Add catalog indexer worker

[stacimc]

Fixes

Fixes #4147 by @stacimc

Description

This PR adds the catalog indexer worker. It does not actually connect to it in the data refresh yet.

Some unit tests are added, but because of the nature of the worker the bulk of the tests will be integration tests modeled after the existing ingestion server integration tests. I was able to get these mostly working locally but pulled them out of this PR as I think it will make more sense to add the integration tests once the data refresh is fully implemented, possibly reusing issue #4340.

Almost all of this code is just a refactor of existing code. Where possible I've noted when files are almost entirely copied over from the ingestion server.

Note that the service is called catalog_indexer_worker to distinguish it from the existing indexer_worker that belongs to the ingestion server, which we must keep around for now.

Testing Instructions

just build && just up

Use elasticvue to Delete All Documents from one of your existing indices. I used audio-init-filtered. The index should now have 0 documents.

Now run just catalog/shell, and you should be able to curl the indexer worker.

# Replace the target_index name if you don't have audio-init-filtered
# Note the start_id/end_id we're using, to make it reindex only 100 records
curl -X POST -d '{"model_name": "audio", "table_name": "audio", "start_id": 0, "end_id": 100, "target_index": "audio-init-filtered"}' -H "Content-Type: application/json" 'http://catalog_indexer_worker:8003/task'

You should get back a response like:

{
  "message": "Successfully scheduled task.", 
  "task_id": "adbec2cbc93a4c2f895caf12ede3eb88", 
  "status_check": "http://catalog_indexer_worker:8003/task/adbec2cbc93a4c2f895caf12ede3eb88"
}

Now curl the status_check endpoint that was just returned:

curl 'http://catalog_indexer_worker:8003/task/adbec2cbc93a4c2f895caf12ede3eb88'

And you should see something like:

{
  "task_id": "adbec2cbc93a4c2f895caf12ede3eb88",
  "active": false,
  "model": "audio",
  "target_index": "audio-init-filtered",
  "progress": 100.0,
  "start_time": "2024-05-14 21:27:54.437745",
  "finish_time": "2024-05-14 21:27:54.575137",
  "error": false
}

In elasticvue, you should now see 100 docs in your target index.

If you used a filtered index, you can now query the API (http://localhost:50281/v1/audio) and see that the results are returned as expected. ⚠️ Important: In Elasticvue you must first refresh the index to make the documents available. Then you can inspect them in Elasticvue, or check the API. This will be handled by the DAG later.

You can also try running the just commands for the indexer worker.

Checklist

• My pull request has a descriptive title (not a vague title likeUpdate index.md).
• My pull request targets the default branch of the repository (main) or a parent feature branch.
• My commit messages follow best practices.
• My code follows the established code style of the repository.
• I added or updated tests for the changes I made (if applicable).
• I added or updated documentation (if applicable).
• I tried running the project locally and verified that there are no visible errors.
• I ran the DAG documentation generator (just catalog/generate-docs for catalog
  PRs) or the media properties generator (just catalog/generate-docs media-props
  for the catalog or just api/generate-docs for the API) where applicable.

Developer Certificate of Origin

Developer Certificate of Origin

Developer Certificate of Origin
Version 1.1

Copyright (C) 2004, 2006 The Linux Foundation and its contributors.
1 Letterman Drive
Suite D4700
San Francisco, CA, 94129

Everyone is permitted to copy and distribute verbatim copies of this
license document, but changing it is not allowed.


Developer's Certificate of Origin 1.1

By making a contribution to this project, I certify that:

(a) The contribution was created in whole or in part by me and I
    have the right to submit it under the open source license
    indicated in the file; or

(b) The contribution is based upon previous work that, to the best
    of my knowledge, is covered under an appropriate open source
    license and I have the right under that license to submit that
    work with modifications, whether created in whole or in part
    by me, under the same open source license (unless I am
    permitted to submit under a different license), as indicated
    in the file; or

(c) The contribution was provided directly to me by some other
    person who certified (a), (b) or (c) and I have not modified
    it.

(d) I understand and agree that this project and the contribution
    are public and that a record of the contribution (including all
    personal information I submit with it, including my sign-off) is
    maintained indefinitely and may be redistributed consistent with
    this project or the open source license(s) involved.

[stacimc]

Copied from ingestion_server without changes to logic.

[stacimc]

Copied from ingestion_server without changes to logic.

[stacimc]

Copied from ingestion_server without changes to logic. This provides the mappings from db record to ES document.

[stacimc]

Copied from ingestion_server without changes to logic.

[sarayourfriend]

@stacimc I've opened a PR targeting this branch that switches Pipenv out for PDM, add a test recipe to the new justfile, and to run indexer worker tests in CI.

#4497

I didn't see a separate issue for implanting CI for the new project, so I wasn't sure where they would go, if not in this PR; but if you want to keep it as a fast-follow, no worries, it doesn't need to be merged into this PR, can be on its own if you'd prefer.

[sarayourfriend]

This LGTM! On #4497 I can run just indexer_worker/test and all the tests pass. On this branch, just indexer_worker/test-local also passes after running through the local venv installation.

I have not spent much time reviewing the code line by line given how much of it is copy/paste. That's fine by me, as well as the decision to push tests off into integration style tests in later issues 👍

[AetherUnbound]

This looks great! I was able to test this locally and it worked as expected. A couple of comments here but nothing blocking, if anything we can make issues going forward for addressing them as non-urgent follow ups (and possibly good first issues!).

[AetherUnbound]

Maybe this is something we can make an issue for, particularly given this might be longer-term code now: it would be great if this could be a dataclass or namedtuple, rather than a dictionary! And then we could have some typing around the self.tasks dictionary as well, which would be neat 😄

[stacimc]

Of course! Sounds like a good first issue as well. Issue: #4534 👍