Update redact_generic.py

redact_generic.py

@@ -1,16 +1,16 @@
 #!/usr/bin/env python3
 """
-通用CSV脱敏脚本
+Generic CSV Redaction Script
 
-用法示例:
+Usage Examples:
   python redact_generic.py input.csv
   python redact_generic.py input.csv --output out.csv --columns log_message
   python redact_generic.py input.csv --full-mask
 
-特性:
-- 默认一组正则规则（URL, IP, email, 手机号, 身份证, 长数字, 设备ID, App名）
-- 支持按列脱敏（通过列名或索引），默认对所有文本列处理
-- 提供部分掩码（默认）或完全替换选项
+Features:
+- Default set of regex rules (URL, IP, email, phone number, ID card, long number, device ID, App name)
+- Supports column-specific redaction (by column name or index), defaults to processing all text columns
+- Provides partial masking (default) or full replacement options
 """
 import re
 import csv
@@ -20,7 +20,7 @@ from typing import List, Pattern
 
 
 def build_rules(full_mask: bool = False):
-    # 返回 (name, pattern, repl_or_callable)
+    # Returns (name, pattern, repl_or_callable)
     rules = []
 
     # URLs
@@ -124,22 +124,22 @@ def process_csv(input_path: Path, output_path: Path, columns: List[str], full_ma
 
 
 def main():
-    ap = argparse.ArgumentParser(description='通用CSV脱敏工具')
-    ap.add_argument('input', help='输入CSV文件')
-    ap.add_argument('--output', '-o', help='输出CSV文件（默认与输入同目录，后缀 .redacted.csv）')
-    ap.add_argument('--columns', '-c', nargs='+', help='要脱敏的列名或列索引（索引从0开始）。不指定则对所有列应用')
-    ap.add_argument('--full-mask', action='store_true', help='使用完全替换而不是部分掩码（对手机号/身份证等）')
+    ap = argparse.ArgumentParser(description='Generic CSV Redaction Tool')
+    ap.add_argument('input', help='Input CSV file')
+    ap.add_argument('--output', '-o', help='Output CSV file (default in the same directory as input with .redacted.csv suffix)')
+    ap.add_argument('--columns', '-c', nargs='+', help='Column names or indices to redact (indices start from 0). If not specified, apply to all columns')
+    ap.add_argument('--full-mask', action='store_true', help='Use full replacement instead of partial masking (for phone numbers, ID cards, etc.)')
 
     args = ap.parse_args()
     inp = Path(args.input)
     if not inp.exists():
-        print('输入文件不存在:', inp)
+        print('Input file does not exist:', inp)
         return
 
     out = Path(args.output) if args.output else inp.with_suffix('.redacted.csv')
 
     process_csv(inp, out, args.columns or [], args.full_mask)
-    print('已写脱敏文件：', out)
+    print('Redacted file written to:', out)
 
 
 if __name__ == '__main__':