# 🔍 data_fetcher.py
from firestore_client import get_firestore_client
from schema_utils import has_field, resolve_field

db = get_firestore_client()

# Utility function to convert Gemini-style filters into Firestore queries
def apply_filters(query_ref, collection_name, filters):
    for f in filters:
        raw_field = f.get("field")
        field = resolve_field(collection_name, raw_field) if raw_field else None
        op = f.get("op", "==")
        value = f.get("value")
        if field and value is not None and has_field(collection_name, field):
            query_ref = query_ref.where(field, op, value)
        # else: silently ignore invalid field for this collection
    return query_ref

def fetch_data_from_firestore(instruction):
    """
    instruction: dict like
    {
      "collections": [
        {
          "name": "participants",
          "filters": [
            {"field": "companyCode", "op": "==", "value": "QTX"},
            {"field": "status", "op": "==", "value": "active"}
          ],
          "limit": 100
        },
        ...
      ]
    }
    """
    results = {}
    for col in instruction.get("collections", []):
        name = col.get("name")
        filters = col.get("filters", [])
        limit = col.get("limit", 50)

        try:
            if name == "users" and any(f["field"] == "uid" for f in filters):
                uid_filter = next(f for f in filters if f["field"] == "uid")
                doc_id = uid_filter["value"]
                doc = db.collection("users").document(doc_id).get()
                if doc.exists:
                    results[name] = [doc.to_dict()]
                else:
                    results[name] = []
            else:
                col_ref = db.collection(name)
                col_ref = apply_filters(col_ref, name, filters)
                docs = col_ref.limit(limit).stream()
                results[name] = [doc.to_dict() for doc in docs]
        except Exception as e:
            results[name] = {"error": str(e)}

    return results