Errors

Every error raised by typesafer inherits from typesafer_error, so you can catch them all with tryCatch(..., typesafer_error = ...). More specific classes:

• typesafer_error_input: invalid arguments, detected before any request is sent.

• typesafer_error_auth: the API key is missing, or the API rejected it (HTTP 401).

• typesafer_error_http: any unsuccessful HTTP response. Has fields status, response_body (parsed JSON, text, or NULL), request_id, and resp (the httr2::response()). Subclasses:

  • typesafer_error_auth: HTTP 401.

  • typesafer_error_validation: HTTP 422. The details field is a tibble with one row per offending field (field, message, type).

  • typesafer_error_rate_limit: HTTP 429 (rate limited) or 529 (overloaded), after retries are exhausted.

  • typesafer_error_server: other 5xx responses, after retries are exhausted.

• typesafer_error_connection: the request never got a response (DNS, connection, or timeout failure), after retries are exhausted.

Requests that fail with 408, 429, or 5xx, or with a connection failure, are retried with exponential backoff (0.5s doubling to at most 5s, with jitter), within a 30 second budget. By default a request is tried up to 3 times; change that with the max_tries argument. retry-after-ms and retry-after response headers are honored.

Examples

tryCatch(
  ts_score("How urgent?", "high"),
  typesafer_error_input = function(e) conditionMessage(e)
)
#> [1] "\033[1m\033[22m`criteria` must have between 2 and 10 levels, not 1."