Find replace in files recursively with regular expressions

find-replace-in-files-regex.py

#!/usr/bin/env python3
#
# Example Usage:
# -------------------------------------------------------------------------------
# modify dates from 30-10-2024 to 2024-10-30
#    find-replace-in-files-regex.py -g "*.html" "(\d\d)-(\d\d)-(\d\d\d\d)" "\3-\2-\1"
# -------------------------------------------------------------------------------


import argparse
import re
import shutil
import sys
from pathlib import Path

Default = "\033[39m"
Black = "\033[30m"
Red = "\033[31m"
Green = "\033[32m"
Yellow = "\033[33m"
Blue = "\033[34m"
Magenta = "\033[35m"
Cyan = "\033[36m"
LightGray = "\033[37m"
DarkGray = "\033[90m"
LightRed = "\033[91m"
LightGreen = "\033[92m"
LightYellow = "\033[93m"
LightBlue = "\033[94m"
LightMagenta = "\033[95m"
LightCyan = "\033[96m"
White = "\033[97m"
Reset = "\033[0m"

def parse_args():
    parser = argparse.ArgumentParser(description="Recursively search and replace text in files.")
    parser.add_argument("search", help="The search regex pattern. use -m to enable multiline patterns.")
    parser.add_argument("replace", help="The replacement regex pattern.")
    parser.add_argument("--path", "-p", default=".", help="Folder path to start the search (default=%(default)s).")
    parser.add_argument("--glob", "-g", default="*.*", help="Glob pattern for files to search (default=%(default)s).")
    parser.add_argument("--force", "-f", action="store_true", help="Replace text in files. (%(default)s)", default=False)
    exclusive_group = parser.add_mutually_exclusive_group()
    exclusive_group.add_argument("--literal", "-l", action="store_true", help="Search/replace are literal strings, not regexes (%(default)s)", default=False)
    exclusive_group.add_argument("--multiline", "-m", action="store_true", help="Enable multiline text matching. (%(default)s)", default=False)
    parser.add_argument('--backup', '-b',
                        action='store_true',
                        help='Create backup files',
                        default=False)
    return parser.parse_args()

  
def process_file(
    conf,
    file_path: Path,
    search_pattern: re.Pattern,
    search_pattern_colorized: re.Pattern
):

    try:
        content: str = file_path.read_text(encoding="utf-8")

        if conf.literal:
            if conf.search not in content:
                return

        matches = list(re.finditer(search_pattern, content))
        if not matches:
            return

        # Display matches
        print(f"{'-' * 60}\n💡{file_path}\n{'-' * 60}")

        for match in matches:
            # Get the entire line where the match occurred
            start_of_line = content.rfind('\n', 0, match.start()) + 1
            end_of_line = content.find('\n', match.end())
            end_of_line = end_of_line if end_of_line != -1 else len(content)
            line = content[start_of_line:end_of_line]

            # Highlight the match within the line
            # note: group() and group(0) are equivalent. They both return the entire matched string

            matched_search_text = match.group(0)

            # Highlight the matched text within a line.
            highlighted_line_before = line.replace(matched_search_text, rf'{Red}{matched_search_text}{Reset}').strip()
            print(f"  {Magenta}BEFORE:{Reset}\n\t{highlighted_line_before}")

            highlighted_line_after =search_pattern.sub(rf'{LightGreen}{args.replace}{Reset}', line).strip()
            print(f"  {Cyan}AFTER:{Reset}\n\t{highlighted_line_after}")

        if conf.force:
            new_content = search_pattern.sub(conf.replace, content)

            if new_content != content:  # Only write if there's a change
                print(f"Writing changes: {file_path}")

                if conf.backup:
                    backup_path: Path = file_path.with_suffix(f'{file_path.suffix}.bak')

                    while backup_path.exists():
                        backup_path = backup_path.with_suffix(f'{backup_path.suffix}.bak')
                    shutil.copyfile(file_path, backup_path)

                file_path.write_text(new_content, encoding='utf-8')
            else:
                print(f"No changes applied to: {file_path}")

    except Exception as e:
        print(f"Error processing {file_path}: {e}", file=sys.stderr)


if __name__ == "__main__":
    args = parse_args()
    print(args)

    # multiline is mutually exclusive with literal
    search_patt: re.Pattern
    search_patt_colorized: re.Pattern

    if args.multiline:
        search_patt = re.compile(args.search, flags=re.MULTILINE | re.DOTALL)
        search_patt_colorized = re.compile("({args.search})", flags=re.MULTILINE | re.DOTALL)
    else:
        if args.literal:
            search_patt = re.compile(re.escape(args.search))
            search_patt_colorized = re.compile(re.escape("({args.search})"))
        else:
            search_patt = re.compile(args.search)
            search_patt_colorized = re.compile("({args.search})")

    root_dir = Path(args.path)

    if not args.force:
        print('\n\nDry run mode! No files will be changed. Use -f to force modifications.')

    for path in root_dir.rglob(args.glob):
        if path.is_file():
            process_file(args, path, search_patt, search_patt_colorized)