Download an export file
Streams the finished export as CSV, one row per attributed order.
Experiment ID carries the test’s numeric ID and Test Group carries
the test group’s index, so a row joins back to
GET /v1/experiments/{id}. Test group names live on the test, not in
the file.
A completed job’s url is the complete download link. Fetch it as
given: its expires and signature parameters authorize the
request.
The link is the credential. This is the one endpoint that takes no bearer token, so treat the whole URL as a secret. It stops working 7 days after the job completed.
Anything wrong with the request answers the same 404
export_not_found: a bad signature, an expired link, an unknown
job, and an unfinished job all read alike, so an unauthenticated
caller learns nothing about which jobs exist.
Path Parameters
Export job ID (opaque string).
Query Parameters
Unix seconds, from the job's url.
Signature over the id and expiry, from the job's url.
Response
The export, as text/csv. Sent as an attachment named
abconvert-orders-{experiment_id}.csv.
The response is of type string.