Nye-TeeOff/backend/scrape_utils.py

import json
from typing import Any, Awaitable, Callable

ProgressCallback = Callable[[dict[str, Any]], Awaitable[None]]


def exclude_discontinued_facilities_clause(facility_table: str = "facilities") -> str:
    return f"""
      AND (
        NOT EXISTS (
          SELECT 1
          FROM courses course_filter
          WHERE course_filter.facility_id = {facility_table}.id
        )
        OR EXISTS (
          SELECT 1
          FROM courses course_filter
          WHERE course_filter.facility_id = {facility_table}.id
            AND COALESCE(course_filter.status, 'ukjent') <> 'nedlagt'
        )
      )
    """


async def emit_progress(progress_callback: ProgressCallback | None, **payload: Any) -> None:
    if progress_callback is None:
        return
    await progress_callback(payload)


def make_progress_event(
    *,
    facility_id: int | None,
    facility_name: str,
    outcome: str,
    message: str,
    processed: int,
    total: int,
) -> dict[str, Any]:
    return {
        "facility_id": facility_id,
        "facility_name": facility_name,
        "outcome": outcome,
        "message": message,
        "processed": processed,
        "total": total,
    }


def parse_llm_json(raw_response: str) -> Any:
    text = (raw_response or "").strip()
    if not text:
        raise ValueError("Tomt svar fra modellen.")

    if text.startswith("```"):
        lines = text.splitlines()
        if lines:
            lines = lines[1:]
        if lines and lines[-1].strip().startswith("```"):
            lines = lines[:-1]
        text = "\n".join(lines).strip()

    candidates = [text]

    first_obj = text.find("{")
    last_obj = text.rfind("}")
    if first_obj != -1 and last_obj != -1 and last_obj > first_obj:
        candidates.append(text[first_obj:last_obj + 1])

    first_arr = text.find("[")
    last_arr = text.rfind("]")
    if first_arr != -1 and last_arr != -1 and last_arr > first_arr:
        candidates.append(text[first_arr:last_arr + 1])

    seen: set[str] = set()
    for candidate in candidates:
        candidate = candidate.strip()
        if not candidate or candidate in seen:
            continue
        seen.add(candidate)
        try:
            parsed = json.loads(candidate)
            if isinstance(parsed, str):
                nested = parsed.strip()
                if nested.startswith("{") or nested.startswith("["):
                    return json.loads(nested)
            return parsed
        except json.JSONDecodeError:
            continue

    raise ValueError("Klarte ikke å tolke gyldig JSON fra modellsvaret.")