Data API

Objects

• Batch: one file loaded into one data set. Status: staged → applied → undone, or held (waits for a person), rejected, discarded. Its rows are the diff: new, changed (old → new), unchanged and rejected (with codes).
• Bundle: a folder (.zip) of feed files, each its own batch, applied in start-of-day order.
• Feed, feed run (one business date's file), schedule, start of day (the chain of steps), data-quality issue.

Load a file

curl -X POST "https://vestoris.com/api/v1/data/prices?filename=prices.csv&apply=auto" \
  -H "Authorization: Bearer $VESTORIS_TOKEN" -H "Content-Type: application/octet-stream" \
  --data-binary @prices.csv

The body is the file itself (CSV, TXT, XLSX or Vestoris XML; gzip or zip). The token needs data:load:<data set> for that data set. The answer is the batch: 201 staged or applied, 422 rejected (the batch with its row codes is in the body). Without apply=auto a person reviews the diff and applies it (POST /batches/{id}/apply), or you apply it yourself with data:write. File uploads are deduplicated by content, so they need no Idempotency-Key.

Typical sequence

1. Load: POST /data/{dataset} (or push to a feed: POST /feeds/{id}/files, or a folder: POST /feeds/bundles).
2. Check: GET /batches/{id} and GET /batches/{id}/rows?status=rejected.
3. Apply, discard or undo: POST /batches/{id}/apply, /discard, /undo.
4. Follow the morning: GET /start-of-day, GET /feeds/runs, webhooks feed.run.updated and sod.status.

The formats and error codes are in the Data Feed Reference.