from __future__ import annotations

import json
import subprocess
from datetime import datetime


def list_objects(gcs_uri: str) -> list[dict]:
    """List objects under a GCS prefix as ``[{name, size, last_modified_ms}]`` via ``gcloud``.

    ``name`` is the object path relative to the bucket; ``last_modified_ms`` is epoch
    milliseconds (from the object's ``update_time``). ``gcloud storage objects list`` only
    recurses under a sub-prefix with the ``**`` recursive wildcard; a bare prefix matches an
    object named exactly that and returns nothing. Shared by the BigQuery extract (manifest)
    and the GCS load (external discovery)."""
    prefix = gcs_uri if gcs_uri.endswith("/") else gcs_uri + "/"
    completed = subprocess.run(
        ["gcloud", "storage", "objects", "list", f"{prefix}**", "--format=json"],
        check=True,
        capture_output=True,
        text=True,
    )
    items = json.loads(completed.stdout or "[]")
    return [
        {"name": item["name"], "size": int(item.get("size", 0)), "last_modified_ms": _epoch_ms(item.get("update_time"))}
        for item in items
    ]


def _epoch_ms(value: str | None) -> int | None:
    if not value:
        return None
    text = value.replace("Z", "+00:00")
    try:
        parsed = datetime.fromisoformat(text)
    except ValueError:
        parsed = datetime.strptime(value, "%Y-%m-%dT%H:%M:%S%z")
    return int(parsed.timestamp() * 1000)