Simplify away safe_text_dupfile/EncodedFile #6875
Closed
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
I tried to understand what the `safe_text_dupfile()` function and `EncodedFile` class do. Outside tests, `EncodedFile` is only used by `safe_text_dupfile`, and `safe_text_dupfile` is only used by `FDCaptureBinary.__init__()`. I then started to eliminate always-true conditions based on the single call site, and in the end nothing was left. The old code had some workarounds: - Ensure `errors` property is set: pytest-dev#555. It is set now, although now it will return `replace`, which is the actual method used also before, instead of `strict`. AFAIU, the issue just wanted it to be set to something, not `strict` specifically. - Infinite recursion trying to pickle `EncodedFile` (f7282b8). Not sure, I don't think this object should be pickle-able at all (and it isn't...). - A bug in `logging.Handler.__repr__` with int `file.name`. pytest-dev#2555 Was fixed in Python>=3.7.4: https://bugs.python.org/issue36015. TODO: This workaround is probably still needed?
bluetech
commented
Mar 7, 2020
| @@ -579,7 +526,7 @@ def _start(self): | |||
|
|
|||
| def snap(self): | |||
| self.tmpfile.seek(0) | |||
| res = self.tmpfile.read() | |||
| res = self.tmpfile.buffer.read() | |||
There was a problem hiding this comment.
This I think is a bug fix anyway?
There was a problem hiding this comment.
Maybe fixes #6871 ?
Might make sense to pull it out then already.
There was a problem hiding this comment.
But would only avoid the indirection via __getattr__, no? (with EncodedFile being used)
| @@ -621,10 +568,10 @@ class FDCapture(FDCaptureBinary): | |||
| EMPTY_BUFFER = str() # type: ignore | |||
|
|
|||
| def snap(self): | |||
| res = super().snap() | |||
There was a problem hiding this comment.
Since tmpfile is now text-mode for reading also (not just for writing like the weird EncodedFile did), I switched to read directly. Also symmetric with the Sys variants which do the same.
|
Hmm, seems like |
|
Replaced by #6899. |
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.
I tried to understand what the
safe_text_dupfile()function andEncodedFileclass do. Outside tests,EncodedFileis only used bysafe_text_dupfile, andsafe_text_dupfileis only used byFDCaptureBinary.__init__(). I then started to eliminate always-true conditions based on the single call site, and in the end nothing was left.The old code had some workarounds:
Ensure
errorsproperty is set: AttributeError: '_io.FileIO' object has no attribute 'errors' #555. It is set now, although now it will returnreplace, which is the actual method used also before, instead ofstrict. AFAIU, the issue just wanted it to be set to something, notstrictspecifically.Infinite recursion trying to pickle
EncodedFile(f7282b8). Not sure, I don't think this object should be pickle-able at all (and it isn't...).A bug in
logging.Handler.__repr__with intfile.name: TypeError with logging.StreamHandler.__repr__ and pytest's EncodedFile #2555Was fixed in Python>=3.7.4: https://bugs.python.org/issue36015.
TODO: This workaround is probably still needed?