Thanks for providing HTML dumps, on the English Wiktionary we've already started to use them to generate lists of "wanted" entries. This works very well, but some links are not taken into account as there are currently no dumps available for the namespaces Appendix (id: 100), Thesaurus (id: 110), Reconstruction (id: 118), and Citations (id: 114).
Would it be possible to include these in the next run?
Apparently the *-NS0-*.json.tar.gz files really do contain only namespace 0 pages even in places where $wgContentNamespaces contains namespaces other than 0, unlike the more flexible *-pages-articles-*.xml* dumps.
$ curl https://dumps.wikimedia.org/other/enterprise_html/runs/20220301/idwikisource-NS0-20220301-ENTERPRISE-HTML.json.tar.gz | tar xzOf - | jq -r ".namespace.identifier" | LANG=C sort -u % Total % Received % Xferd Average Speed Time Time Time Current Dload Upload Total Spent Left Speed 100 39.2M 100 39.2M 0 0 4333k 0 0:00:09 0:00:09 --:--:-- 4591k 0
Not sure what logic was followed to decide the contents of the JSON dumps but it's clear they don't map the XML dumps, so I can't tell whether that's a bug. Either way there are two separate tasks here (+1):
This would be also really useful for the German Wiktionary, where the Flexion namespace contains a lot of grammatical information. The namespace is also included in the normal articles-XML dump (however not easily usable without the expanded templates).
On the English Wiktionary we now use HTML dumps to generate our stats. Some of our content is not in the mainspace and therefore not reflected in the statistics. There are also problems generating information related to proto-languages, these live in the Reconstruction: namespace.