Track top-level module imports in the semantic model #9775
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
charliermarsh
force-pushed
the
charlie/imp
branch
from
76203ea
to
508aec6
Compare
CodSpeed Performance ReportMerging #9775 will improve performances by 14.55%Comparing Summary
Benchmarks breakdown
|
charliermarsh
force-pushed
the
charlie/imp
branch
2 times, most recently
from
2c94285
to
747f8e0
Compare
|
I suspect the gains here are actually a little under-estimated, because most of our benchmark files import NumPy, and files that don't import NumPy should get a huge boost from this change. |
|
| code | total | + violation | - violation | + fix | - fix |
|---|---|---|---|---|---|
| S101 | 117 | 0 | 117 | 0 | 0 |
| ANN202 | 70 | 0 | 70 | 0 | 0 |
| FA100 | 55 | 0 | 55 | 0 | 0 |
| AIR001 | 40 | 0 | 40 | 0 | 0 |
| ANN201 | 38 | 0 | 38 | 0 | 0 |
| ANN001 | 28 | 0 | 28 | 0 | 0 |
| D103 | 24 | 0 | 24 | 0 | 0 |
| T201 | 23 | 0 | 23 | 0 | 0 |
| PLC0415 | 18 | 0 | 18 | 0 | 0 |
| COM812 | 14 | 0 | 14 | 0 | 0 |
| SIM117 | 12 | 0 | 12 | 0 | 0 |
| FBT001 | 6 | 0 | 6 | 0 | 0 |
| FBT002 | 5 | 0 | 5 | 0 | 0 |
| PT012 | 5 | 0 | 5 | 0 | 0 |
| PTH118 | 5 | 0 | 5 | 0 | 0 |
| PTH123 | 4 | 0 | 4 | 0 | 0 |
| FIX002 | 4 | 0 | 4 | 0 | 0 |
| TD002 | 4 | 0 | 4 | 0 | 0 |
| TD003 | 4 | 0 | 4 | 0 | 0 |
| D106 | 4 | 0 | 4 | 0 | 0 |
| RET505 | 3 | 0 | 3 | 0 | 0 |
| D101 | 3 | 0 | 3 | 0 | 0 |
| D205 | 3 | 0 | 3 | 0 | 0 |
| ANN401 | 3 | 0 | 3 | 0 | 0 |
| B015 | 2 | 0 | 2 | 0 | 0 |
| CPY001 | 2 | 0 | 2 | 0 | 0 |
| D202 | 2 | 0 | 2 | 0 | 0 |
| D212 | 2 | 0 | 2 | 0 | 0 |
| PLW1514 | 2 | 0 | 2 | 0 | 0 |
| D107 | 2 | 0 | 2 | 0 | 0 |
| D400 | 2 | 0 | 2 | 0 | 0 |
| D415 | 2 | 0 | 2 | 0 | 0 |
| C408 | 2 | 0 | 2 | 0 | 0 |
| PERF401 | 1 | 0 | 1 | 0 | 0 |
| A002 | 1 | 0 | 1 | 0 | 0 |
| A001 | 1 | 0 | 1 | 0 | 0 |
| D100 | 1 | 0 | 1 | 0 | 0 |
| D401 | 1 | 0 | 1 | 0 | 0 |
| D404 | 1 | 0 | 1 | 0 | 0 |
| PTH100 | 1 | 0 | 1 | 0 | 0 |
| PTH120 | 1 | 0 | 1 | 0 | 0 |
| D209 | 1 | 0 | 1 | 0 | 0 |
| PLR0913 | 1 | 0 | 1 | 0 | 0 |
| PLR0917 | 1 | 0 | 1 | 0 | 0 |
| C901 | 1 | 0 | 1 | 0 | 0 |
| PLR0915 | 1 | 0 | 1 | 0 | 0 |
| PLR1702 | 1 | 0 | 1 | 0 | 0 |
| PLR6201 | 1 | 0 | 1 | 0 | 0 |
| PLR2004 | 1 | 0 | 1 | 0 | 0 |
| TD004 | 1 | 0 | 1 | 0 | 0 |
| RET504 | 1 | 0 | 1 | 0 | 0 |
| PLR0914 | 1 | 0 | 1 | 0 | 0 |
| ARG001 | 1 | 0 | 1 | 0 | 0 |
|
The ecosystem changes are because I gated some of the Pandas rules to require a Pandas import. I can roll that back. It honestly might be an improvement... It's trading false positives for false negatives. |
MichaReiser
approved these changes
Feb 2, 2024
There was a problem hiding this comment.
Very Clever and nice perf improvement, especially for projects that have the frameworks specific rules enabled without using the framework.
I struggle a bit with the "see" and "seen" terminology. It's very generic and to me its unclear what "semantic.see(module)" would mean.
Should we also consider' builtins'? For if pandas is globally imported, run the panda rules.
I fear that this otherwise breaks some workflows. For example, I recently recommended the use of builtins to cover the use case where they use %run other_module.py where other_module.py imports pandas globally.
Comment on lines
+56
to
+59
| if str::is_lowercase(name) { | ||
| return; | ||
| } | ||
|
|
There was a problem hiding this comment.
What if the entire performance win is because of this unrelated lint rule change 😄
|
The ecosystem changes all look like previous false positives. +1 from me. |
charliermarsh
force-pushed
the
charlie/imp
branch
from
747f8e0
to
e09bf0a
Compare
|
@zanieb - I decided to roll back that specific change. We can definitely consider it but it felt wrong to include in this optimization PR. (Some (not all) of the changes were false negatives.) |
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.
Summary
This is a simple idea to avoid unnecessary work in the linter, especially for rules that run on all name and/or all attribute nodes. Imagine a rule like the NumPy deprecation check. If the user never imported
numpy, we should be able to skip that rule entirely -- whereas today, we do aresolve_call_pathcheck on every name in the file. It turns out that there's basically a finite set of modules that we care about, so we now track imports on those modules as explicit flags on the semantic model. In rules that can only ever trigger if those modules were imported, we add a dedicated and extremely cheap check to the top of the rule.We could consider generalizing this to all modules, but I would expect that not to be much faster than
resolve_call_path, which is just a hash map lookup onTextSizeanyway.It would also be nice to make this declarative, such that rules could declare the modules they care about, the analyzers could call the rules as appropriate. But, I don't think such a design should block merging this.