add-copyright

#!/usr/bin/env python3
#
# Copyright (c) 2023 MariaDB plc
#
# Use of this software is governed by the Business Source License included
# in the LICENSE.TXT file and at www.mariadb.com/bsl11.
#
# Change Date: 2024-06-03
#
# On the date above, in accordance with the Business Source License, use
# of this software will be governed by version 2 or later of the General
# Public License.
#

import getopt
import os
import re
import shutil
import sys
import tempfile

# A string literal would be a match and cause an unintended modification.
pattern="Copyright "
pattern+=".*"
pattern+="MariaDB .*"

# The Copyright should be in the blurb at the beginning. We will not
# do modifications after this threshold.
line_threshold=20

recd=re.compile(pattern)

def print_usage_and_exit(name, code):
    print("usage: " + name + " -c copyright [path]+")
    print()
    print("Add a specfic copyright line after the \"MariaDB\"")
    print("copyright line. If path is a directory, the copyright will ")
    print("recursively be added to all files in that directory hierarchy.")
    print()
    print("Example: " + name + " -c \"2023 MariaDB plc, Finnish Branch\" somefile.cc")
    sys.exit(code)

def update_copyright_in_file(copyright, filename):
    print(filename)
    tmp=tempfile.mkstemp()
    fout=os.fdopen(tmp[0], "w")
    with open(filename) as fin:
        modified=False
        try:
            line=fin.readline()
            line_number=1
            while line:
                match=recd.search(line)
                if match:
                    # Some MaxGui files (namely Vue templates) are forced to
                    # have the license blurb after actual code.
                    if line_number <= line_threshold or filename.endswith('.vue'):
                        next_line=''
                        while True:
                            fout.write(line)
                            next_line=fin.readline()
                            line_number=line_number + 1
                            if next_line:
                                next_match=recd.search(next_line)
                                if next_match:
                                    line=next_line
                                    match=next_match
                                    continue
                                else:
                                    break
                            else:
                                break

                        # 14 is the length of "Copyright (c) " that an existing MariaDB copyright
                        # is assumed to begin with.
                        cutoff=match.span()[0] + 14
                        prefix=line[0:cutoff]
                        suffix=line[cutoff:]
                        if suffix != copyright + "\n":
                            new_line=prefix + copyright + "\n"
                            fout.write(new_line)
                            modified=True
                        if next_line:
                            fout.write(next_line)
                    else:
                        print("WARNING: Ignoring 'Copyright' encountered on line " + str(line_number) + ".")
                        fout.write(line)
                else:
                    fout.write(line)
                line = fin.readline()
                line_number=line_number + 1
        except UnicodeDecodeError:
            print("WARNING: Could not decode file, ignoring.")
            modified=False
        except:
            print("WARNING: Could not read file, ignoring.")
            modified=False

        fin.close()
        fout.close()

        if modified:
            shutil.copyfile(tmp[1], filename)

        os.remove(tmp[1])

def update_copyright_in_dir(copyright, dir):
    for file in os.listdir(dir):
        path=dir + "/" + file
        update_copyright_in_path(copyright, path)

def update_copyright_in_path(copyright, path):
    if not os.path.exists(path):
        print("WARNING: " + path + " does not exist.")
    elif os.path.isdir(path):
        update_copyright_in_dir(copyright, path)
    elif os.path.isfile(path):
        update_copyright_in_file(copyright, path)
    else:
        print("WARNING: '" + path + "' is not a regular file or directory, ignoring.")

def update_copyright_in_paths(copyright, paths):
    for path in paths:
        update_copyright_in_path(copyright, path)

def main(argv):
    name=argv[0]
    try:
        opts, args = getopt.getopt(argv[1:], "hc:")
    except getopt.GetoptError:
        print_usage_and_exit(name, 1)

    copyright=''

    for opt, arg in opts:
        if opt == '-h':
            print_usage_and_exit(name, 0)
        elif opt == '-c':
            copyright = arg

    if copyright == '':
        print_usage_and_exit(name, 1)

    update_copyright_in_paths(copyright, args)

if __name__ == "__main__":
    main(sys.argv)