkey_idkey18
impincl_impincl16
match_match10
match_match15
match_match16
match_match17
match_match18
namespace_namespace08
node_node01
numbering_numbering01
position_position70
position_position73
position_position78
position_position79
position_position93
select_select15
select_select16
select_select84
sort_sort27
sort_sort28
whitespace_whitespace02
whitespace_whitespace03
whitespace_whitespace05
whitespace_whitespace12

Progressions:

axes_axes86
copy_copy10
copy_copy11
copy_copy42
copy_copy50
dlft_dlft02
dlft_dlft03
dlft_dlft04
idkey_idkey29
position_position32
position_position33
position_position34
predicate_predicate06
predicate_predicate12
predicate_predicate20
predicate_predicate43
select_select50
select_select79
string_string14
string_string107
string_string122

> Did you figure out why they are happening?

I synchronized the non-transform classes in my test tree with the 
Classpath versions prior to running this last test suite. I believe the 
changes are due to the prefix/namespace changes introduced in Classpath 
to fix some bugs in Kaffe a few months ago.

> Do you feel these are showstoppers for the next snapshot release?

No. The vast majority of these tests are designed to catch errors on 
the extreme edge of what is possible with XSLT/XPath. A less than 0.5% 
regression is not really significant.
--
Chris Burdess