POST /entity/<type>/_summarize

Params

Sample requests

Grouped, which is the call worth making:

ARR = {"Content-Type": "application/vnd+shotgun.api3_array+json"}
r = c.post("/entity/versions/_summarize", headers=ARR,
           json={"filters": [["project", "is", {"type": "Project", "id": 70}]],
                 "summary_fields": [{"field": "id", "type": "count"}],
                 "grouping": [{"field": "sg_status_list", "type": "exact", "direction": "asc"}]})

{
  "data": {
    "summaries": { "id": 100 },
    "groups": [
      { "group_name": "na", "group_value": "na", "summaries": { "id": 28 } }
    ]
  }
}

Ungrouped, where the whole body is 45 bytes:

r = c.post("/entity/versions/_summarize", headers=ARR,
           json={"filters": [["project", "is", {"type": "Project", "id": 70}]],
                 "summary_fields": [{"field": "id", "type": "count"}]})

{ "summaries": { "id": 100 }, "groups": [] }

Response codes

Edge cases

• Summarizing an unsummarizable field, image, answers 200 with a 37-byte body and no summary. It does not 400. Test that the key you asked for is in summaries before reading it.
• group_name is the rendered label and group_value the raw one. For a timecode field the rendered form is HH:MM:SS:FF, which is how the frame rate is recovered when no field exposes it.
• One call per field at about 300ms. Over 71 fields that is 21 seconds. Rank a shortlist by fill rate first and summarize only the candidates.