From ba6315d2f6fe550912ab2ba87f165329d9e2f834 Mon Sep 17 00:00:00 2001 From: fanquake Date: Fri, 19 Dec 2025 16:43:42 +0000 Subject: [PATCH] contrib: remove copyright_header.py --- contrib/devtools/README.md | 59 --- contrib/devtools/copyright_header.py | 601 --------------------------- 2 files changed, 660 deletions(-) delete mode 100755 contrib/devtools/copyright_header.py diff --git a/contrib/devtools/README.md b/contrib/devtools/README.md index 3309057b6f3..50e807814c5 100644 --- a/contrib/devtools/README.md +++ b/contrib/devtools/README.md @@ -61,65 +61,6 @@ the script should be called from the git root folder as follows. git diff -U0 HEAD~1.. | ./contrib/devtools/clang-format-diff.py -p1 -i -v ``` -copyright\_header.py -==================== - -Provides utilities for managing copyright headers of `The Bitcoin Core -developers` in repository source files. It has three subcommands: - -``` -$ ./copyright_header.py report [verbose] -$ ./copyright_header.py update -$ ./copyright_header.py insert -``` -Running these subcommands without arguments displays a usage string. - -copyright\_header.py report \ [verbose] ---------------------------------------------------------- - -Produces a report of all copyright header notices found inside the source files -of a repository. Useful to quickly visualize the state of the headers. -Specifying `verbose` will list the full filenames of files of each category. - -copyright\_header.py update \ [verbose] ---------------------------------------------------------- -Updates all the copyright headers of `The Bitcoin Core developers` which were -changed in a year more recent than is listed. For example: -``` -// Copyright (c) - The Bitcoin Core developers -``` -will be updated to: -``` -// Copyright (c) - The Bitcoin Core developers -``` -where `` is obtained from the `git log` history. - -This subcommand also handles copyright headers that have only a single year. In -those cases: -``` -// Copyright (c) The Bitcoin Core developers -``` -will be updated to: -``` -// Copyright (c) - The Bitcoin Core developers -``` -where the update is appropriate. - -copyright\_header.py insert \ ------------------------------------- -Inserts a copyright header for `The Bitcoin Core developers` at the top of the -file in either Python or C++ style as determined by the file extension. If the -file is a Python file and it has `#!` starting the first line, the header is -inserted in the line below it. - -The copyright dates will be set to be `-` where -`` is according to the `git log` history. If -`` is equal to ``, it will be set as a single -year rather than two hyphenated years. - -If the file already has a copyright for `The Bitcoin Core developers`, the -script will exit. - gen-manpages.py =============== diff --git a/contrib/devtools/copyright_header.py b/contrib/devtools/copyright_header.py deleted file mode 100755 index aa77de8aa6f..00000000000 --- a/contrib/devtools/copyright_header.py +++ /dev/null @@ -1,601 +0,0 @@ -#!/usr/bin/env python3 -# Copyright (c) 2016-present The Bitcoin Core developers -# Distributed under the MIT software license, see the accompanying -# file COPYING or http://www.opensource.org/licenses/mit-license.php. - -import re -import fnmatch -import sys -import subprocess -import datetime -import os - -################################################################################ -# file filtering -################################################################################ - -EXCLUDE = [ - # auto generated: - 'src/qt/bitcoinstrings.cpp', - 'src/chainparamsseeds.h', - # other external copyrights: - 'src/test/fuzz/FuzzedDataProvider.h', - 'src/tinyformat.h', - 'src/bench/nanobench.h', - # python init: - '*__init__.py', -] -EXCLUDE_COMPILED = re.compile('|'.join([fnmatch.translate(m) for m in EXCLUDE])) - -EXCLUDE_DIRS = [ - # git subtrees - "src/crypto/ctaes/", - "src/leveldb/", - "src/minisketch", - "src/secp256k1/", - "src/crc32c/", -] - -INCLUDE = ['*.h', '*.cpp', '*.cc', '*.c', '*.mm', '*.py', '*.sh', '*.bash-completion'] -INCLUDE_COMPILED = re.compile('|'.join([fnmatch.translate(m) for m in INCLUDE])) - -def applies_to_file(filename): - for excluded_dir in EXCLUDE_DIRS: - if filename.startswith(excluded_dir): - return False - return ((EXCLUDE_COMPILED.match(filename) is None) and - (INCLUDE_COMPILED.match(filename) is not None)) - -################################################################################ -# obtain list of files in repo according to INCLUDE and EXCLUDE -################################################################################ - -GIT_LS_CMD = 'git ls-files --full-name'.split(' ') -GIT_TOPLEVEL_CMD = 'git rev-parse --show-toplevel'.split(' ') - -def