Better story for import redefinitions #13969
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
This changes our importing logic to be more consistent and to treat import statements more like assignments. Fixes python#13803 The primary motivation for this is when typing modules as protocols, as in the linked issue. But it turns out we already allowed redefinition with from imports, so this just seems like a nice consistency win. We move shared logic from visit_import_all and visit_import_from (via process_imported_symbol) into add_imported_symbol. We then reuse it in visit_import. To simplify stuff, we inline the code from add_module_symbol into visit_import. Then we copy over logic from add_symbol, because MypyFile is not a SymbolTableNode, but this isn't the worst thing ever. Finally, we now need to check non-from import statements like assignments, which was a thing we weren't doing earlier.
hauntsaninja
commented
Oct 31, 2022
| [stale] | ||
| [out2] | ||
| tmp/b.py:4: error: Name "a" already defined on line 3 | ||
| [stale b] |
There was a problem hiding this comment.
Not fully sure what's going on with this particular test case, maybe we should delete it since the behaviour now looks pretty similar to ignore. Alternatively, could add some new special case logic in checker.py to disallow assignments via import x to things typed Any (although I don't really like the idea of doing that).
There was a problem hiding this comment.
We can keep this test case, though it's not super interesting. The changes to the test case look reasonable to me. There's no particular reason to generate an error here.
This comment has been minimized.
This comment has been minimized.
|
Okay, primer looks pretty good. The one change is some type ignores go from error code no-redef to assignment. I think this is okay — it's more consistent with what would happen with "from" imports. |
JukkaL
reviewed
Oct 31, 2022
| [stale] | ||
| [out2] | ||
| tmp/b.py:4: error: Name "a" already defined on line 3 | ||
| [stale b] |
There was a problem hiding this comment.
We can keep this test case, though it's not super interesting. The changes to the test case look reasonable to me. There's no particular reason to generate an error here.
| def f(self) -> str: ... | ||
|
|
||
| cond: bool | ||
| t: P |
There was a problem hiding this comment.
What about something like c: Type[Cls] and then conditionally importing two different classes? Also, f: Callable[[int], str] and then importing functions, perhaps. Or maybe these are covered by existing test cases?
|
Diff from mypy_primer, showing the effect of this PR on open source code: psycopg (https://github.com/psycopg/psycopg)
+ psycopg/psycopg/pq/__init__.py:89: error: Unused "type: ignore" comment
+ psycopg/psycopg/pq/__init__.py:89: error: Incompatible import of "module" (imported name has type Module, local name has type "object") [assignment]
+ psycopg/psycopg/pq/__init__.py:89: note: Error code "assignment" not covered by "type: ignore" comment
+ tests/test_copy.py:312: error: Unused "type: ignore" comment
+ tests/test_copy.py:312: error: Incompatible import of "BaseDumper" (imported name has type "Type[StrBinaryDumper]", local name has type "Type[StrDumper]") [assignment]
+ tests/test_copy.py:312: note: Error code "assignment" not covered by "type: ignore" comment
+ tests/test_copy_async.py:305: error: Unused "type: ignore" comment
+ tests/test_copy_async.py:305: error: Incompatible import of "BaseDumper" (imported name has type "Type[StrBinaryDumper]", local name has type "Type[StrDumper]") [assignment]
+ tests/test_copy_async.py:305: note: Error code "assignment" not covered by "type: ignore" comment
cwltool (https://github.com/common-workflow-language/cwltool)
+ cwltool/utils.py:252: error: Unused "type: ignore" comment
+ cwltool/utils.py:254: error: Unused "type: ignore" comment
werkzeug (https://github.com/pallets/werkzeug)
+ src/werkzeug/utils.py:267: error: Unused "type: ignore" comment
dragonchain (https://github.com/dragonchain/dragonchain)
+ dragonchain/transaction_processor/transaction_processor.py:45:9: error: Name "processor" already defined (by an import) [no-redef]
+ dragonchain/transaction_processor/transaction_processor.py:47:9: error: Name "processor" already defined (by an import) [no-redef]
+ dragonchain/transaction_processor/transaction_processor.py:49:9: error: Name "processor" already defined (by an import) [no-redef]
+ dragonchain/transaction_processor/transaction_processor.py:52:9: error: Name "processor" already defined (by an import) [no-redef]
+ dragonchain/transaction_processor/transaction_processor.py:54:9: error: Module has no attribute "setup" [attr-defined]
pandas (https://github.com/pandas-dev/pandas)
+ pandas/core/frame.py:2695: error: Unused "type: ignore" comment
+ pandas/core/frame.py:2695: error: Incompatible import of "statawriter" (imported name has type "Type[StataWriter117]", local name has type "Type[StataWriter]") [assignment]
+ pandas/core/frame.py:2695: note: Error code "assignment" not covered by "type: ignore" comment
+ pandas/core/frame.py:2700: error: Unused "type: ignore" comment
+ pandas/core/frame.py:2700: error: Incompatible import of "statawriter" (imported name has type "Type[StataWriterUTF8]", local name has type "Type[StataWriter]") [assignment]
+ pandas/core/frame.py:2700: note: Error code "assignment" not covered by "type: ignore" comment
pytest (https://github.com/pytest-dev/pytest)
- testing/_py/test_local.py:1156: error: Name "isimportable" already defined (possibly by an import) [no-redef]
jinja (https://github.com/pallets/jinja)
+ src/jinja2/utils.py:185: error: Unused "type: ignore" comment
aiohttp (https://github.com/aio-libs/aiohttp)
+ aiohttp/client_reqrep.py:75: error: Unused "type: ignore" comment
|
wxtim
added a commit
to wxtim/cylc
that referenced
this pull request
Feb 7, 2023
datamel
added a commit
to cylc/cylc-flow
that referenced
this pull request
Feb 7, 2023
fix mypy fail caused by python/mypy#13969
wxtim
added a commit
to wxtim/cylc
that referenced
this pull request
Feb 17, 2023
…lc into fix_localhost_platform_matching * 'fix_localhost_platform_matching' of github.com:wxtim/cylc: Fix a bug preventing `cylc vip --workflow-name=foo` from working. (cylc#5349) fix no rose vars in cylc view (cylc#5367) Cylc lint fixes (cylc#5363) data store: support unsatisfied ext_trigger fix mypy fail caused by python/mypy#13969
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.
This changes our importing logic to be more consistent and to treat import statements more like assignments.
Fixes #13803, fixes #13914, fixes half of #12965, probably fixes #12574
The primary motivation for this is when typing modules as protocols, as in #13803. But it turns out we already allowed redefinition with "from" imports, so this also seems like a nice consistency win.
We move shared logic from visit_import_all and visit_import_from (via process_imported_symbol) into add_imported_symbol. We then reuse it in visit_import.
To simplify stuff, we inline the code from add_module_symbol into visit_import. Then we copy over logic from add_symbol, because MypyFile is not a SymbolTableNode, but this isn't the worst thing ever.
Finally, we now need to check non-from import statements like assignments, which was a thing we weren't doing earlier.