Open GoogleCodeExporter opened 9 years ago
Original comment by thomas.m...@gmail.com
on 13 Jul 2011 at 7:18
I have a patch for this which I've attached.
This doesn't attempt to parse the nested templates, just changes the parser to
allow them within a template block. This turned out to be a bit fiddly because
it needs to ensure the nested braces are balanced, so can no longer be done by
the lexer alone.
NB:
* This is my first time messing with JavaCC so definitely needs review (e.g. do
I need LOOKAHEAD directives somewhere? is there a quicker way to achieve this?)
* You'll need to ant -f RebuildScanners.xml after applying the patch (I didn't
include diffs to the auto-generated code)
If anyone has time to review this I'd appreciate it, may be able to contribute
some more fixes to other issues if I'm comfortable that I'm able to make
changes to the grammar in a way that people are happy with.
Original comment by matt...@swiftkey.com
on 26 Jun 2013 at 5:39
Attachments:
As a bit of motivation for this being a priority by the way: a significant
proportion of high-profile wikipedia pages *do* have nested templates, usually
within an infobox template. Without a fix for this issue, parsing these pages
results in a lot of infobox template markup being displayed as broken text,
which hurts all sorts of applications.
Original comment by matt...@swiftkey.com
on 26 Jun 2013 at 5:47
Original issue reported on code.google.com by
jerome.v...@gmail.com
on 12 Jul 2011 at 2:47