Merge pull request #16 from pre-commit/large_added_files

Add check-added-large-files hook

.pre-commit-config.yaml

@@ -1,11 +1,12 @@
 -   repo: git@github.com:pre-commit/pre-commit-hooks
-    sha: 6343700aa063fe30acc319d2dc84353a35a3d6d0
+    sha: e0a6e66528c8335599408cbefe7ebf9e7e15ccb3
     hooks:
     -   id: trailing-whitespace
     -   id: end-of-file-fixer
     -   id: autopep8-wrapper
         args: ['-i', '--ignore=E265,E309,E501']
     -   id: check-json
+    -   id: check-added-large-files
     -   id: check-yaml
     -   id: debug-statements
     -   id: name-tests-test

README.md

@@ -23,6 +23,7 @@ Add this to your `.pre-commit-config.yaml`
 ### Hooks available
 
 - `autopep8-wrapper` - Runs autopep8 over python source.
+' `check-added-large-files` - Prevent giant files from being committed.
 - `check-json` - Attempts to load all json files to verify syntax.
 - `check-yaml` - Attempts to load all yaml files to verify syntax.
 - `debug-statements` - Check for pdb / ipdb / pudb statements in code.

hooks.yaml

@@ -5,6 +5,13 @@
     language: python
     files: \.py$
     args: [-i]
+-   id: check-added-large-files
+    name: Check for added large files
+    language: python
+    entry: check-added-large-files
+    description: Prevent giant files from being committed
+    # Match all files
+    files: ''
 -   id: check-json
     name: Check JSON
     description: This hook checks json files for parseable syntax.

pre_commit_hooks/check_added_large_files.py

@@ -0,0 +1,49 @@
+from __future__ import absolute_import
+from __future__ import division
+from __future__ import print_function
+from __future__ import unicode_literals
+
+import argparse
+import math
+import os
+import sys
+
+from plumbum import local
+
+
+def find_large_added_files(filenames, maxkb):
+    # Find all added files that are also in the list of files pre-commit tells
+    # us about
+    filenames = set(local['git'](
+        'diff', '--staged', '--name-only', '--diff-filter', 'A',
+    ).splitlines()) & set(filenames)
+
+    retv = 0
+    for filename in filenames:
+        kb = int(math.ceil(os.stat(filename).st_size / 1024))
+        if kb > maxkb:
+            print('{0} ({1} KB) exceeds {2} KB.'.format(filename, kb, maxkb))
+            retv = 1
+
+    return retv
+
+
+def main(argv=None):
+    argv = argv if argv is not None else sys.argv[1:]
+
+    parser = argparse.ArgumentParser()
+    parser.add_argument(
+        'filenames', nargs='*',
+        help='Filenames pre-commit believes are changed.'
+    )
+    parser.add_argument(
+        '--maxkb', type=int, default=500,
+        help='Maxmimum allowable KB for added files',
+    )
+
+    args = parser.parse_args(argv)
+    return find_large_added_files(args.filenames, args.maxkb)
+
+
+if __name__ == '__main__':
+    exit(main())

setup.py

@@ -38,6 +38,7 @@ setup(
         'console_scripts': [
             'autopep8-wrapper = pre_commit_hooks.autopep8_wrapper:main',
             'check-json = pre_commit_hooks.check_json:check_json',
+            'check-added-large-files = pre_commit_hooks.check_added_large_files:main',
             'check-yaml = pre_commit_hooks.check_yaml:check_yaml',
             'debug-statement-hook = pre_commit_hooks.debug_statement_hook:debug_statement_hook',
             'end-of-file-fixer = pre_commit_hooks.end_of_file_fixer:end_of_file_fixer',

tests/check_added_large_files_test.py

@@ -0,0 +1,79 @@
+from __future__ import absolute_import
+from __future__ import unicode_literals
+
+import io
+
+import pytest
+from plumbum import local
+
+from pre_commit_hooks.check_added_large_files import find_large_added_files
+from pre_commit_hooks.check_added_large_files import main
+
+
+@pytest.yield_fixture
+def temp_git_dir(tmpdir):
+    git_dir = tmpdir.join('gits').strpath
+    local['git']('init', git_dir)
+    yield git_dir
+
+
+def write_file(filename, contents):
+    """Hax because coveragepy chokes on nested context managers."""
+    with io.open(filename, 'w') as file_obj:
+        file_obj.write(contents)
+
+
+def test_nothing_added(temp_git_dir):
+    with local.cwd(temp_git_dir):
+        assert find_large_added_files(['f.py'], 0) == 0
+
+
+def test_adding_something(temp_git_dir):
+    with local.cwd(temp_git_dir):
+        write_file('f.py', "print('hello world')")
+        local['git']('add', 'f.py')
+
+        # Should fail with max size of 0
+        assert find_large_added_files(['f.py'], 0) == 1
+
+
+def test_add_something_giant(temp_git_dir):
+    with local.cwd(temp_git_dir):
+        write_file('f.py', 'a' * 10000)
+
+        # Should not fail when not added
+        assert find_large_added_files(['f.py'], 0) == 0
+
+        local['git']('add', 'f.py')
+
+        # Should fail with strict bound
+        assert find_large_added_files(['f.py'], 0) == 1
+
+        # Should also fail with actual bound
+        assert find_large_added_files(['f.py'], 9) == 1
+
+        # Should pass with higher bound
+        assert find_large_added_files(['f.py'], 10) == 0
+
+
+def test_added_file_not_in_pre_commits_list(temp_git_dir):
+    with local.cwd(temp_git_dir):
+        write_file('f.py', "print('hello world')")
+        local['git']('add', 'f.py')
+
+        # Should pass even with a size of 0
+        assert find_large_added_files(['g.py'], 0) == 0
+
+
+def test_integration(temp_git_dir):
+    with local.cwd(temp_git_dir):
+        assert main(argv=[]) == 0
+
+        write_file('f.py', 'a' * 10000)
+        local['git']('add', 'f.py')
+
+        # Should not fail with default
+        assert main(argv=['f.py']) == 0
+
+        # Should fail with --maxkb
+        assert main(argv=['--maxkb', '9', 'f.py']) == 1