I declined, because it is (currently) out of scope. We can revisit if/once this extension were to get deployed on other projects.

This extension is intended specifically for Wikipedia, and using "article" instead of "page" was a conscious design decision. See T317950 for more backstory.

Looks like some discussion is needed before this can move forward (T338452#8913695); moving to blocked.

When proposing this, I was thinking along the lines of:

New mwparserfromhtml version released per https://gitlab.wikimedia.org/repos/research/html-dumps/-/issues/119#note_43212
Dependency updated in https://gitlab.wikimedia.org/repos/structured-data/section-topics/-/commit/ab135be8ac5f086c10a5b5104d1a4a1270c758f0

https://gitlab.wikimedia.org/repos/structured-data/section-topics/-/blob/main/section_topics/pipeline.py#L131 (and duplicate code in other repos) will probably need some fixing to properly handle <ref>

All notifications for the same event_page_id & same event_agent_id are bundled. That's also explicitly the basis on which they're being bundled by (so even if we were to send multiple series of notifications for the same page, they'd still end up being bundled)

Note: added an example in T330931#8989720

FYI for testing, execute these commands to generate all expected types of notifications:

This is done: section level suggestion notifications have been delivered successfully (except for fiwiki, tracked here: T341112)

No (I was wrong when I acknowledged that before); they're all individual echo_event entries with unique event_id values, but their event_agent_id & event_page_id will be the same.

(Moving to blocked until @cchen closes, or more detail is needed)

Now that this has been implemented, here are the details for how these notifications end up being stored:

As part of T339129, I believe I'll end up deduplicating section_titles_denylist.json.

Patch is merged; should be fixed when mwparserfromhtml publishes a new release, after which we should update the version number for the package in section-topics repo.
Moving to Blocked to keep an eye out for that.

MR for mwparserfromhtml here: https://gitlab.wikimedia.org/repos/research/html-dumps/-/merge_requests/18

supported image extensions

List of wikipedias

Note: leaving this open until T330931 has been completed to make sure above assumption are correct (or do what is needed otherwise)

Quick note that this topic came up on Slack last week.

@Ladsgroup can you CR+2 this one: https://gerrit.wikimedia.org/r/c/operations/puppet/+/924877/
That change will instruct the maintenance script to submit batches to job queue instead of executing immediately (relevant code here)

Quick update: main patch & a couple of tiny follow-ups have been merged (one more minor - not functionally relevant - remaining)
A manual (synchronous; not via job queue) dry-run of all the new logic succeeded. The scheduled weekly runs completed successfully as well.
Now that we know all the refactored logic works out, we'll let it process over job queue from next week on.

Cleaned up ‘mlitn’

LGTM