Discover metadata
Use discovery when you need to iterate through a dataset rather than look up a known set of IDs. Each dataset is divided into partitions, and each partition is returned in pages.
Endpoints
Section titled “Endpoints”| Method | Route | Dataset |
|---|---|---|
GET |
/api/:store/app/discovery |
Apps. |
GET |
/api/:store/publisher/discovery |
Publishers. |
GET |
/api/:store/sdk/discovery |
SDKs. |
GET |
/api/:store/app_sdks/discovery |
Installed SDKs per app. |
GET |
/api/parent_company/discovery |
Parent companies. |
Query parameters
Section titled “Query parameters”| Parameter | Required | Description |
|---|---|---|
total_partitions |
Yes | Integer from 100 to 10,000. Keep this value unchanged throughout the run. |
partition |
Yes | Integer from 1 to total_partitions. |
page_token |
No | Omit for the first page. For later pages, use the preceding response’s next_page_token. |
Read the response
Section titled “Read the response”The response contains up to 500 records in result_rows. This abbreviated example shows the envelope for app discovery:
{ "result_rows": [ { "id": 366691344, "name": "Packing ~ for holiday and business", "publisher_id": 286879564 } ], "next_page_token": "NEXT_PAGE_TOKEN"}Use next_page_token as page_token on the next request to the same partition. Continue until the response omits next_page_token.
Treat page tokens as opaque strings. Do not reuse them in another partition. You can process partitions concurrently, but each partition’s pages must be fetched in order.
Use the attribute links in the endpoint table for field definitions. ID types can differ from lookup responses; see Stores & identifiers. For ingestion guidance, see Choose an access pattern.