#!/usr/bin/env python3
"""Lightweight HTML audit for common a11y/markup issues.

Scans both *.html and *.php files. PHP blocks are stripped before HTML parsing
so they don't confuse html.parser.
"""

from __future__ import annotations

import re
from html.parser import HTMLParser
from pathlib import Path

ROOT = Path(__file__).resolve().parents[1]

SKIP_DIRS = {"partials", ".git", ".agents", ".claude", "tools",
             "pub", "docs", "locales", "admin", "api"}
SKIP_FILES = {
    "test.html",
    "test_jp.html",
    "test_zh.html",
    "startpage/test.html",
    "vibe/admin_vibe.php",
    "changelog/admin.php",
}

# Replace <?php ... ?> and <?= ... ?> blocks with a placeholder before HTML
# parsing. Using a non-empty placeholder (rather than "") preserves attributes
# like title="<?= ... ?>" so they don't look empty to the audit.
RE_PHP = re.compile(r"<\?.*?\?>", re.S)
PHP_PLACEHOLDER = "phpval"

# IDs that legitimately appear multiple times in a file because only one
# instance is rendered at runtime (e.g. branched by `if ($lang === 'en')`).
# Map of file -> {ids} to suppress the duplicate-ids check for.
DUP_ID_ALLOWLIST: dict[str, set[str]] = {
    "index.php": {"guestbook-form-iframe", "guestbook-comments-iframe"},
}


class AuditParser(HTMLParser):
    def __init__(self) -> None:
        super().__init__()
        self.ids: dict[str, int] = {}
        self.duplicate_ids: set[str] = set()
        self.missing_alt: list[str] = []
        self.missing_iframe_title: list[str] = []
        self.blank_rel: list[str] = []

    def handle_starttag(self, tag: str, attrs: list[tuple[str, str | None]]) -> None:
        attr_map = {k.lower(): (v or "") for k, v in attrs}

        if "id" in attr_map:
            ident = attr_map["id"]
            if ident:
                if ident in self.ids:
                    self.duplicate_ids.add(ident)
                self.ids[ident] = self.ids.get(ident, 0) + 1

        if tag == "img":
            if "alt" not in attr_map:
                src = attr_map.get("src", "")
                self.missing_alt.append(src)

        if tag == "iframe":
            if not attr_map.get("title", ""):
                src = attr_map.get("src", "")
                self.missing_iframe_title.append(src)

        if tag == "a":
            if attr_map.get("target", "") == "_blank":
                rel = attr_map.get("rel", "")
                if "noopener" not in rel:
                    href = attr_map.get("href", "")
                    self.blank_rel.append(href)


def main() -> int:
    issues = []

    files: list[Path] = []
    for ext in ("*.html", "*.php"):
        files.extend(ROOT.rglob(ext))

    for src_file in files:
        if any(part in SKIP_DIRS for part in src_file.parts):
            continue
        rel = src_file.relative_to(ROOT).as_posix()
        if rel in SKIP_FILES:
            continue

        text = src_file.read_text(encoding="utf-8", errors="ignore")
        text = RE_PHP.sub(PHP_PLACEHOLDER, text)

        parser = AuditParser()
        parser.feed(text)

        dup_ids = parser.duplicate_ids - DUP_ID_ALLOWLIST.get(rel, set())
        if dup_ids:
            issues.append((rel, "duplicate-ids", sorted(dup_ids)))
        if parser.missing_alt:
            issues.append((rel, "img-missing-alt", parser.missing_alt))
        if parser.missing_iframe_title:
            issues.append((rel, "iframe-missing-title", parser.missing_iframe_title))
        if parser.blank_rel:
            issues.append((rel, "target-blank-missing-noopener", parser.blank_rel))

    if not issues:
        print("OK: no audit issues found")
        return 0

    print("HTML audit issues:")
    for rel, kind, items in issues:
        print(f"- {rel}: {kind}")
        for item in items[:10]:
            print(f"  - {item}")
        if len(items) > 10:
            print(f"  - ... ({len(items) - 10} more)")
    return 1


if __name__ == "__main__":
    raise SystemExit(main())