hooks to test if a file contains BOM, compatible with python 2 and 3

pre-commit.py

#!/usr/bin/env python
# -*- coding: utf-8 -*-
from __future__ import absolute_import
from __future__ import print_function
from __future__ import unicode_literals
import os.path
import sys
import subprocess

UTF8_BOM = b'\xef\xbb\xbf'
UTF16_BOM = b'\xff\xfe'


def _show_working_dir():
    return subprocess.check_output(
        [
            'git',
            'rev-parse',
            '--show-toplevel',
        ]
    )


def _diff_index():
    return subprocess.check_output(
        [
            'git',
            'diff-index',
            '--cached',
            '--diff-filter=AM',
            '--name-only',
            'HEAD',
        ]
    )


def files_staged_for_commit():
    result = _diff_index().decode().split('\n')
    return result[0:-1]


def has_bom(bs):
    if bs.startswith(UTF8_BOM) or bs.startswith(UTF16_BOM):
        return True
    return False

prefix = _show_working_dir().decode().strip()
for file in files_staged_for_commit():
    with open(os.path.join(prefix, file), 'rb') as f:
        if has_bom(f.read(3)):
            print('Found BOM in', f.name)
            sys.exit(1)

pre-receive.py

#!/usr/bin/env python

from __future__ import print_function
import fileinput
import sys
import subprocess

UTF8_BOM = b'\xef\xbb\xbf'
UTF16_BOM = b'\xff\xfe'
ZERO_COMMIT = '0000000000000000000000000000000000000000'


def diff_tree(rev):
    # <mode> SP <type> SP <object> TAB <file>
    result = subprocess.check_output(
        [
            'git',
            'diff-tree',
            '--pretty=format:',
            '--diff-filter=AM',
            rev
        ]
    )
    return result.decode().strip().replace('\t', ' ')


def cat_file(object_id):
    result = subprocess.check_output(
        [
            'git',
            'cat-file',
            'blob',
            object_id
        ]
    )
    return result


def rev_list(prev, current):
    if prev == ZERO_COMMIT:
        command = [
            'git',
            'rev-list',
            current,
            '--all',
            '--not'
        ]
    else:
        command = [
            'git',
            'rev-list',
            '%s..%s' % (prev, current),
            '--all',
            '--not'
        ]

    result = subprocess.check_output(command)
    return result.decode()


def has_bom(bs):
    if bs.startswith(UTF8_BOM) or bs.startswith(UTF16_BOM):
        return True
    return False


def process(id):
    return has_bom(cat_file(id)[0:3])


if __name__ == '__main__':
    # <old-value> SP <new-value> SP <ref-name> LF
    (old_rev, new_rev, ref_name) = fileinput.input().readline().split(' ')
    for rev in rev_list(old_rev, new_rev).splitlines():
        for line in diff_tree(rev).splitlines():
            mode, t, prev_object_id, object_id, status, name = line.split(' ')
            if process(object_id):
                print('Error: found BOM in', name)
                sys.exit(1)