dumps.wikimedia.org should have a machine readable interface for listing all sites, all types of dumps for a given site, and all versions of a given dump. This "web API" could just consist of static JSON files. An index.json in each directory, with file names and descriptions, would be a good start.
Feature wish list: https://www.mediawiki.org/wiki/Wikimedia_MediaWiki_Core_Team/Backlog/Improve_dumps#Dump_Organization_and_Discovery
| Status | Subtype | Assigned | Task | ||
|---|---|---|---|---|---|
| Open | None | T88728 Improve Wikimedia dumping infrastructure | |||
| Resolved | ArielGlenn | T92966 Machine readable interface for dumps.wikimedia.org |
The Dumps project on Wikimedia Labs is currently using such an API internally in automating the process of archiving to the Internet Archive. I am currently working on exposing this API to the web, so do voice out what you wish to have in this API, thanks!
There's a dumpruninfo.json, dumpstatus.json and report.json produced for each wiki and date as dumps run; is this sufficient for folks' needs?
I'm going to close this as done, given the existence of these json files for some time (years!)