Use posix_prefix for Apple's system Python #10256
Merged
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
|
Thanks! I'd love to test :) haven't done this before though, how do I upgrade my pip to this branch? |
|
Try |
|
Cool, I've installed it, tested by un/reinstalling the Meraki python package but it's still coming up: I'm available if you want to do live testing via Skype/Zoom |
|
Hmm OkK, so we made progress (the warning is different) but are not quite there. Thanks, I'll continue investigating. |
|
Caused by a stupid mistake 🤦 Could you install the latest change? (same command) |
|
Nice catch, here's the output now: |
|
Hmm, the output didn’t change? That doesn’t sound right. Could you try adding |
|
It looks like that worked! |
|
uranusjr
force-pushed
the
locations-osx-framework-not-default-prefix
branch
from
d9bf6e1
to
051de2f
Compare
SylvainDe
reviewed
Aug 1, 2021
|
@SylvainDe I’d still be happy to know what you think! Although it seems you didn’t post any comments? |
SylvainDe
reviewed
Aug 2, 2021
| @@ -43,8 +69,7 @@ def _infer_prefix() -> str: | |||
| """ | |||
| if _PREFERRED_SCHEME_API: | |||
| return _PREFERRED_SCHEME_API("prefix") | |||
| os_framework_global = is_osx_framework() and not running_under_virtualenv() | |||
| if os_framework_global and "osx_framework_library" in _AVAILABLE_SCHEMES: | |||
| if _should_use_osx_framework_prefix(): | |||
There was a problem hiding this comment.
I have the feeling that this somehow hides the:
if val1 in _AVAILABLE_SCHEMES:
return val1
if val2 in _AVAILABLE_SCHEMES:
return val2
In fact, I'd even be tempted to rewrite the code in such a way that the check if performed in a single place.
def yield_candidates(): # or a better name :)
if is_osx_framework() and not running_under_virtualenv(): # or _should_use_osx_framework_prefix()
yield "osx_framework_library"
yield f"{sys.implementation.name}_{os.name}"
yield sys.implementation.name
yield f"{os.name}_prefix"
yield os.name # On Windows, prefx is just called "nt".
for val in yield_candidates:
if val in _AVAILABLE_SCHEMES:
return val
return "posix_prefix"
This is probably going out of scope for a code review comment.
There was a problem hiding this comment.
The choice to not always call _AVAILABLE_SCHEMES is intentional; those that aren’t checked are from upstream CPython, and a sysconfig without those values is seriously broken, so it’s better to panic-abort than silently trying something else IMO.
| # Special case: When installing into a custom prefix, use posix_prefix | ||
| # instead of osx_framework_library. See _should_use_osx_framework_prefix() | ||
| # docstring for details. | ||
| if prefix is not None and scheme_name == "osx_framework_library": |
There was a problem hiding this comment.
If my understanding is correct, this can only correspond to the situation where we've gone into the scheme_name = _infer_prefix() case and that particular function is only called from here.
Hence, I wonder if it'd make sense to have that particular piece of logic directly in the _infer_prefix rather than as step performed afterward.
This could correspond to doing something like:
- adding a param to the _infer_prefix function about the prefix being provided or not
- writing
return "osx_framework_library" if provided_prefix is None else "posix_prefix"in the _infer_prefix function
There was a problem hiding this comment.
My intention behind the current approach is to make those_infer_* functions act similarly to sysconfig.get_preferred_scheme (because eventually we’re going to replace the ad-hoc inference with it). If we put the logic in _infer_prefix, when we remove _infer_user and _infer_home, we’ll still be left with one short _infer_prefix function. I don’t think either structure is better though, this is just how I mentally structure this logic (first I ask the macOS Python what its preferred scheme is, and if it says osx_framework_library, oh I’m going to ignore it).
Merged
Sign up for free
to subscribe to this conversation on GitHub.
Already have an account?
Sign in.
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.
For #10151 (comment)
This adds additional logic to NOT use Apple’s
osx_framework_libraryforpip install --prefix, matching distutils’s behaviour. Using a framework scheme under a custom prefix doesn’t really work, so this makes sense.cc @stiewie33 for testing.