Migrate from Bigeye

# bigeye_to_odcs.py — best-effort translation of Bigeye monitors to ODCS YAML.
# Requires a Bigeye workspace API token. Treat the output as a starting point;
# verify field names against your Bigeye tenant before running at scale.
import os
import requests
import yaml

BIGEYE_API = "https://app.bigeye.com/api/v1"
TOKEN = os.environ["BIGEYE_API_TOKEN"]
WAREHOUSE_ID = int(os.environ.get("BIGEYE_WAREHOUSE_ID", "0"))

session = requests.Session()
session.headers.update({"Authorization": f"apikey {TOKEN}"})

def list_monitors(warehouse_id):
    # The Bigeye metric/monitor list endpoint has evolved between API versions.
    # Adjust the path for your tenant if the response shape differs.
    r = session.get(
        f"{BIGEYE_API}/metrics",
        params={"warehouseId": warehouse_id, "pageSize": 500},
    )
    r.raise_for_status()
    return r.json().get("metrics", [])

# Map common Bigeye metric types to ODCS validity entries.
# Bigeye metric names may differ slightly per tenant; extend this map as needed.
TYPE_MAP = {
    "PERCENT_NULL": "not_null",
    "PERCENT_UNIQUE": "unique",
    "FRESHNESS": "freshness",
    "ROW_COUNT": "row_count",
    "MIN": "range_min",
    "MAX": "range_max",
}

validity = []
for m in list_monitors(WAREHOUSE_ID):
    t = m.get("metricType") or m.get("type")
    col = m.get("column", {}).get("name") if m.get("column") else None
    rule = TYPE_MAP.get(t)
    if rule is None:
        print(f"[warn] unmapped metric type: {t} (name={m.get('name')})")
        continue
    entry = {"rule": rule}
    if col:
        entry["column"] = col
    # Preserve Bigeye-configured thresholds where present.
    if "thresholds" in m:
        entry["thresholds"] = m["thresholds"]
    validity.append(entry)

odcs = {
    "apiVersion": "v3.1.0",
    "kind": "DataContract",
    "name": "bigeye_migration",
    "customProperties": {"anomalyarmor": {"validity": validity}},
}
print(yaml.safe_dump(odcs, sort_keys=False))