디자인 포트폴리오 구성·첨삭

#!/usr/bin/env python3"""Validate structural and credibility markers in an HTML portfolio deck."""from __future__ import annotationsimport argparseimport jsonimport reimport sysfrom pathlib import PathREGISTERED = {f"ID{i:02d}" for i in range(1, 19)}PLACEHOLDERS = (    "[PORTFOLIO_TITLE]",    "[DESIGNER_NAME]",    "[PORTFOLIO_META]",    "PORTFOLIO_SLIDES",    "TODO",    "TBD",    "Lorem ipsum",)GENERIC_PLACEHOLDER = re.compile(r"\[[A-Z][A-Z0-9_ -]{2,}\]")def issue(level: str, message: str) -> tuple[str, str]:    return level, messagedef validate(path: Path) -> list[tuple[str, str]]:    text = path.read_text(encoding="utf-8")    findings: list[tuple[str, str]] = []    for token in PLACEHOLDERS:        if token.lower() in text.lower():            findings.append(issue("P1", f"Unresolved placeholder: {token}"))    for token in sorted(set(GENERIC_PLACEHOLDER.findall(text)) - set(PLACEHOLDERS)):        findings.append(issue("P1", f"Unresolved placeholder: {token}"))    preset_match = re.search(r"<html\b[^>]*\bdata-style=[\"']([^\"']+)[\"']", text, re.I)    html_preset = preset_match.group(1) if preset_match else None    if html_preset is None:        findings.append(issue("P1", "The root html element has no data-style preset."))    slides = re.findall(r"<section\b[^>]*class=[\"'][^\"']*\bslide\b[^\"']*[\"'][^>]*>", text, re.I)    if not slides:        findings.append(issue("P1", "No slide sections found."))    layout_counts: dict[str, int] = {}    for index, tag in enumerate(slides, start=1):        match = re.search(r"data-layout=[\"']([^\"']+)[\"']", tag, re.I)        if not match:            findings.append(issue("P1", f"Slide {index} has no data-layout."))            continue        layout = match.group(1)        layout_counts[layout] = layout_counts.get(layout, 0) + 1        if layout not in REGISTERED:            findings.append(issue("P1", f"Slide {index} uses unregistered layout {layout}."))    image_tags = re.findall(r"<img\b[^>]*>", text, re.I)    for index, tag in enumerate(image_tags, start=1):        if not re.search(r"\balt=[\"'][^\"']+[\"']", tag, re.I):            findings.append(issue("P1", f"Image {index} has missing or empty alt text."))        direct_src = re.search(r"(?<!data-)\bsrc=[\"']([^\"']+)[\"']", tag, re.I)        if direct_src and not direct_src.group(1).lower().startswith("data:") and not re.search(            r"(?:generated|ai-)", direct_src.group(1), re.I        ):            findings.append(issue("P2", f"Image {index} loads eagerly; prefer data-src for the lightweight page window."))    generated_refs = re.findall(r"(?:(?:src|data-src)=[\"'][^\"']*(?:generated|ai-)[^\"']*[\"']|AI-assisted|AI generated)", text, re.I)    disclosures = re.findall(r"AI-assisted concept visualization|Not engineering evidence", text, re.I)    if generated_refs and not disclosures:        findings.append(issue("P0", "Generated imagery appears without an AI disclosure."))    generated_paths = re.findall(r"data-src=[\"']([^\"']*images/generated/[^\"']+)[\"']", text, re.I)    direct_generated = re.findall(r"(?<!data-)src=[\"']([^\"']*images/generated/[^\"']+)[\"']", text, re.I)    for ref in direct_generated:        findings.append(issue("P1", f"Generated image bypasses deferred loading: {ref}"))    for ref in sorted(set(generated_paths)):        if Path(ref).suffix.lower() != ".webp":            findings.append(issue("P1", f"Browser-facing generated image is not WebP: {ref}"))        if not (path.parent / ref).is_file():            findings.append(issue("P1", f"Deferred image file is missing: {ref}"))    slide_blocks = re.findall(r"<section\b[^>]*class=[\"'][^\"']*\bslide\b[^\"']*[\"'][^>]*>.*?</section>", text, re.I | re.S)    for index, block in enumerate(slide_blocks, start=1):        if re.search(r"data-src=[\"'][^\"']*images/generated/", block, re.I) and not re.search(            r"AI-assisted|Not engineering evidence|Not research evidence|Not test evidence", block, re.I        ):            findings.append(issue("P0", f"Slide {index} has generated imagery without a nearby disclosure."))    manifest_path = path.parent / "portfolio_manifest.json"    manifest: dict | None = None    if manifest_path.is_file():        try:            loaded_manifest = json.loads(manifest_path.read_text(encoding="utf-8"))            if isinstance(loaded_manifest, dict):                manifest = loaded_manifest            else:                findings.append(issue("P1", "Could not verify HTML against manifest: root must be an object."))        except (json.JSONDecodeError, OSError) as exc:            findings.append(issue("P1", f"Could not verify HTML against manifest: {exc}"))    if manifest is not None:        manifest_preset = manifest.get("portfolio", {}).get("style_preset")        if html_preset and manifest_preset and html_preset != manifest_preset:            findings.append(                issue(                    "P1",                    f"HTML data-style {html_preset!r} does not match manifest style_preset {manifest_preset!r}.",                )            )        if generated_paths:            manifest_paths = {                item.get("file") for item in manifest.get("generated_assets", []) if isinstance(item, dict)            }            missing = sorted(set(generated_paths) - manifest_paths)            stale = sorted(manifest_paths - set(generated_paths))            for ref in missing:                findings.append(issue("P1", f"Generated image is missing from manifest: {ref}"))            for ref in stale:                findings.append(issue("P1", f"Manifest generated asset is not used by HTML: {ref}"))    if len(slides) >= 12 and len(layout_counts) < 7:        findings.append(issue("P2", "A 12+ page case uses fewer than seven distinct layouts."))    sequences = []    for tag in slides:        match = re.search(r"data-layout=[\"'](ID\d{2})[\"']", tag, re.I)        if match:            sequences.append(match.group(1))    for i in range(len(sequences) - 2):        if sequences[i] == sequences[i + 1] == sequences[i + 2]:            findings.append(issue("P2", f"Layout {sequences[i]} repeats three times consecutively."))            break    if "prefers-reduced-motion" not in text:        findings.append(issue("P1", "No reduced-motion CSS handling found."))    if "@media print" not in text:        findings.append(issue("P3", "No print/PDF CSS found."))    if not re.search(r"<title>[^<]+</title>", text, re.I):        findings.append(issue("P1", "Document title is missing."))    return findingsdef main() -> int:    parser = argparse.ArgumentParser(description=__doc__)    parser.add_argument("html", type=Path, help="Path to portfolio index.html")    args = parser.parse_args()    if not args.html.is_file():        print(f"P0: File not found: {args.html}")        return 2    findings = validate(args.html)    for level in ("P0", "P1", "P2", "P3"):        for item_level, message in findings:            if item_level == level:                print(f"{level}: {message}")    blockers = sum(1 for level, _ in findings if level in {"P0", "P1"})    print(f"Checked {args.html}: {len(findings)} finding(s), {blockers} blocker(s).")    return 1 if blockers else 0if __name__ == "__main__":    sys.exit(main())