call_git_ls(base_directory): - out = subprocess.check_output([*GIT_LS_CMD, base_directory], text=True) - return [f for f in out.split('\n') if f != ''] - -def call_git_toplevel(): - "Returns the absolute path to the project root" - return subprocess.check_output(GIT_TOPLEVEL_CMD, text=True).strip() - -def get_filenames_to_examine(base_directory): - "Returns an array of absolute paths to any project files in the base_directory that pass the include/exclude filters" - root = call_git_toplevel() - filenames = call_git_ls(base_directory) - return sorted([os.path.join(root, filename) for filename in filenames if - applies_to_file(filename)]) - -################################################################################ -# define and compile regexes for the patterns we are looking for -################################################################################ - - -COPYRIGHT_WITH_C = r'Copyright \(c\)' -COPYRIGHT_WITHOUT_C = 'Copyright' -ANY_COPYRIGHT_STYLE = '(%s|%s)' % (COPYRIGHT_WITH_C, COPYRIGHT_WITHOUT_C) - -YEAR = "20[0-9][0-9]" -YEAR_RANGE = '(%s)(-%s)?' % (YEAR, YEAR) -YEAR_LIST = '(%s)(, %s)+' % (YEAR, YEAR) -ANY_YEAR_STYLE = '(%s|%s)' % (YEAR_RANGE, YEAR_LIST) -ANY_COPYRIGHT_STYLE_OR_YEAR_STYLE = ("%s %s" % (ANY_COPYRIGHT_STYLE, - ANY_YEAR_STYLE)) - -ANY_COPYRIGHT_COMPILED = re.compile(ANY_COPYRIGHT_STYLE_OR_YEAR_STYLE) - -def compile_copyright_regex(copyright_style, year_style, name): - return re.compile(r'%s %s,? %s( +\*)?\n' % (copyright_style, year_style, name)) - -EXPECTED_HOLDER_NAMES = [ - r"Satoshi Nakamoto", - r"The Bitcoin Core developers", - r"BitPay Inc\.", - r"Pieter Wuille", - r"Wladimir J\. van der Laan", - r"Jeff Garzik", - r"Jan-Klaas Kollhof", - r"ArtForz -- public domain half-a-node", - r"Intel Corporation ?", - r"The Zcash developers", - r"Jeremy Rubin", -] - -DOMINANT_STYLE_COMPILED = {} -YEAR_LIST_STYLE_COMPILED = {} -WITHOUT_C_STYLE_COMPILED = {} - -for holder_name in EXPECTED_HOLDER_NAMES: - DOMINANT_STYLE_COMPILED[holder_name] = ( - compile_copyright_regex(COPYRIGHT_WITH_C, YEAR_RANGE, holder_name)) - YEAR_LIST_STYLE_COMPILED[holder_name] = ( - compile_copyright_regex(COPYRIGHT_WITH_C, YEAR_LIST, holder_name)) - WITHOUT_C_STYLE_COMPILED[holder_name] = ( - compile_copyright_regex(COPYRIGHT_WITHOUT_C, ANY_YEAR_STYLE, - holder_name)) - -################################################################################ -# search file contents for copyright message of particular category -################################################################################ - -def get_count_of_copyrights_of_any_style_any_holder(contents): - return len(ANY_COPYRIGHT_COMPILED.findall(contents)) - -def file_has_dominant_style_copyright_for_holder(contents, holder_name): - match = DOMINANT_STYLE_COMPILED[holder_name].search(contents) - return match is not None - -def file_has_year_list_style_copyright_for_holder(contents, holder_name): - match = YEAR_LIST_STYLE_COMPILED[holder_name].search(contents) - return match is not None - -def file_has_without_c_style_copyright_for_holder(contents, holder_name): - match = WITHOUT_C_STYLE_COMPILED[holder_name].search(contents) - return match is not None - -################################################################################ -# get file info -################################################################################ - -def read_file(filename): - return open(filename, 'r').read() - -def gather_file_info(filename): - info = {} - info['filename'] = filename - c = read_file(filename) - info['contents'] = c - - info['all_copyrights'] = get_count_of_copyrights_of_any_style_any_holder(c) - - info['classified_copyrights'] = 0 - info['dominant_style'] = {} - info['year_list_style'] = {} - info['without_c_style'] = {} - for holder_name in EXPECTED_HOLDER_NAMES: - has_dominant_style = ( - file_has_dominant_style_copyright_for_holder(c, holder_name)) - has_year_list_style = ( - file_has_year_list_style_copyright_for_holder(c, holder_name)) - has_without_c_style = ( - file_has_without_c_style_copyright_for_holder(c, holder_name)) - info['dominant_style'][holder_name] = has_dominant_style - info['year_list_style'][holder_name] = has_year_list_style - info['without_c_style'][holder_name] = has_without_c_style - if has_dominant_style or has_year_list_style or has_without_c_style: - info['classified_copyrights'] = info['classified_copyrights'] + 1 - return info - -################################################################################ -# report execution -################################################################################ - -SEPARATOR = '-'.join(['' for _ in range(80)]) - -def print_filenames(filenames, verbose): - if not