)]}'
{"swift/common/middleware/mpu.py":[{"author":{"_account_id":15343,"name":"Tim Burke","email":"tburke@nvidia.com","username":"tburke"},"change_message_id":"c087e6d8e7362e5a262577187afa17a2d9a190f9","unresolved":true,"context_lines":[{"line_number":90,"context_line":""},{"line_number":91,"context_line":""},{"line_number":92,"context_line":"class MPUId(object):"},{"line_number":93,"context_line":"    ID_DELIMITER \u003d \u0027:\u0027"},{"line_number":94,"context_line":""},{"line_number":95,"context_line":"    __slots__ \u003d (\u0027uuid\u0027, \u0027timestamp\u0027)"},{"line_number":96,"context_line":""}],"source_content_type":"text/x-python","patch_set":3,"id":"f2c1e331_97bc6c31","line":93,"range":{"start_line":93,"start_character":20,"end_line":93,"end_character":21},"updated":"2024-10-01 15:49:42.000000000","message":"Is there maybe some advantage to be had by ensuring this is something that won\u0027t need to be `quote()`d? I think `-` could work if we do a `.split(\u0027-\u0027, 1)` in `parse` similar to what was there before. Or maybe `~`?","commit_id":"d117fa1dd6b79c2b36e2557cb863e54b5d1462fc"},{"author":{"_account_id":15343,"name":"Tim Burke","email":"tburke@nvidia.com","username":"tburke"},"change_message_id":"72c47471ec44ced0629947de2616fbe92bc80e0f","unresolved":true,"context_lines":[{"line_number":90,"context_line":""},{"line_number":91,"context_line":""},{"line_number":92,"context_line":"class MPUId(object):"},{"line_number":93,"context_line":"    ID_DELIMITER \u003d \u0027:\u0027"},{"line_number":94,"context_line":""},{"line_number":95,"context_line":"    __slots__ \u003d (\u0027uuid\u0027, \u0027timestamp\u0027)"},{"line_number":96,"context_line":""}],"source_content_type":"text/x-python","patch_set":3,"id":"db3be06f_29a66aca","line":93,"range":{"start_line":93,"start_character":20,"end_line":93,"end_character":21},"in_reply_to":"ba629f64_9126c26b","updated":"2024-10-03 20:33:52.000000000","message":"\u003e I\u0027d avoid \u0027-\u0027 because it can appear in the base64 encoded random part.\n\nYeah, but again, with a `.split(\u0027-\u0027, 1)` we\u0027d be good -- since `-` can\u0027t appear in the timestamp. Come to think of it, `_` _could_ work, if we used `rsplit` instead of `split`...\n\nBut sure, `~` is good too, and dodges the issue nicely.","commit_id":"d117fa1dd6b79c2b36e2557cb863e54b5d1462fc"},{"author":{"_account_id":7847,"name":"Alistair Coles","email":"alistairncoles@gmail.com","username":"acoles"},"change_message_id":"6970c612568d2d69e11152e6ed66efa51c79cde5","unresolved":true,"context_lines":[{"line_number":90,"context_line":""},{"line_number":91,"context_line":""},{"line_number":92,"context_line":"class MPUId(object):"},{"line_number":93,"context_line":"    ID_DELIMITER \u003d \u0027:\u0027"},{"line_number":94,"context_line":""},{"line_number":95,"context_line":"    __slots__ \u003d (\u0027uuid\u0027, \u0027timestamp\u0027)"},{"line_number":96,"context_line":""}],"source_content_type":"text/x-python","patch_set":3,"id":"ba629f64_9126c26b","line":93,"range":{"start_line":93,"start_character":20,"end_line":93,"end_character":21},"in_reply_to":"f2c1e331_97bc6c31","updated":"2024-10-02 13:36:46.000000000","message":"good point. I like the idea that we could add parts to the upload id relatively easily so I\u0027d avoid \u0027-\u0027 because it can appear in the base64 encoded random part. ~ seems the (only?) good option.","commit_id":"d117fa1dd6b79c2b36e2557cb863e54b5d1462fc"}]}