de-conf plugin request

[hgiesel]

I've been trying to parse Wiktionary pages like this one
However wtf does not parse the template string correctly: It fails to read {{de-conj|ab.tun<irreg>}} and skips the <irreg> part.

Another template it fails to parse correctly is on this page.

This:

{{de-conj|[[sich]]<accpron>_[[auf]]_[[sein]]en<pron>_[[Lorbeer]]en_[[aus.ruhen]]}

is turned by wtf into:

sich _auf_seinen _Lorbeeren_aus.ruhen

[spencermountain]

hey Henrik, this is a toughie.
de-conj results are actually generated by a script, somewhere inside wiktionary. They auto-create easier conjugations, and allow users to set exceptions.
Conjugating german verbs is beyond the scope of wtf_wikipedia, but could be a candidate for a plugin.
You can see we're generating conjugations at de-compromise, if that's what your looking for.
cheers

[hgiesel]

I don't mean that it should generate the actual results. I actually intend to do that myself.
I mean that if I parse the Wiktionary page with wtf, it seems like it drops some parts from the document.

After parsing the page, the I want to have this text: ab.tun<irreg>, however it mutilates it to this: ab.tun, and skips the <irreg> part.

[spencermountain]

ahh, ya. I see what you mean.
First- that sounds cool that you're reproducing the results. Please share-back what you can.

Yeah, as you suspected, it's the angle-brackets. The library involves a lot of xml tags, which by default, pass-through.
This also runs before the template parser.

It would be easy to support <irreg> but i'm just looking at the template doc, and see things like this:

{{de-conj|schwimmen<schwamm:schwomm[archaic; used up through the 19th century],geschwommen,schwämme:schwömme[rare]>}}

So yikes, I didn't know about this syntax. I'm not sure how to do it, to be honest.
You may find there is a solution somewhere in the kill_xml file - but I can't think of one right now.
cheers

[MarketingPip]

@hgiesel - the wiki templates parsing for Wiktionary pages needs some love keep in mind... (not just for the irregg). As there is lot's of issue's in regards to proper parsing - tho I do believe @spencermountain is already aware of this.

If you happen to play with the English Wiktionary you will run into a lot of issues with improper parsing. That said - you are more than welcome to contribute any fixes you find - as I know @spencermountain is a very busy guy.