Function get_providers

optimade/utils.py:139–210

Retrieve Materials-Consortia providers (from https://providers.optimade.org/v1/links). Fallback order if providers.optimade.org is not available: 1. Try Materials-Consortia/providers on GitHub. 2. Try submodule `providers`' list of providers. 3. Log warning that providers list from

(
    add_mongo_id: bool = False, session: requests.Session | None = None
)

def get_providers(
    add_mongo_id: bool = False, session: requests.Session | None = None
) -> list:
    """Retrieve Materials-Consortia providers (from https://providers.optimade.org/v1/links).

    Fallback order if providers.optimade.org is not available:

    1. Try Materials-Consortia/providers on GitHub.
    2. Try submodule `providers`' list of providers.
    3. Log warning that providers list from Materials-Consortia is not included in the
       `/links`-endpoint.

    Arguments:
        add_mongo_id: Whether to populate the `_id` field of the provider with MongoDB
            ObjectID.
        session: An optional `requests.Session` to use for the request, allowing custom
            HTTP configuration (e.g. proxies). Defaults to the module-level `requests`.

    Returns:
        List of raw JSON-decoded providers including MongoDB object IDs.

    """
    import json

    _get = session.get if session is not None else requests.get

    for provider_list_url in PROVIDER_LIST_URLS:
        try:
            providers = _get(provider_list_url, timeout=10).json()
        except (
            requests.exceptions.ConnectionError,
            requests.exceptions.ConnectTimeout,
            json.JSONDecodeError,
            requests.exceptions.SSLError,
        ):
            pass
        else:
            break
    else:
        try:
            from optimade.server.data import providers  # type: ignore
        except ImportError:
            from optimade.server.logger import LOGGER

            LOGGER.warning(
                """Could not retrieve a list of providers!

    Tried the following resources:

{}
    The list of providers will not be included in the `/links`-endpoint.
""".format("".join([f"    * {_}\n" for _ in PROVIDER_LIST_URLS]))
            )
            return []

    providers_list = []
    for provider in providers.get("data", []):
        # Remove/skip "exmpl"