add-default-license.py

#!/usr/bin/env python3

import os
import re


DEFAULT_COPYRIGHT_NOTICE = [
    'Copyright (c) Ansible Project',
    'GNU General Public License v3.0+ (see LICENSES/GPL-3.0-or-later.txt or https://www.gnu.org/licenses/gpl-3.0.txt)',
    'SPDX-License-Identifier: GPL-3.0-or-later',
]


EXTENSION_HANDLERS = {}

LICENSE_PATH_MATCHER = re.compile('(\(see )[^ ]+( or https?://[^)]+\))')


def read_lines(path):
    with open(path, 'rb') as f:
        content = f.read().decode('utf-8')
    lines = content.splitlines()
    if content.endswith('\n'):
        lines.append('')
    return lines


def write_lines(path, lines):
    content = '\n'.join(lines)
    with open(path, 'wb') as f:
        f.write(content.encode('utf-8'))
        

def identify_license(line):
    if 'GNU General Public License' in line:
        if 'v3.0+' in line:
            return 'GPL-3.0-or-later'
        if 'version 3 or later' in line:
            return 'GPL-3.0-or-later'
    if 'Simplified BSD License' in line:
        return 'BSD-2-Clause'
    if 'Apache License 2.0' in line:
        return 'Apache-2.0'
    if 'PSF License' in line or 'Python-2.0' in line:
        return 'PSF-2.0'
    if 'MIT License' in line:
        return 'MIT'
    return None


def adjust_license_path(line, license):
    return LICENSE_PATH_MATCHER.sub(lambda m: '%sLICENSES/%s.txt%s' % (m.group(1), license, m.group(2)), line)


def handle_line_comments(path, line_comment_start, shebang_start=None, shebang=None, add_shebang=False):
    lines = read_lines(path)
    found_license = False
    for index, line in enumerate(lines):
        if not line.startswith(line_comment_start):
            continue
        if 'SPDX-License-Identifier: ' in line:
            found_license = True
            break
    if found_license:
        return

    first_content_line = 0
    if lines:
        if shebang_start is not None:
            if lines[first_content_line].startswith(shebang_start):
                first_content_line += 1
        if shebang is not None:
            if lines[first_content_line] == shebang:
                first_content_line += 1
    if add_shebang and first_content_line == 0:
        lines.insert(0, shebang)
        first_content_line += 1
    for line in DEFAULT_COPYRIGHT_NOTICE:
        lines.insert(first_content_line, '%s %s' % (line_comment_start, line))
        first_content_line += 1
    if first_content_line < len(lines) and lines[first_content_line] != '':
        lines.insert(first_content_line, '')
    write_lines(path, lines)


EXTENSION_HANDLERS['.py'] = lambda path: handle_line_comments(path, '#', shebang_start='#!')
EXTENSION_HANDLERS['.sh'] = lambda path: handle_line_comments(path, '#', shebang_start='#!')
EXTENSION_HANDLERS['.yml'] = lambda path: handle_line_comments(path, '#', shebang='---', add_shebang=True)
EXTENSION_HANDLERS['.yaml'] = lambda path: handle_line_comments(path, '#', shebang='---', add_shebang=True)
EXTENSION_HANDLERS['.ps1'] = lambda path: handle_line_comments(path, '#')
EXTENSION_HANDLERS['.pem'] = lambda path: handle_line_comments(path, '#')


def process(path):
    basename = os.path.basename(path)
    extension = os.path.splitext(basename)[1]
    handler = EXTENSION_HANDLERS.get(extension)
    if basename == 'aliases':
        handler = lambda path: handle_line_comments(path, '#')
    if handler:
        try:
            handler(path)
        except Exception as exc:
            print('ERROR while processing %s: %s' % (path, exc))


current = os.getcwd()

for real_path, _, files in os.walk(current):
    path = os.path.relpath(real_path, current)
    if path.startswith('.git/') or path == '.git':
        continue
    if '/__pycache__' in path:
        continue
    for file in files:
        real_file_path = os.path.join(real_path, file)
        process(real_file_path)

add-licenses.py

#!/usr/bin/env python3

import os
import re


EXTENSION_HANDLERS = {}

LICENSE_PATH_MATCHER = re.compile('(\(see )[^ ]+( or https?://[^)]+\))')


def read_lines(path):
    with open(path, 'rb') as f:
        content = f.read().decode('utf-8')
    lines = content.splitlines()
    if content.endswith('\n'):
        lines.append('')
    return lines


def write_lines(path, lines):
    content = '\n'.join(lines)
    with open(path, 'wb') as f:
        f.write(content.encode('utf-8'))
        

def identify_license(line):
    if 'GNU General Public License' in line:
        if 'v3.0+' in line:
            return 'GPL-3.0-or-later'
        if 'version 3 or later' in line:
            return 'GPL-3.0-or-later'
    if 'Simplified BSD License' in line:
        return 'BSD-2-Clause'
    if 'Apache License 2.0' in line:
        return 'Apache-2.0'
    if 'PSF License' in line or 'Python-2.0' in line:
        return 'PSF-2.0'
    if 'MIT License' in line:
        return 'MIT'
    return None


def adjust_license_path(line, license):
    return LICENSE_PATH_MATCHER.sub(lambda m: '%sLICENSES/%s.txt%s' % (m.group(1), license, m.group(2)), line)


def handle_line_comments(path, line_comment_start):
    lines = read_lines(path)
    has_license = None
    for index, line in enumerate(lines):
        if not line.startswith(line_comment_start):
            continue
        pos = line.find('SPDX-License-Identifier: ')
        if pos >= 0:
            has_license = line[pos + len('SPDX-License-Identifier: '):]
    found_license = False
    for index, line in enumerate(lines):
        if not line.startswith(line_comment_start):
            continue
        line = line[len(line_comment_start):]
        license = identify_license(line)
        if license is not None:
            if has_license is not None and license != has_license:
                raise Exception('SPDX license identifier (%s) does not match detected license (%s) from line: %s' % (has_license, license, line))
            if found_license:
                raise Exception('Found more than one license line. The second line: %s' % (line, ))
            found_license = True
            lines[index] = '%s%s' % (line_comment_start, adjust_license_path(line, license))
            if not has_license:
                lines.insert(index + 1, '%s SPDX-License-Identifier: %s' % (line_comment_start, license))
    write_lines(path, lines)


EXTENSION_HANDLERS['.py'] = lambda path: handle_line_comments(path, '#')
EXTENSION_HANDLERS['.yml'] = lambda path: handle_line_comments(path, '#')
EXTENSION_HANDLERS['.yaml'] = lambda path: handle_line_comments(path, '#')
EXTENSION_HANDLERS['.ps1'] = lambda path: handle_line_comments(path, '#')


def process(path):
    extension = os.path.splitext(path)[1]
    handler = EXTENSION_HANDLERS.get(extension)
    if handler:
        try:
            handler(path)
        except Exception as exc:
            print('ERROR while processing %s: %s' % (path, exc))


current = os.getcwd()

for real_path, _, files in os.walk(current):
    path = os.path.relpath(real_path, current)
    if path.startswith('.git/') or path == '.git':
        continue
    if '/__pycache__' in path:
        continue
    for file in files:
        real_file_path = os.path.join(real_path, file)
        process(real_file_path)