the activity API endpoint at api/v1/orgs/orgname/activities/feeds
lacks detail
#34291
Labels
topic/api
Concerns mainly the API
type/proposal
The new feature has not been accepted yet but needs to be discussed first.
Description
for blender we mirror the activity feed on our chat platform as an easy way for devs to keep up with activity on the gitea instance, we had this since the phabricator days. When we switched to gitea there was no api end point for this and we just scraped the html from https://projects.blender.org/org/blender/dashboard and were on our way. This had some drawbacks as the html format sometimes changed and it required a logged in user to get to that page, so we stole some cookies from a browser session and were once more on our way but... cookies expire... so at random intervals we had to steal new cookies.. all in all, not ideal.
in recent versions of gitea https://projects.blender.org/api/v1/orgs/blender/activities/feeds became available, which somehow doesn't require a logged in user but hey, not complaining there, what i am complaining about is a lack of detail, here's a random example, but virtually all events in the json feed miss "something"
for the following event on the html page
there's this information in the api end point
while there is an incredible amount of detail available about the user performing the action, there's other rather basic information that is seemingly completely absent, like what issue/pull is this about? what is its id? what's the title? who made it? when?
the issue ID can only be extracted from the
issue_url
or/andhtml_url
fields, the title is completely lost.the
content
field is generally filled with some arbitrary data, sometimes you get lucky and you get the ID + title of a pull/issue in there, other times it's just the ID, sometimes it's the ID + last comment, and sometimes (thecommit_repo
op_type) there's another json object string encoded into it... the field is just full of surprises, you never know what you're gonna find in it. which while exciting, not ideal.all in all, it's currently impossible to retire our html scraper and switch to the gitea api end point for this information.
Gitea Version
1.23.6
Can you reproduce the bug on the Gitea demo site?
Yes
Log Gist
No response
Screenshots
No response
Git Version
No response
Operating System
No response
How are you running Gitea?
not in control of the projects.blender.org server, i don't have this information.
Database
None
The text was updated successfully, but these errors were encountered: