)]}'
{"swift/common/middleware/tempauth.py":[{"author":{"_account_id":4608,"name":"Kota Tsuyuzaki","email":"bloodeagle40234@gmail.com","username":"tsuyuzaki-kota"},"change_message_id":"08236d9c3ddcbc34ad1bc2f33b1904f725ede31e","unresolved":false,"context_lines":[{"line_number":399,"context_line":"            user \u003d self.users[account_user]"},{"line_number":400,"context_line":"            account \u003d account_user.split(\u0027:\u0027, 1)[0]"},{"line_number":401,"context_line":"            account_id \u003d user[\u0027url\u0027].rsplit(\u0027/\u0027, 1)[-1]"},{"line_number":402,"context_line":"            if not s3_auth_details[\u0027check_signature\u0027](user[\u0027key\u0027]):"},{"line_number":403,"context_line":"                return None"},{"line_number":404,"context_line":"            env[\u0027PATH_INFO\u0027] \u003d env[\u0027PATH_INFO\u0027].replace("},{"line_number":405,"context_line":"                account_user, account_id, 1)"}],"source_content_type":"text/x-python","patch_set":1,"id":"dfeb2761_674f9892","line":402,"range":{"start_line":402,"start_character":11,"end_line":402,"end_character":67},"updated":"2017-05-11 21:37:41.000000000","message":"I don\u0027t think this is always safe. Because we have 2 version which has whether auth_details includes \"check_signature\" or not at swift3 itself.\n\nCould we log a warning if check_signature doesn\u0027t exist in the s3_auth_details dict? Or keep old handling if no check_signature is not in the request.","commit_id":"081bc305420979f28cbddb69fbf3978a41c201f8"},{"author":{"_account_id":15343,"name":"Tim Burke","email":"tburke@nvidia.com","username":"tburke"},"change_message_id":"9d2b6395cf822791fd8062909f3576fdd0c933d8","unresolved":false,"context_lines":[{"line_number":399,"context_line":"            user \u003d self.users[account_user]"},{"line_number":400,"context_line":"            account \u003d account_user.split(\u0027:\u0027, 1)[0]"},{"line_number":401,"context_line":"            account_id \u003d user[\u0027url\u0027].rsplit(\u0027/\u0027, 1)[-1]"},{"line_number":402,"context_line":"            if not s3_auth_details[\u0027check_signature\u0027](user[\u0027key\u0027]):"},{"line_number":403,"context_line":"                return None"},{"line_number":404,"context_line":"            env[\u0027PATH_INFO\u0027] \u003d env[\u0027PATH_INFO\u0027].replace("},{"line_number":405,"context_line":"                account_user, account_id, 1)"}],"source_content_type":"text/x-python","patch_set":1,"id":"3f044301_863807a0","line":402,"range":{"start_line":402,"start_character":11,"end_line":402,"end_character":67},"in_reply_to":"dfeb2761_674f9892","updated":"2017-05-11 21:51:14.000000000","message":"I\u0027m inclined to say that we don\u0027t care about non-master swift3 here, never mind some window of previous releases. Tempauth only really serves two purposes:\n\n- allow us to easily test and\n- provide an example for new auth middlewares.\n\nOn the first, if a feature is no longer required for that testing, we ought to remove it. No reason to keep that cognitive overhead when it isn\u0027t being used.\n\nOn the second, I wouldn\u0027t feel bad at all recommending a hypothetical new auth middleware to just use the provided check_signature function, and declare that they only work with new-enough versions of swift3. Although swauth (if I ever get around to submitting a patch for this there) probably *should* be more cautious and fall back to their previous v2-only behavior.\n\nThat said, you\u0027re probably right about logging a warning; if it\u0027s not available, we ought to recommend that the user upgrade swift3 (or install an older version of swift, I suppose...)","commit_id":"081bc305420979f28cbddb69fbf3978a41c201f8"}]}