--- Comment #1 from Markus Krötzsch <mar...@semantic-mediawiki.org> 2010-08-29
21:31:02 UTC ---
Good catch, but a tricky one, too. The reason for the problem is that query
descriptions serialise themselves writing <q> and </q> by means of HTML
entities < and >. This is not understood by the current query parser, who
simply ignores the respective parts including the OR, so it only evaluates the
parts in square brackets, with the conjunctive default interpretation. The
errors that are encountered are not reported since it is assumed that errors in
concept queries are shown on the concept page anyway.
The easy way to "fix" this would be to replace HTML entities in the query text,
but I am not sure if this is actually valid. The problem here is that this
would also replace those entities in all other occurrences in the query, which
may not be right (depending on whether the entities have been escaped before).
If this is so, then the query would not be reconstructible from the
part-escaped string and the serialisation would need to be changed.
Unfortunately, this would affect other code, so care is needed.
A valid question to ask here is whether this has ever worked. Anyway, some more
time is needed to look into this.
Configure bugmail: https://bugzilla.wikimedia.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.
Wikibugs-l mailing list