llm_utility.py

import json, re, ast
from typing import Any

try:
    from json_repair import repair_json
    HAS_JSON_REPAIR = True
except ImportError:
    HAS_JSON_REPAIR = False

def safe_parse_json(raw: str) -> Any:
    """
    Attempt to parse JSON from an LLM response, regardless of format issues.
    Tries multiple strategies until one succeeds.
    Returns Python object (dict/list) or {} if nothing works.
    """

    # 1. Direct parse
    try:
        return json.loads(raw)
    except Exception:
        pass

    # 2. Regex extract array or object
    try:
        match = re.search(r"(\[.*\])", raw, re.DOTALL)
        if match:
            return json.loads(match.group(1))
        match = re.search(r"(\{.*\})", raw, re.DOTALL)
        if match:
            return json.loads(match.group(1))
    except Exception:
        pass

    # 3. Repair JSON if library is available
    if HAS_JSON_REPAIR:
        try:
            repaired = repair_json(raw)
            return json.loads(repaired)
        except Exception:
            pass

    # 4. Try Python literal_eval (handles single quotes)
    try:
        return ast.literal_eval(raw)
    except Exception:
        pass

    # 5. Fallback: return empty
    return {}