verbose: - return - for filename in filenames: - print("\t%s" % filename) - -def print_report(file_infos, verbose): - print(SEPARATOR) - examined = [i['filename'] for i in file_infos] - print("%d files examined according to INCLUDE and EXCLUDE fnmatch rules" % - len(examined)) - print_filenames(examined, verbose) - - print(SEPARATOR) - print('') - zero_copyrights = [i['filename'] for i in file_infos if - i['all_copyrights'] == 0] - print("%4d with zero copyrights" % len(zero_copyrights)) - print_filenames(zero_copyrights, verbose) - one_copyright = [i['filename'] for i in file_infos if - i['all_copyrights'] == 1] - print("%4d with one copyright" % len(one_copyright)) - print_filenames(one_copyright, verbose) - two_copyrights = [i['filename'] for i in file_infos if - i['all_copyrights'] == 2] - print("%4d with two copyrights" % len(two_copyrights)) - print_filenames(two_copyrights, verbose) - three_copyrights = [i['filename'] for i in file_infos if - i['all_copyrights'] == 3] - print("%4d with three copyrights" % len(three_copyrights)) - print_filenames(three_copyrights, verbose) - four_or_more_copyrights = [i['filename'] for i in file_infos if - i['all_copyrights'] >= 4] - print("%4d with four or more copyrights" % len(four_or_more_copyrights)) - print_filenames(four_or_more_copyrights, verbose) - print('') - print(SEPARATOR) - print('Copyrights with dominant style:\ne.g. "Copyright (c)" and ' - '"" or "-":\n') - for holder_name in EXPECTED_HOLDER_NAMES: - dominant_style = [i['filename'] for i in file_infos if - i['dominant_style'][holder_name]] - if len(dominant_style) > 0: - print("%4d with '%s'" % (len(dominant_style), - holder_name.replace('\n', '\\n'))) - print_filenames(dominant_style, verbose) - print('') - print(SEPARATOR) - print('Copyrights with year list style:\ne.g. "Copyright (c)" and ' - '", , ...":\n') - for holder_name in EXPECTED_HOLDER_NAMES: - year_list_style = [i['filename'] for i in file_infos if - i['year_list_style'][holder_name]] - if len(year_list_style) > 0: - print("%4d with '%s'" % (len(year_list_style), - holder_name.replace('\n', '\\n'))) - print_filenames(year_list_style, verbose) - print('') - print(SEPARATOR) - print('Copyrights with no "(c)" style:\ne.g. "Copyright" and "" or ' - '"-":\n') - for holder_name in EXPECTED_HOLDER_NAMES: - without_c_style = [i['filename'] for i in file_infos if - i['without_c_style'][holder_name]] - if len(without_c_style) > 0: - print("%4d with '%s'" % (len(without_c_style), - holder_name.replace('\n', '\\n'))) - print_filenames(without_c_style, verbose) - - print('') - print(SEPARATOR) - - unclassified_copyrights = [i['filename'] for i in file_infos if - i['classified_copyrights'] < i['all_copyrights']] - print("%d with unexpected copyright holder names" % - len(unclassified_copyrights)) - print_filenames(unclassified_copyrights, verbose) - print(SEPARATOR) - -def exec_report(base_directory, verbose): - filenames = get_filenames_to_examine(base_directory) - file_infos = [gather_file_info(f) for f in filenames] - print_report(file_infos, verbose) - -################################################################################ -# report cmd -################################################################################ - -REPORT_USAGE = """ -Produces a report of all copyright header notices found inside the source files -of a repository. - -Usage: - $ ./copyright_header.py report [verbose] - -Arguments: - - The base directory of a bitcoin source code repository. - [verbose] - Includes a list of every file of each subcategory in the report. -""" - -def report_cmd(argv): - if len(argv) == 2: - sys.exit(REPORT_USAGE) - - base_directory = argv[2] - if not os.path.exists(base_directory): - sys.exit("*** bad : %s" % base_directory) - - if len(argv) == 3: - verbose = False - elif argv[3] == 'verbose': - verbose = True - else: - sys.exit("*** unknown argument: %s" % argv[2]) - - exec_report(base_directory, verbose) - -################################################################################ -# query git for year of last change -################################################################################ - -GIT_LOG_CMD = "git log --pretty=format:%%ai %s" - -def call_git_log(filename): - out = subprocess.check_output((GIT_LOG_CMD % filename).split(' '), text=True) - return out.split('\n') - -def get_git_change_years(filename): - git_log_lines = call_git_log(filename) - if len(git_log_lines) == 0: - return [datetime.date.today().year] - # timestamp is in ISO 8601 format. e.g. "2016-09-05 14:25:32 -0600" - return [line.split(' ')[0].split('-')[0] for line in git_log_lines] - -def get_most_recent_git_change_year(filename): - return