Page MenuHomeSoftware Heritage

swh.journal.publisher: Process origins
ClosedPublic

Authored by ardumont on Wed, Nov 21, 5:29 PM.

Details

Reviewers
olasd
douardda
Group Reviewers
Reviewers
Summary

This simply take the origin provided by the journal listener and
forwards it to the topic of reified objects.

Depends on D693

Test Plan

tox

Diff Detail

Repository
rDJNL Journal infrastructure
Branch
master
Lint
No Linters Available
Unit
No Unit Test Coverage
Build Status
Buildable 2609
Build 3242: tox-on-jenkinsJenkins
Build 3241: arc lint + arc unit

Event Timeline

ardumont created this revision.Wed, Nov 21, 5:29 PM
douardda accepted this revision.Thu, Nov 22, 9:48 AM
douardda added a subscriber: douardda.

Same as D692; LGTM but I really don't *see* what this diff is actually *doing*. Just reading this, seems like a convoluted way of testing the equality of a dict with itself, but I guess I miss some bigger picture...

This revision is now accepted and ready to land.Thu, Nov 22, 9:48 AM
ardumont edited the summary of this revision. (Show Details)Thu, Nov 22, 10:21 AM

but I really don't *see* what this diff is actually *doing*. Just reading this, seems like a convoluted way of testing the equality of a dict with itself, but I guess I miss some bigger picture...

More details on answer [1]

[1] https://forge.softwareheritage.org/D692#14417

ardumont added a comment.EditedThu, Nov 22, 10:30 AM

convoluted way of testing the equality of a dict with itself,

Since i did not have time to unstuck me on running a real kafka instance in test context (out of scope for this week ;)
I still needed to check that the data is indeed flowing through ;)

The publisher's process_objects method is the one running at the core of the publisher.
Independently of the topic subscription/publishing step (which is stuck in tests for now).

So yeah, in the end, that not much. That's better than nothing though.

ardumont closed this revision.Mon, Nov 26, 3:37 PM