third-party/pre-commit
1
0
Fork 0
mirror of https://github.com/pre-commit/pre-commit.git synced 2026-02-17 08:14:42 +04:00
Code Issues Projects Releases Packages Wiki Activity Actions

Speed up filename filtering.

Browse source 
Before there was a `getcwd` syscall for every filename which was filtered.

Instead this is now cached per-run.

- When all files are identified by filename only: ~45% improvement
- When no files are identified by filename only: ~55% improvement

This makes little difference to overall execution, the bigger win is
eliminating the `memoize_by_cwd` hack.  Just removing the memoization would
have *increased* the runtime by 300-500%.
This commit is contained in:
Anthony Sottile 2019-01-23 20:42:27 -08:00
parent 38308dc02d
commit b1389603e0
6 changed files with 61 additions and 103 deletions
Download patch file Download diff file Expand all files Collapse all files

19
tests/commands/run_test.py
View file

@ -11,9 +11,10 @@ import pytest
import pre_commit.constants as C
from pre_commit.commands.install_uninstall import install
from pre_commit.commands.run import _compute_cols
from pre_commit.commands.run import _filter_by_include_exclude
from pre_commit.commands.run import _get_skips
from pre_commit.commands.run import _has_unmerged_paths
from pre_commit.commands.run import Classifier
from pre_commit.commands.run import filter_by_include_exclude
from pre_commit.commands.run import run
from pre_commit.util import cmd_output
from pre_commit.util import make_executable
@ -748,18 +749,22 @@ def test_fail_fast(cap_out, store, repo_with_failing_hook):
assert printed.count(b'Failing hook') == 1
def test_classifier_removes_dne():
classifier = Classifier(('this_file_does_not_exist',))
assert classifier.filenames == []
@pytest.fixture
def some_filenames():
return (
'.pre-commit-hooks.yaml',
'im_a_file_that_doesnt_exist.py',
'pre_commit/git.py',
'pre_commit/main.py',
)
def test_include_exclude_base_case(some_filenames):
ret = _filter_by_include_exclude(some_filenames, '', '^$')
ret = filter_by_include_exclude(some_filenames, '', '^$')
assert ret == [
'.pre-commit-hooks.yaml',
'pre_commit/git.py',
@ -771,22 +776,22 @@ def test_include_exclude_base_case(some_filenames):
def test_matches_broken_symlink(tmpdir):
with tmpdir.as_cwd():
os.symlink('does-not-exist', 'link')
ret = _filter_by_include_exclude({'link'}, '', '^$')
ret = filter_by_include_exclude({'link'}, '', '^$')
assert ret == ['link']
def test_include_exclude_total_match(some_filenames):
ret = _filter_by_include_exclude(some_filenames, r'^.*\.py$', '^$')
ret = filter_by_include_exclude(some_filenames, r'^.*\.py$', '^$')
assert ret == ['pre_commit/git.py', 'pre_commit/main.py']
def test_include_exclude_does_search_instead_of_match(some_filenames):
ret = _filter_by_include_exclude(some_filenames, r'\.yaml$', '^$')
ret = filter_by_include_exclude(some_filenames, r'\.yaml$', '^$')
assert ret == ['.pre-commit-hooks.yaml']
def test_include_exclude_exclude_removes_files(some_filenames):
ret = _filter_by_include_exclude(some_filenames, '', r'\.py$')
ret = filter_by_include_exclude(some_filenames, '', r'\.py$')
assert ret == ['.pre-commit-hooks.yaml']

34
tests/util_test.py
View file

@ -1,17 +1,14 @@
from __future__ import unicode_literals
import os.path
import random
import pytest
from pre_commit.util import CalledProcessError
from pre_commit.util import clean_path_on_failure
from pre_commit.util import cmd_output
from pre_commit.util import memoize_by_cwd
from pre_commit.util import parse_version
from pre_commit.util import tmpdir
from testing.util import cwd
def test_CalledProcessError_str():
@ -42,37 +39,6 @@ def test_CalledProcessError_str_nooutput():
)
@pytest.fixture
def memoized_by_cwd():
@memoize_by_cwd
def func(arg):
return arg + str(random.getrandbits(64))
return func
def test_memoized_by_cwd_returns_same_twice_in_a_row(memoized_by_cwd):
ret = memoized_by_cwd('baz')
ret2 = memoized_by_cwd('baz')
assert ret is ret2
def test_memoized_by_cwd_returns_different_for_different_args(memoized_by_cwd):
ret = memoized_by_cwd('baz')
ret2 = memoized_by_cwd('bar')
assert ret.startswith('baz')
assert ret2.startswith('bar')
assert ret != ret2
def test_memoized_by_cwd_changes_with_different_cwd(memoized_by_cwd):
ret = memoized_by_cwd('baz')
with cwd('.git'):
ret2 = memoized_by_cwd('baz')
assert ret != ret2
def test_clean_on_failure_noop(in_tmpdir):
with clean_path_on_failure('foo'):
pass