max(get_git_change_years(filename)) - -################################################################################ -# read and write to file -################################################################################ - -def read_file_lines(filename): - with open(filename, 'r') as f: - file_lines = f.readlines() - return file_lines - -def write_file_lines(filename, file_lines): - with open(filename, 'w') as f: - f.write(''.join(file_lines)) - -################################################################################ -# update header years execution -################################################################################ - -COPYRIGHT = r'Copyright \(c\)' -YEAR = "20[0-9][0-9]" -YEAR_RANGE = '(%s)(-%s)?' % (YEAR, YEAR) -HOLDER = 'The Bitcoin Core developers' -UPDATEABLE_LINE_COMPILED = re.compile(' '.join([COPYRIGHT, YEAR_RANGE, HOLDER])) - -def get_updatable_copyright_line(file_lines): - index = 0 - for line in file_lines: - if UPDATEABLE_LINE_COMPILED.search(line) is not None: - return index, line - index = index + 1 - return None, None - -def parse_year_range(year_range): - year_split = year_range.split('-') - start_year = year_split[0] - if len(year_split) == 1: - return start_year, start_year - return start_year, year_split[1] - -def year_range_to_str(start_year, end_year): - if start_year == end_year: - return start_year - return "%s-%s" % (start_year, end_year) - -def create_updated_copyright_line(line, last_git_change_year): - copyright_splitter = 'Copyright (c) ' - copyright_split = line.split(copyright_splitter) - # Preserve characters on line that are ahead of the start of the copyright - # notice - they are part of the comment block and vary from file-to-file. - before_copyright = copyright_split[0] - after_copyright = copyright_split[1] - - space_split = after_copyright.split(' ') - year_range = space_split[0] - start_year, end_year = parse_year_range(year_range) - if end_year >= last_git_change_year: - return line - return (before_copyright + copyright_splitter + - year_range_to_str(start_year, last_git_change_year) + ' ' + - ' '.join(space_split[1:])) - -def update_updatable_copyright(filename): - file_lines = read_file_lines(filename) - index, line = get_updatable_copyright_line(file_lines) - if not line: - print_file_action_message(filename, "No updatable copyright.") - return - last_git_change_year = get_most_recent_git_change_year(filename) - new_line = create_updated_copyright_line(line, last_git_change_year) - if line == new_line: - print_file_action_message(filename, "Copyright up-to-date.") - return - file_lines[index] = new_line - write_file_lines(filename, file_lines) - print_file_action_message(filename, - "Copyright updated! -> %s" % last_git_change_year) - -def exec_update_header_year(base_directory): - for filename in get_filenames_to_examine(base_directory): - update_updatable_copyright(filename) - -################################################################################ -# update cmd -################################################################################ - -UPDATE_USAGE = """ -Updates all the copyright headers of "The Bitcoin Core developers" which were -changed in a year more recent than is listed. For example: - -// Copyright (c) - The Bitcoin Core developers - -will be updated to: - -// Copyright (c) - The Bitcoin Core developers - -where is obtained from the 'git log' history. - -This subcommand also handles copyright headers that have only a single year. In those cases: - -// Copyright (c) The Bitcoin Core developers - -will be updated to: - -// Copyright (c) - The Bitcoin Core developers - -where the update is appropriate. - -Usage: - $ ./copyright_header.py update - -Arguments: - - The base directory of a bitcoin source code repository. -""" - -def print_file_action_message(filename, action): - print("%-52s %s" % (filename, action)) - -def update_cmd(argv): - if len(argv) != 3: - sys.exit(UPDATE_USAGE) - - base_directory = argv[2] - if not os.path.exists(base_directory): - sys.exit("*** bad base_directory: %s" % base_directory) - exec_update_header_year(base_directory) - -################################################################################ -# inserted copyright header format -################################################################################ - -def get_header_lines(header, start_year, end_year): - lines = header.split('\n')[1:-1] - lines[0] = lines[0] % year_range_to_str(start_year, end_year) - return [line + '\n' for line in lines] - -CPP_HEADER = ''' -// Copyright (c) %s The Bitcoin Core developers -// Distributed under the MIT software license, see the accompanying -// file COPYING or http://www.opensource.org/licenses/mit-license.php. -''' - -def