electric-horses-infra/stacks/eh-search/app/search/static_pages.py

"""Static page registry + tag-to-page bridge.

Hardcoded list of top-level Astro pages that aren't in Directus.
The tag bridge maps blog tags (real, curated by the team) to these pages,
turning the blog content taxonomy into a search vocabulary for the site.
"""
from dataclasses import dataclass, field
from typing import Iterable


@dataclass
class StaticPage:
    slug: str          # URL path, e.g. "/werkstatt"
    title: str         # Display title
    snippet: str       # Description shown in search results
    keywords: list[str] = field(default_factory=list)  # Direct synonyms


# Top-level pages of electric-horses.de that live as Astro routes,
# not in Directus. Order matters for tie-breaking (earlier = preferred).
STATIC_PAGES: list[StaticPage] = [
    StaticPage(
        slug="/werkstatt",
        title="Werkstatt",
        snippet="Spezialwerkstatt fuer E-Roller und E-Motorraeder. Inspektion, Reparatur, Akku-Service.",
        keywords=["werkstatt", "reparatur", "service", "inspektion", "wartung"],
    ),
    StaticPage(
        slug="/ersatzteile",
        title="Ersatzteile",
        snippet="Original-Ersatzteile und Zubehoer fuer E-Roller, E-Motorraeder und E-Scooter.",
        keywords=["ersatzteile", "teile", "zubehoer", "originalteile"],
    ),
    StaticPage(
        slug="/fahrzeuge",
        title="Fahrzeuge",
        snippet="Unser kompletter Bestand an E-Rollern, E-Motorraedern und E-Autos.",
        keywords=["fahrzeuge", "bestand", "angebot", "uebersicht", "alle"],
    ),
    StaticPage(
        slug="/marken",
        title="Marken",
        snippet="Alle Marken im Ueberblick: Askoll, RAY, ZERO, Energica, Nissan und mehr.",
        keywords=["marken", "hersteller", "brands"],
    ),
    StaticPage(
        slug="/vermietung",
        title="Vermietung",
        snippet="E-Motorrad-Vermietung in Wendelstein bei Nuernberg. Tagestouren, Wochenmiete.",
        keywords=["vermietung", "mieten", "leihen", "verleih", "rental"],
    ),
    StaticPage(
        slug="/kontakt",
        title="Kontakt",
        snippet="Kontakt zu Electric Horses: Telefon, E-Mail, Anfahrt, Oeffnungszeiten.",
        keywords=["kontakt", "anfahrt", "adresse", "telefon", "email", "oeffnungszeiten"],
    ),
    StaticPage(
        slug="/ueber-uns",
        title="Ueber uns",
        snippet="Electric Horses ist die E-Mobilitaets-Sparte des Autohaus Richter & Zech.",
        keywords=["ueber uns", "team", "geschichte", "richter zech", "wendelstein"],
    ),
    StaticPage(
        slug="/blog",
        title="Blog",
        snippet="News, Ratgeber und Erfahrungsberichte rund um Elektromobilitaet.",
        keywords=["blog", "news", "artikel", "ratgeber"],
    ),
]


# Tag → Page bridge.
# These tags exist on real blog posts; matching one of them surfaces the
# linked page (in addition to the blog posts that carry the tag).
# Source: actual tag pool extracted from blog_posts.tags.
TAG_TO_PAGE_BRIDGE: dict[str, str] = {
    # ─── Werkstatt-Themen (Technik, Service, Reifen, Akku) ───────
    "akku": "/werkstatt",
    "akkuladung": "/werkstatt",
    "ladedauer": "/werkstatt",
    "reichweite": "/werkstatt",
    "verbrauch": "/werkstatt",
    "bremsen": "/werkstatt",
    "reifen": "/werkstatt",
    "metzeler": "/werkstatt",
    "pirelli": "/werkstatt",
    "heidenau": "/werkstatt",
    "wartung": "/werkstatt",
    "service": "/werkstatt",
    "installation": "/werkstatt",
    "montage": "/werkstatt",
    "werkzeug": "/werkstatt",
    "trittbrettverbreiterung": "/werkstatt",
    "radnabenmotor": "/werkstatt",
    "propilot": "/werkstatt",
    "effizienz": "/werkstatt",
    "ledersitze": "/werkstatt",

    # ─── Beratung & Foerderung (Kontakt) ─────────────────────────
    "versicherung": "/kontakt",
    "foerderung": "/kontakt",
    "foerderungen": "/kontakt",
    "praemie": "/kontakt",
    "preisvorteil": "/kontakt",
    "innovationspraemie": "/kontakt",
    "umweltpraemie": "/kontakt",
    "kosten": "/kontakt",
    "vergleich": "/kontakt",
    "vorfuehrer": "/kontakt",
    "preis": "/kontakt",
    "goelectric": "/kontakt",

    # ─── Fuehrerschein & Legales (Werkstatt-Section "Recht") ─────
    "abe": "/werkstatt",
    "ekfv": "/werkstatt",
    "klassea1": "/werkstatt",
    "klasseb": "/werkstatt",
    "schluesselzahl196": "/werkstatt",
    "fuehrerschein": "/werkstatt",
    "strassenzulassung": "/werkstatt",
    "verkehrsregeln": "/werkstatt",
    "verordnung": "/werkstatt",
    "legal": "/werkstatt",
    "ekickroller": "/werkstatt",
    "ekfv": "/werkstatt",

    # ─── Vermietung-Themen ───────────────────────────────────────
    "touristik": "/vermietung",
    "freizeit": "/vermietung",

    # ─── Marken-Indikatoren (linken auf /marken) ─────────────────
    "marknunabhaengig": "/marken",
}


def search_static_pages(query: str, limit: int = 5) -> list[dict]:
    """Search static pages by title, slug, keywords, snippet substring.

    Returns dicts in the same shape as content search results.
    """
    q = (query or "").strip().lower()
    if not q:
        return []
    results = []
    for page in STATIC_PAGES:
        score = 0.0
        # Exact title match -> very high
        if page.title.lower() == q:
            score = 10.0
        # Title prefix
        elif page.title.lower().startswith(q):
            score = 7.0
        # Slug match (without leading slash)
        elif page.slug.lstrip("/").lower() == q:
            score = 9.0
        # Title substring (only for queries >= 3 chars to avoid noise like 'ze' in 'fahrzeuge')
        elif len(q) >= 3 and q in page.title.lower():
            score = 5.0
        # Keyword exact / prefix
        elif any(kw == q for kw in page.keywords):
            score = 6.0
        elif any(kw.startswith(q) for kw in page.keywords):
            score = 4.0
        # Snippet substring (weakest, also requires >= 3 chars)
        elif len(q) >= 3 and q in page.snippet.lower():
            score = 2.0

        if score > 0:
            results.append({
                "type": "page",
                "title": page.title,
                "slug": page.slug,
                "snippet": page.snippet,
                "score": score,
                "matched_via": "static_page",
            })

    results.sort(key=lambda r: -r["score"])
    return results[:limit]


def get_pages_for_tags(tags: Iterable[str]) -> set[str]:
    """Given an iterable of blog tags, return the set of page slugs they map to."""
    if not tags:
        return set()
    return {TAG_TO_PAGE_BRIDGE[t.lower()] for t in tags if t and t.lower() in TAG_TO_PAGE_BRIDGE}


def page_by_slug(slug: str) -> StaticPage | None:
    for p in STATIC_PAGES:
        if p.slug == slug:
            return p
    return None