3e814bc702
The issue is that the default `limit` seems to be 30, and not 50. Due to the special case for `page == 1` and the `len(result_page) < limit` check, the `url_json_get_all_pages` function may return before all pages have been loaded. The fix is to simply remove the `limit` parameter. It wasn't used anyway. Using it correctly is somewhat tricky if we can't pass in the limit in the `page == 1` case. This may result in a couple more API requests but it's probably not a problem in practice. If it becomes a problem, we should figure out in which cases the `page == 1` special case is actually required (it was not in the cases that I tested). Note that the first link returns fewer results than the second link: * https://projects.blender.org/api/v1/users/jacqueslucke/activities/feeds?only-performed-by=true&date=2024-04-22 * https://projects.blender.org/api/v1/users/jacqueslucke/activities/feeds?only-performed-by=true&date=2024-04-22&limit=50 Pull Request: https://projects.blender.org/blender/blender/pulls/120948