fix(_streaming_download): handle retry-level JSON error responses
#2329
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
cfm
force-pushed
the
2232-error-on-retry
branch
from
01201a7 to
3b27867
Compare
legoktm
reviewed
Dec 12, 2024
| if contents[0:1] == b"{": | ||
| logger.error(f"Retry {retry}, received JSON error response.") | ||
| return self._handle_json_response(contents) | ||
| # ...and then at the retry level: | ||
| partial = contents[bytes_written_prev:bytes_written] |
There was a problem hiding this comment.
Should/can we skip this check if bytes_written_prev == 0 (a no-retry request)?
There was a problem hiding this comment.
I wondered about this. In terms of control flow:
- In a retry request:
bytes_written_prev < len(contents)andbytes_written = len(contents), solen(partial) < len(contents), sopartial[0:1] != contents[0:1]and is worth checking,- unless we've already returned on line 220.
- In a non-retry request:
bytes_written_prev = 0andbytes_written = len(contents), solen(partial) == len(contents), sopartial[0:1] == contents[0:1]and is indeed redundant,- unless we've already returned on line 220.
So the "should" here for me comes down to: Is this check too expensive to do twice for a successful request? Probably not. But it is confusing, so let me make it explicit.
There was a problem hiding this comment.
Done in c5d0b87. I've kept the conditionals parallel between contents and partial, but let me know if you'd prefer a nested one for partial.
cfm
force-pushed
the
2232-error-on-retry
branch
from
3b27867 to
7448448
Compare
…uest halfway through actual file-size Now that we can respond correctly to retry-level errors, this change should allow running "make regenerate-sdk-cassettes" on any valid "make dev" instance, not only one with submissions above a hard-coded file-size.
cfm
force-pushed
the
2232-error-on-retry
branch
from
7448448 to
d233eb3
Compare
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Status
Ready for review.
Description
Fixes #2232 by checking for JSON responses the proxy may have returned midstream in a
Rangeretry, not only at the very beginning of the expected response.#2330 is a limitation of the test suite I've discovered in the course of this investigation. It's out of the scope of this fix.
Test Plan
make regenerate-sdk-cassettessucceeds against unparameterizedmake dev..pyfiles for easier review!Checklist
If these changes modify code paths involving cryptography, the opening of files in VMs or network (via the RPC service) traffic, Qubes testing in the staging environment is required. For fine tuning of the graphical user interface, testing in any environment in Qubes is required. Please check as applicable:
If these changes add or remove files other than client code, the AppArmor profile may need to be updated. Please check as applicable:
If these changes modify the database schema, you should include a database migration. Please check as applicable:
mainand confirmed that the migration is self-contained and applies cleanlymainand would like the reviewer to do so