1"""
2Command line tool to validate Office document XML files against XSD schemas and tracked changes.
3 
4Usage:
5    python validate.py <path> [--original <original_file>] [--auto-repair] [--author NAME]
6 
7The first argument can be either:
8- An unpacked directory containing the Office document XML files
9- A packed Office file (.docx/.pptx/.xlsx) which will be unpacked to a temp directory
10 
11Auto-repair fixes:
12- paraId/durableId values that exceed OOXML limits
13- Missing xml:space="preserve" on w:t elements with whitespace
14"""
15 
16import argparse
17import sys
18import tempfile
19import zipfile
20from pathlib import Path
21 
22from validators import DOCXSchemaValidator, PPTXSchemaValidator, RedliningValidator
23 
24 
25def main():
26    parser = argparse.ArgumentParser(description="Validate Office document XML files")
27    parser.add_argument(
28        "path",
29        help="Path to unpacked directory or packed Office file (.docx/.pptx/.xlsx)",
30    )
31    parser.add_argument(
32        "--original",
33        required=False,
34        default=None,
35        help="Path to original file (.docx/.pptx/.xlsx). If omitted, all XSD errors are reported and redlining validation is skipped.",
36    )
37    parser.add_argument(
38        "-v",
39        "--verbose",
40        action="store_true",
41        help="Enable verbose output",
42    )
43    parser.add_argument(
44        "--auto-repair",
45        action="store_true",
46        help="Automatically repair common issues (hex IDs, whitespace preservation)",
47    )
48    parser.add_argument(
49        "--author",
50        default="Claude",
51        help="Author name for redlining validation (default: Claude)",
52    )
53    args = parser.parse_args()
54 
55    path = Path(args.path)
56    assert path.exists(), f"Error: {path} does not exist"
57 
58    original_file = None
59    if args.original:
60        original_file = Path(args.original)
61        assert original_file.is_file(), f"Error: {original_file} is not a file"
62        assert original_file.suffix.lower() in [".docx", ".pptx", ".xlsx"], (
63            f"Error: {original_file} must be a .docx, .pptx, or .xlsx file"
64        )
65 
66    file_extension = (original_file or path).suffix.lower()
67    assert file_extension in [".docx", ".pptx", ".xlsx"], (
68        f"Error: Cannot determine file type from {path}. Use --original or provide a .docx/.pptx/.xlsx file."
69    )
70 
71    if path.is_file() and path.suffix.lower() in [".docx", ".pptx", ".xlsx"]:
72        temp_dir = tempfile.mkdtemp()
73        with zipfile.ZipFile(path, "r") as zf:
74            zf.extractall(temp_dir)
75        unpacked_dir = Path(temp_dir)
76    else:
77        assert path.is_dir(), f"Error: {path} is not a directory or Office file"
78        unpacked_dir = path
79 
80    match file_extension:
81        case ".docx":
82            validators = [
83                DOCXSchemaValidator(unpacked_dir, original_file, verbose=args.verbose),
84            ]
85            if original_file:
86                validators.append(
87                    RedliningValidator(unpacked_dir, original_file, verbose=args.verbose, author=args.author)  
88                )
89        case ".pptx":
90            validators = [
91                PPTXSchemaValidator(unpacked_dir, original_file, verbose=args.verbose),
92            ]
93        case _:
94            print(f"Error: Validation not supported for file type {file_extension}")
95            sys.exit(1)
96 
97    if args.auto_repair:
98        total_repairs = sum(v.repair() for v in validators)
99        if total_repairs:
100            print(f"Auto-repaired {total_repairs} issue(s)")
101 
102    success = all(v.validate() for v in validators)
103 
104    if success:
105        print("All validations PASSED!")
106 
107    sys.exit(0 if success else 1)
108 
109 
110if __name__ == "__main__":
111    main()
112