38f4c44fd0
group-review: add more unit tests
go-generate-check / go-generate-check (pull_request) Failing after 28s
2026-01-07 11:56:04 +01:00
605d3dee06
group-review: add notification unit tests
2026-01-07 11:32:04 +01:00
6f26bcdccc
group-review: add mocked unit test
2026-01-07 10:42:39 +01:00
fffdf4fad3
group-review: add additional unit tests
2026-01-07 10:29:54 +01:00
f6d2239f4d
group-review: move globals to struct
...
This enables easier testing
2026-01-07 10:27:12 +01:00
913fb7c046
group-review: add systemd file
2026-01-05 17:07:27 +01:00
79318dc169
group-review: add env variables
...
instead of using command-line parameters, we can use env variables
Very helpful for services.
2026-01-05 16:53:30 +01:00
377ed1c37f
group-review: set correct comment text on negative review
2026-01-05 14:13:53 +01:00
51b0487b29
pr: parse ADD issue body
...
Validation of repository and org and other names is for the
consumer of this call. Maybe this can change later.
2025-12-16 18:33:22 +01:00
49e32c0ab1
PR: actually subscribe to PRComments
...
In previous fix attempt, there were changes to process IssueComments
but PRComments are their instance of IssueComments distinct from
IssueComments webhook events
Fixes: a418b48809
2025-12-12 16:41:25 +01:00
01e4f5f59e
pr: fix error reporting for timeline PrjGit parse
2025-12-11 18:45:56 +01:00
19d9fc5f1e
pr: request staging only when staging.config is there
...
If the project git does not have a staging.config, then there is
no reason to request reviews by the staging bot.
2025-12-11 17:01:15 +01:00
c4e184140a
pr: handle case when reviews is nil
2025-12-09 17:19:38 +01:00
56c492ccdf
PR: do not remove maintainer if also a reviewer
...
Maintainer review is only required if the PR is created by non-maintainer
or ProjetGit PR is created by non-bot. If maintainer review is not required,
but the maintainer is also listed as a Reviewer, then we cannot remove
this review request from the PR.
2025-12-08 18:03:00 +01:00
3a6009a5a3
pr: review requests cannot be stale
...
Yes, gitea marks these as stale too, but we should just ignore this
flag in case of requests. Stale requests are still pending.
2025-12-05 10:15:02 +01:00
2c4d25a5eb
pr: remove maintainers if submitter is maintainer
...
If we have case where reviews are already there, for whatever reason
(eg. upgrade from older bot), remove the pending reviews if they
are not needed
2025-12-05 09:36:38 +01:00
052ab37412
common: Loading pending reviews when loading reviews
2025-12-04 19:02:21 +01:00
925f546272
pr: check all reviews, not just ones tagged reviewers
2025-12-04 18:21:32 +01:00
71fd32a707
pr: fix debug statements
2025-12-04 18:04:51 +01:00
581131bdc8
pr: request and unrequest reviewers
...
Move the function to request and unrequest reviewers to a different
function. This will allow later simplification of the function
that determines if all reviews are complete.
Unrequesting of reviews is only possible in case of bot issued
review requests. The rest are left as-is.
2025-12-03 18:55:10 +01:00
495ed349ea
common: refactor: FetchPRSet also fetches Reviews
2025-12-03 18:55:10 +01:00
350a255d6e
pr: allow to fetch reviews in PRSet loader
2025-12-03 18:55:10 +01:00
e3087e46c2
PR: skip maintainer review if not needed
...
If project or package maintainer already reviewed the PR, as
appropriate, they are no longer re-added to the PR. We also need
to remove reviewers, but only if they were previously requested
by the bot and not something else.
2025-12-03 18:55:10 +01:00
ae6b638df6
pr: handle case of … and ... elided title
...
Also do not change title of PR if not created by bot
2025-12-03 12:41:50 +01:00
2c73cc683a
status: placeholder for factory sample data tests
2025-11-28 18:45:24 +01:00
32adfb1111
doc: fix table
2025-11-28 17:46:49 +01:00
fe8fcbae96
status: add test data
2025-11-28 17:22:12 +01:00
5756f7ceea
pr: only update PR if elided title not changed
...
Gitea trims long titles so we need to compare if the trimmed length
is same, not entire string that will always differ.
2025-11-28 12:25:58 +01:00
2be0f808d2
pr: make sure issue list is consistent
2025-11-28 12:08:54 +01:00
7a0f651eaf
direct: Gitea can send messages with no default branch
...
When a repository is created, there appears to be a race condition
where the default branch is not yet set in the message webhook
event.
We should additionally take care if the submodule is "registered"
but it wasn't correctly added, mostly due to earlier error. So,
always deinit submodules
2025-11-25 10:32:59 +01:00
2e47104b17
direct: improve commit messages in auto-updates
2025-11-24 12:58:46 +01:00
76bfa612c5
direct: use local branch name, instead of remote
...
If we fetch only one commit and force fetch the branch to local,
it seems that the remote head ref is not actually set. So, we should
just use the local version anyway, as it's updated.
2025-11-24 11:38:54 +01:00
71aa0813ad
devel: sync migrated project list
2025-11-20 20:30:38 +01:00
cc675c1b24
devel: ignore dot files
...
magic hash is when no files exist and echo "" is passed to md5sum
2025-11-20 19:57:35 +01:00
44e4941120
devel: handle build.specials.obscpio
2025-11-20 19:40:49 +01:00
86acfa6871
pr: set staging auto label according to config
...
falls back to staging/Auto if nothing is set
2025-11-20 16:25:06 +01:00
7f09b2d2d3
common: match project config before packages
...
We need to cycle through all project configs before we try to
match non-project config branches/packages. If we have multiple
project gits in one org, this coudl match wrong config
2025-11-20 13:22:40 +01:00
f3a37f1158
pr: case fold
2025-11-19 19:32:55 +01:00
9d6db86318
pr: log case of no config in verification check
2025-11-19 17:10:08 +01:00
e11993c81f
pr: do not stop processing if failed on some pacakge during check
2025-11-19 16:36:48 +01:00
4bd259a2a0
pr: ignore pkg PR if not open and no PrjGit PR
2025-11-19 16:35:17 +01:00
162ae11cdd
common: init the cache so not null
2025-11-19 15:29:52 +01:00
8431b47322
group-review: allow dots in org and package names
2025-11-19 10:00:39 +01:00
3ed5ecc3f0
pr: add staging/Auto on new PRs
...
also, cache timeline fetches
2025-11-17 11:01:45 +01:00
d08ab3efd6
direct: support reverts
2025-11-13 23:52:05 +01:00
a4f6628e52
direct: always deinit, even if dirty
2025-11-13 23:51:18 +01:00
25073dd619
direct: use correct remote name for submodules
...
should be "origin"
2025-11-13 22:15:00 +01:00
4293181b4e
pr: improve logging of review errors
...
when user is missing, log only the missing user
2025-11-12 21:39:26 +01:00
551a4ef577
direct: use origin for submodule checkout
2025-11-10 11:03:54 +01:00
6afb18fc58
direct: use correct repo for default branch
2025-11-10 10:24:33 +01:00
