1"""Pack a directory into a DOCX, PPTX, or XLSX file.
2 
3Validates with auto-repair, condenses XML formatting, and creates the Office file.
4 
5Usage:
6    python pack.py <input_directory> <output_file> [--original <file>] [--validate true|false]
7 
8Examples:
9    python pack.py unpacked/ output.docx --original input.docx
10    python pack.py unpacked/ output.pptx --validate false
11"""
12 
13import argparse
14import sys
15import shutil
16import tempfile
17import zipfile
18from pathlib import Path
19 
20import defusedxml.minidom
21 
22from validators import DOCXSchemaValidator, PPTXSchemaValidator, RedliningValidator
23 
24def pack(
25    input_directory: str,
26    output_file: str,
27    original_file: str | None = None,
28    validate: bool = True,
29    infer_author_func=None,
30) -> tuple[None, str]:
31    input_dir = Path(input_directory)
32    output_path = Path(output_file)
33    suffix = output_path.suffix.lower()
34 
35    if not input_dir.is_dir():
36        return None, f"Error: {input_dir} is not a directory"
37 
38    if suffix not in {".docx", ".pptx", ".xlsx"}:
39        return None, f"Error: {output_file} must be a .docx, .pptx, or .xlsx file"
40 
41    if validate and original_file:
42        original_path = Path(original_file)
43        if original_path.exists():
44            success, output = _run_validation(
45                input_dir, original_path, suffix, infer_author_func
46            )
47            if output:
48                print(output)
49            if not success:
50                return None, f"Error: Validation failed for {input_dir}"
51 
52    with tempfile.TemporaryDirectory() as temp_dir:
53        temp_content_dir = Path(temp_dir) / "content"
54        shutil.copytree(input_dir, temp_content_dir)
55 
56        for pattern in ["*.xml", "*.rels"]:
57            for xml_file in temp_content_dir.rglob(pattern):
58                _condense_xml(xml_file)
59 
60        output_path.parent.mkdir(parents=True, exist_ok=True)
61        with zipfile.ZipFile(output_path, "w", zipfile.ZIP_DEFLATED) as zf:
62            for f in temp_content_dir.rglob("*"):
63                if f.is_file():
64                    zf.write(f, f.relative_to(temp_content_dir))
65 
66    return None, f"Successfully packed {input_dir} to {output_file}"
67 
68 
69def _run_validation(
70    unpacked_dir: Path,
71    original_file: Path,
72    suffix: str,
73    infer_author_func=None,
74) -> tuple[bool, str | None]:
75    output_lines = []
76    validators = []
77 
78    if suffix == ".docx":
79        author = "Claude"
80        if infer_author_func:
81            try:
82                author = infer_author_func(unpacked_dir, original_file)
83            except ValueError as e:
84                print(f"Warning: {e} Using default author 'Claude'.", file=sys.stderr)
85 
86        validators = [
87            DOCXSchemaValidator(unpacked_dir, original_file),
88            RedliningValidator(unpacked_dir, original_file, author=author),
89        ]
90    elif suffix == ".pptx":
91        validators = [PPTXSchemaValidator(unpacked_dir, original_file)]
92 
93    if not validators:
94        return True, None
95 
96    total_repairs = sum(v.repair() for v in validators)
97    if total_repairs:
98        output_lines.append(f"Auto-repaired {total_repairs} issue(s)")
99 
100    success = all(v.validate() for v in validators)
101 
102    if success:
103        output_lines.append("All validations PASSED!")
104 
105    return success, "\n".join(output_lines) if output_lines else None
106 
107 
108def _condense_xml(xml_file: Path) -> None:
109    try:
110        with open(xml_file, encoding="utf-8") as f:
111            dom = defusedxml.minidom.parse(f)
112 
113        for element in dom.getElementsByTagName("*"):
114            if element.tagName.endswith(":t"):
115                continue
116 
117            for child in list(element.childNodes):
118                if (
119                    child.nodeType == child.TEXT_NODE
120                    and child.nodeValue
121                    and child.nodeValue.strip() == ""
122                ) or child.nodeType == child.COMMENT_NODE:
123                    element.removeChild(child)
124 
125        xml_file.write_bytes(dom.toxml(encoding="UTF-8"))
126    except Exception as e:
127        print(f"ERROR: Failed to parse {xml_file.name}: {e}", file=sys.stderr)
128        raise
129 
130 
131if __name__ == "__main__":
132    parser = argparse.ArgumentParser(
133        description="Pack a directory into a DOCX, PPTX, or XLSX file"
134    )
135    parser.add_argument("input_directory", help="Unpacked Office document directory")
136    parser.add_argument("output_file", help="Output Office file (.docx/.pptx/.xlsx)")
137    parser.add_argument(
138        "--original",
139        help="Original file for validation comparison",
140    )
141    parser.add_argument(
142        "--validate",
143        type=lambda x: x.lower() == "true",
144        default=True,
145        metavar="true|false",
146        help="Run validation with auto-repair (default: true)",
147    )
148    args = parser.parse_args()
149 
150    _, message = pack(
151        args.input_directory,
152        args.output_file,
153        original_file=args.original,
154        validate=args.validate,
155    )
156    print(message)
157 
158    if "Error" in message:
159        sys.exit(1)
160