After all that has been said about xpath 2.0, I'm not sure
which is more
readable...
Using <http://nicexsl.sourceforge.net> resolves the quotes and brackets
levels of readability:
(After automated XSLT to NiceXSL translation, I had to manually prettify the
XPaths,
since whitespace is preserved in converted XPaths.)
stylesheet version="2.0" {
match(/) {
variable spanspec := /table/tgroup/spanspec;
<!-- constucts a sequence of the two colspecs, note the 'as' -->
variable first : element()+ {
sequence (/table//colspec[(preceding-sibling::colspan[(_at_)colname
=
$spanspec/@namest]
and
following-sibling::colspan[(_at_)colname = $spanspec/@nameend])
or @colname = $spanspec/@namest
or @colname = $spanspec/@nameend]);
}
<!-- sums the columns specified by the start and end column
by the total number of columns, and then
formats it to two decimal places
and concats on a percentage symbol -->
value-of (concat(string(format-number(
(sum(for $x in $first/@colwidth return
number(translate($x,'m','')))
div sum(for $x in $first/../colspec/@colwidth return
number(translate($x,'m','')))) * 100,
'00.00')),'%'));
}
}
A more plausible implementation should factor the XPaths using intermediate
vatriables
or stylesheet functions, so that each XPath line is naturally one reasonable
length line.
This will raise the line count slightly but give some nice names to the
intermediates.
Regards
Ed Willink
------------------------------------------------------------------------
E.D.Willink, Email:
mailto:EdWillink(_at_)iee(_dot_)org
Thales Research and Technology (UK) Ltd, Tel: +44 118 923 8278 (direct)
Worton Drive, or +44 118 986 8601 (ext 8278)
Worton Grange Business Park, Fax: +44 118 923 8399
Reading, RG2 0SB
ENGLAND http://www.computing.surrey.ac.uk/personal/pg/E.Willink
------------------------------------------------------------------------