Stop using json-rev everywhere

[marco-c]

It has become superslow in the latest versions of Mercurial, as it is now returning all changes in the revision.

We can instead use json-automationrelevance or json-info.

[shubham-y]

Hi @marco-c , can you guide me on this ? I would like to work on this issue

[marco-c]

Hello @shubham-y, you'll need to change

code-coverage/backend/code_coverage_backend/hgmo.py

Line 20 in 5653932

def hgmo_revision_details(repository, changeset):

to use json-automationrelevance or json-info (json-info is probably preferable) instead of json-rev.

1. Modify HGMO_REVISION_URL to use json-automationrelevance or json-info instead of json-rev
2. Make sure the result of the hgmo_revision_details function is the same before and after your change, for a given repository and revision (you can use mozilla-central as repository, and 0058e174c5d0a4f751398b5aefcce42adc511535 as revision). Luckily you can test it in isolation, as this function doesn't depend on other parts of the project.

[mugunthanramesh]

Is anyone taking this issue ? Can i take this and i have few doubts @marco-c i am new here please correct me if i am wrong

1. i checked json-info but it doesn't return pushid
2. json-automationrelevance may also slow down. It information on what changesets are relevant to automated consumers given a source changeset. (from the doc)

[marco-c]

Is anyone taking this issue ? Can i take this

Issues are taken/assigned when there is a PR linked to them.

i checked json-info but it doesn't return pushid

OK, we can't use it then.

json-automationrelevance may also slow down. It information on what changesets are relevant to automated consumers given a source changeset. (from the doc)

It is slower than json-info, but way faster than json-rev, so let's switch to json-automationrelevance.

[mugunthanramesh]

ok i will make a PR with json-automationrelevance.

update :
PR raised

[mugunthanramesh]

Need to update the tests checks are failing

[Anupam-USP]

Hey @marco-c Is the issue still open I want to contribute?

[marco-c]

@Anupam-USP the issue is still open, there is a PR linked to it but it has probably been abandoned.