"""Wiki page titles and Gitea filename encoding."""

from __future__ import annotations

import hashlib
import re
from urllib.parse import quote, unquote

from export1c_help.discover import HelpPage

_INVALID = re.compile(r'[\\/:*?"<>|#\[\]]+')
_SPACES = re.compile(r"\s+")
# Keep slug short so percent-encoded form fits in 255-byte FS names
_MAX_SLUG_BYTES = 72


def sanitize_title(title: str) -> str:
    title = _SPACES.sub(" ", title).strip()
    title = _INVALID.sub(" ", title)
    title = _SPACES.sub(" ", title).strip()
    return title or "untitled"


def _truncate_utf8(s: str, max_bytes: int) -> str:
    raw = s.encode("utf-8")
    if len(raw) <= max_bytes:
        return s
    raw = raw[:max_bytes]
    while raw:
        try:
            return raw.decode("utf-8").rstrip("- ")
        except UnicodeDecodeError:
            raw = raw[:-1]
    return "page"


def _short_hash(s: str) -> str:
    return hashlib.sha1(s.encode("utf-8")).hexdigest()[:8]


def wiki_filename(title: str, *, meta_key: str | None = None) -> str:
    """
    Gitea wiki stores non-ASCII page names as percent-encoded «Slug.md».

    Slug = title with spaces → «-», truncated so encoded length ≤ 255.
    """
    slug = sanitize_title(title).replace(" ", "-")
    slug = _truncate_utf8(slug, _MAX_SLUG_BYTES)
    if meta_key:
        # stable disambiguator when truncated
        pass
    encoded = quote(slug, safe="-_.") + ".md"
    if len(encoded) > 255 and meta_key:
        slug = _truncate_utf8(slug, 50) + "-" + _short_hash(meta_key)
        encoded = quote(slug, safe="-_.") + ".md"
    return encoded


def decode_wiki_filename(name: str) -> str:
    base = name[:-3] if name.endswith(".md") else name
    return unquote(base)


def assign_titles(
    pages: list[HelpPage],
    h1_by_meta: dict[str, str | None],
) -> tuple[dict[str, str], dict[str, str]]:
    """Returns (meta_key → title, meta_key → filename)."""
    used_title: set[str] = set()
    used_file: set[str] = set()
    titles: dict[str, str] = {}
    files: dict[str, str] = {}

    for p in pages:
        h1 = h1_by_meta.get(p.meta_key)
        base = sanitize_title(h1 or p.synonym or p.meta_key)
        title = base
        n = 2
        while title.casefold() in used_title:
            if p.is_form and p.form_name and n == 2:
                title = sanitize_title(f"{base} — {p.form_name}")
            elif n == 2:
                title = sanitize_title(f"{base} — {p.object_name}")
            else:
                title = sanitize_title(f"{base} ({n})")
            n += 1
            if n > 5000:
                title = sanitize_title(p.meta_key)
                break

        fname = wiki_filename(title, meta_key=p.meta_key)
        if fname.casefold() in used_file:
            slug = _truncate_utf8(sanitize_title(title).replace(" ", "-"), 50)
            fname = quote(f"{slug}-{_short_hash(p.meta_key)}", safe="-_.") + ".md"
        guard = 0
        while fname.casefold() in used_file:
            fname = quote(f"page-{_short_hash(p.meta_key + str(guard))}", safe="-_.") + ".md"
            guard += 1

        used_title.add(title.casefold())
        used_file.add(fname.casefold())
        titles[p.meta_key] = title
        files[p.meta_key] = fname

    return titles, files