recipe-importer/app/scraper.py

"""Recipe scraper — parses Hungarian recipe sites into a structured dict.

Each supported site has a parser registered via _PARSERS.
Unsupported sites fall back to generic schema.org / og-tag extraction.
"""

import re
import requests
from bs4 import BeautifulSoup

_HEADERS = {
    "User-Agent": "RecipeImporter/1.0 (Hungarian recipe scraper)",
    "Accept-Language": "hu-HU,hu;q=0.9,en;q=0.5",
}

# Maps a substring of the hostname to a parser function.
# Order matters: first match wins.
_PARSERS: list[tuple[str, "callable"]] = []


def _register(host_substring: str):
    """Decorator: register a parser for URLs whose hostname contains *host_substring*."""
    def decorator(fn):
        _PARSERS.append((host_substring, fn))
        return fn
    return decorator


# ---------------------------------------------------------------------------
# Public API
# ---------------------------------------------------------------------------


def scrape(url: str) -> dict:
    """Fetch *url* and return a recipe dict.

    Returns::

        {
            "title": str,
            "description": str,
            "image_url": str | None,
            "ingredients": [{"quantity": str, "unit": str, "food": str, "extra": str}, ...],
            "instructions": [str, ...],
            "tags": [str, ...],
            "original_url": str,
        }

    Raises ValueError on unsupported sites or parse failures.
    """
    resp = requests.get(url, headers=_HEADERS, timeout=30)
    resp.raise_for_status()
    resp.encoding = resp.apparent_encoding or "utf-8"
    soup = BeautifulSoup(resp.text, "lxml")

    host = _host(url)
    for substring, parser in _PARSERS:
        if substring in host:
            return parser(soup, url)

    # Fallback: try generic schema.org / og-tag extraction
    return _parse_generic(soup, url)


def supported_sites() -> list[str]:
    """Return list of supported site hostname substrings."""
    return [s for s, _ in _PARSERS]


# ---------------------------------------------------------------------------
# mindmegette.hu
# ---------------------------------------------------------------------------


@_register("mindmegette")
def _parse_mindmegette(soup: BeautifulSoup, url: str) -> dict:
    title = _og(soup, "og:title") or _text(soup.find("title"))
    # Strip " | Mindmegette.hu" suffix
    if title:
        title = re.sub(r"\s*\|\s*Mindmegette\.hu$", "", title).strip()

    description = _og(soup, "og:description") or ""
    image_url = _og(soup, "og:image")

    # --- Ingredients ---
    # Multiple div.ingredients containers may exist (one per group).
    # Group title: <strong class="ingredients-group">A habaráshoz:</strong>
    ingredients = []
    for ing_container in soup.find_all("div", class_="ingredients"):
        # Check for a group title
        group_el = ing_container.find("strong", class_="ingredients-group")
        group_name = _text(group_el).rstrip(":").strip() if group_el else ""
        if group_name:
            ingredients.append({"group": group_name})

        for row in ing_container.find_all("div", class_="ingredients-meta"):
            # Actual HTML: <strong>qty</strong> <span>unit</span>
            #              <a class="ingredients-link">name</a> <small>(extra)</small>
            qty_el = row.find("strong")
            unit_el = None
            for sp in row.find_all("span"):
                if not sp.get("class"):
                    unit_el = sp
                    break
            name_el = row.find("a", class_="ingredients-link")
            extra_el = row.find("small") or row.find("span", class_="extra")

            qty = _text(qty_el)
            unit = _text(unit_el)
            food = _text(name_el)
            extra = _text(extra_el).strip("() ")

            if not food:
                # Fallback: grab whole row text
                food = row.get_text(separator=" ", strip=True)

            if food:
                ingredients.append({
                    "quantity": qty,
                    "unit": unit,
                    "food": food,
                    "extra": extra,
                })

    # --- Instructions ---
    instructions = []
    wysiwyg = soup.find("mindmegette-wysiwyg-box")
    if wysiwyg:
        for li in wysiwyg.find_all("li"):
            txt = _text(li)
            if txt:
                instructions.append(txt)
    # Fallback: look for block-content divs
    if not instructions:
        for div in soup.find_all("div", class_="block-content"):
            ol = div.find("ol")
            if ol:
                for li in ol.find_all("li"):
                    txt = _text(li)
                    if txt:
                        instructions.append(txt)

    # --- Tags ---
    tags = []
    tag_wrapper = soup.select_one("div.desktop-wrapper")
    if tag_wrapper:
        for a in tag_wrapper.select("a.tag"):
            tag_text = a.get_text(strip=True)
            if tag_text:
                tags.append(tag_text)

    return {
        "title": title or "Ismeretlen recept",
        "description": description,
        "image_url": image_url,
        "ingredients": ingredients,
        "instructions": instructions,
        "tags": tags,
        "original_url": url,
    }


# ---------------------------------------------------------------------------
# Generic fallback (og-tags + schema.org microdata)
# ---------------------------------------------------------------------------


def _parse_generic(soup: BeautifulSoup, url: str) -> dict:
    title = _og(soup, "og:title") or _text(soup.find("title")) or "Ismeretlen recept"
    description = _og(soup, "og:description") or ""
    image_url = _og(soup, "og:image")

    ingredients = []
    instructions = []
    tags = []

    # Try schema.org JSON-LD
    for script in soup.find_all("script", type="application/ld+json"):
        try:
            import json
            data = json.loads(script.string or "")
            if isinstance(data, list):
                data = data[0]
            if data.get("@type") == "Recipe":
                for line in data.get("recipeIngredient", []):
                    ingredients.append({
                        "quantity": "", "unit": "", "food": line, "extra": "",
                    })
                raw_instructions = data.get("recipeInstructions", [])
                for item in raw_instructions:
                    if isinstance(item, str):
                        instructions.append(item)
                    elif isinstance(item, dict):
                        instructions.append(item.get("text", ""))
                # Extract keywords
                kw = data.get("keywords", "")
                if isinstance(kw, str):
                    tags = [k.strip() for k in kw.split(",") if k.strip()]
                elif isinstance(kw, list):
                    tags = [str(k).strip() for k in kw if str(k).strip()]
                break
        except (json.JSONDecodeError, TypeError, AttributeError):
            continue

    return {
        "title": title,
        "description": description,
        "image_url": image_url,
        "ingredients": ingredients,
        "instructions": instructions,
        "tags": tags,
        "original_url": url,
    }


# ---------------------------------------------------------------------------
# Helpers
# ---------------------------------------------------------------------------


def _host(url: str) -> str:
    from urllib.parse import urlparse
    return urlparse(url).hostname or ""


def _og(soup: BeautifulSoup, prop: str) -> str | None:
    tag = soup.find("meta", property=prop)
    if tag and tag.get("content"):
        return tag["content"]
    return None


def _text(el) -> str:
    if el is None:
        return ""
    return el.get_text(strip=True)