#914 Section headers with templates are not correctly recognised

[Anonymous]

Sometimes bot removes valid interwiki which lads to anchor in other article.
See http://cs.wikipedia.org/w/index.php?title=Platnost_\(pr%C3%A1vo)&action=history

[Merlijn S. van Deen]

valhallasw@dorthonion:~/src/pywikipedia/trunk$ python interwiki.py cs:Platnost_%28právo%29
Getting 1 page from wikipedia:cs...
[[cs:Platnost (právo)]]: [[cs:Platnost (právo)]] gives new interwiki [[de:Gültigkeit#Gültigkeit im Recht]]
Getting 1 page from wikipedia:de...
NOTE: [[de:Gültigkeit#Gültigkeit im Recht]] does not exist. Skipping.
======Post-processing [[cs:Platnost (právo)]]======
Updating links on page [[cs:Platnost (právo)]].
Changes to be made: Robot: Removing [[de:Gültigkeit#Gültigkeit im Recht]]
- [[de:Gültigkeit#Gültigkeit im Recht]]

ERROR: Found incorrect link to de in [[cs:Platnost (právo)]]
Submit? ([y]es, [n]o, open in [b]rowser, [g]ive up, [a]lways)

tries to refer to
=== {{Anker|Rechtsg\xfcltig}}G\xfcltigkeit im Recht ===\

using #G.C3.BCltigkeit_im_Recht instead does not help

removing {{Anker|...}} does work, if the #Gültigkeit im Recht version is used...

[Merlijn S. van Deen]

Based on the wikitext, it's hard to determine whether the section title is there (evil regexp). We could combine it with a fallback to the API with action=parse, i.e.
http://de.wikipedia.org/w/api.php?action=parse&text=\{{:G%C3%BCltigkeit}}

the rewrite doesn't raise SectionErrors altogether.

options:
- stripping the check
- try to get the regexp working
- keep a simple regexp with an API fallback

questions
- how to implement this in the rewrite?

[Merlijn S. van Deen]

Long story short: it's impossible to do without another API query. We can use

http://en.wikipedia.org/w/api.php?action=parse&prop=sections&page=Help:Editing

to do this. We cannot do it using regexps because of template expansions.