Add linkage_subtitles.json for German Wiktionary

[empiriker]

After taking a closer look at the French linkage.py, I realized that I should probably follow the conventions there.

The PR:

• Adds linkage_subtitles.json for German Wiktionary
• Renames "semantic_relations" to "linkages"
• Adds support for the different --capture_X options
• Sorts the French linkage_subtitles.json alphabetically.

---

@xxyzz It seems to me that the linkage.py already captures most of the semantic relations (linkages). Is there something that I missed and still required implementation?

[xxyzz]

I don't remember I had written that file... I might missed some cases, I only handled two templates. Maybe you'll find more linkage templates.

The French extractor has more rooms for improvement, for example, the inflection sections are not processed. I only find a few pages that have "déclinaison" or "conjugaison" section, and all of them are huge tables. And I already created a complex function to parse small inflection tables, I'm afraid this function will grew out of control if I add more lines to it.

[empiriker]

The French linkage.py seems to be pretty broad and is pretty much what I intended to cover. I noted some problems where there are multiple links in one list item, e.g. * [[laçage]] ou [[lacement]] in the page lacet, but generally it does a good job.

The project that allows me to contribute to this repo is mostly interested in increasing the coverage of wiktextract for other editions while doing "well enough" in some fields (mainly glosses, examples, pronunciation, translations and linkages). So I'll have to prioritize these. I might then look to improve other bits out of personal interest. 🤞