get_cpp_header_lines_to_insert(start_year, end_year): - return reversed(get_header_lines(CPP_HEADER, start_year, end_year)) - -SCRIPT_HEADER = ''' -# Copyright (c) %s The Bitcoin Core developers -# Distributed under the MIT software license, see the accompanying -# file COPYING or http://www.opensource.org/licenses/mit-license.php. -''' - -def get_script_header_lines_to_insert(start_year, end_year): - return reversed(get_header_lines(SCRIPT_HEADER, start_year, end_year)) - -################################################################################ -# query git for year of last change -################################################################################ - -def get_git_change_year_range(filename): - years = get_git_change_years(filename) - return min(years), max(years) - -################################################################################ -# check for existing core copyright -################################################################################ - -def file_already_has_core_copyright(file_lines): - index, _ = get_updatable_copyright_line(file_lines) - return index is not None - -################################################################################ -# insert header execution -################################################################################ - -def file_has_hashbang(file_lines): - if len(file_lines) < 1: - return False - if len(file_lines[0]) <= 2: - return False - return file_lines[0][:2] == '#!' - -def insert_script_header(filename, file_lines, start_year, end_year): - if file_has_hashbang(file_lines): - insert_idx = 1 - else: - insert_idx = 0 - header_lines = get_script_header_lines_to_insert(start_year, end_year) - for line in header_lines: - file_lines.insert(insert_idx, line) - write_file_lines(filename, file_lines) - -def insert_cpp_header(filename, file_lines, start_year, end_year): - file_lines.insert(0, '\n') - header_lines = get_cpp_header_lines_to_insert(start_year, end_year) - for line in header_lines: - file_lines.insert(0, line) - write_file_lines(filename, file_lines) - -def exec_insert_header(filename, style): - file_lines = read_file_lines(filename) - if file_already_has_core_copyright(file_lines): - sys.exit('*** %s already has a copyright by The Bitcoin Core developers' - % (filename)) - start_year, end_year = get_git_change_year_range(filename) - if style in ['python', 'shell']: - insert_script_header(filename, file_lines, start_year, end_year) - else: - insert_cpp_header(filename, file_lines, start_year, end_year) - -################################################################################ -# insert cmd -################################################################################ - -INSERT_USAGE = """ -Inserts a copyright header for "The Bitcoin Core developers" at the top of the -file in either Python or C++ style as determined by the file extension. If the -file is a Python file and it has a '#!' starting the first line, the header is -inserted in the line below it. - -The copyright dates will be set to be: - -"-" - -where is according to the 'git log' history. If - is equal to , the date will be set to be: - -"" - -If the file already has a copyright for "The Bitcoin Core developers", the -script will exit. - -Usage: - $ ./copyright_header.py insert - -Arguments: - - A source file in the bitcoin repository. -""" - -def insert_cmd(argv): - if len(argv) != 3: - sys.exit(INSERT_USAGE) - - filename = argv[2] - if not os.path.isfile(filename): - sys.exit("*** bad filename: %s" % filename) - _, extension = os.path.splitext(filename) - if extension not in ['.h', '.cpp', '.cc', '.c', '.py', '.sh']: - sys.exit("*** cannot insert for file extension %s" % extension) - - if extension == '.py': - style = 'python' - elif extension == '.sh': - style = 'shell' - else: - style = 'cpp' - exec_insert_header(filename, style) - -################################################################################ -# UI -################################################################################ - -USAGE = """ -copyright_header.py - utilities for managing copyright headers of 'The Bitcoin -Core developers' in repository source files. - -Usage: - $ ./copyright_header - -Subcommands: - report - update - insert - -To see subcommand usage, run them without arguments. -""" - -SUBCOMMANDS = ['report', 'update', 'insert'] - -if __name__ == "__main__": - if len(sys.argv) == 1: - sys.exit(USAGE) - subcommand = sys.argv[1] - if subcommand not in SUBCOMMANDS: - sys.exit(USAGE) - if subcommand == 'report': - report_cmd(sys.argv) - elif subcommand == 'update': - update_cmd(sys.argv) - elif subcommand == 'insert': - insert_cmd(sys.argv)