Push from curl

curl -T data.csv https://rows.page. PUT /<id>[/<name>] pushes a new version. Parameters also work as X-Rows-* headers (X-Rows-Title, X-Rows-Id-Column, ...). Plain-text answer unless Accept: application/json or ?json=1. Chunked bodies (curl -T -) are fine.

put/{name}

Request

  • Base URL: https://rows.page
  • URL: https://rows.page/{name}
  • Auth: one of:
    • none
    • HTTP bearer
    • API key in header X-Rows-Token

Path parameters

namestring required

File name, dataset id, or -

Query parameters

format'csv' | 'tsv' | 'json' | 'jsonl' | 'ndjson' | 'parquet'

Decides the parse; otherwise the file extension, then Content-Type, then content sniffing

namestring

File name, e.g. orders.csv

titlestring

Dataset title (max 200)

summarystring

What this is and what to look at (max 4000)

keystring

Upsert one named dataset per account (API key required)

id_columnstring

Column identifying a row across versions, for changed-row diffs

datasetstring

Push a new version of this dataset (needs its token or the owning account)

sourcestring

Where the data came from

runstring

Run id of the producing job

commitstring

Commit sha of the producing code

visibility'public' | 'private'

private needs Pro or Max

json'1' | 'true'

PUT only: answer with JSON instead of plain text

Response

New version, or unchanged: true when the bytes match the latest version

idstring
versioninteger
urlstring
owner_urlstring nullable

url#t=<token>; only on creation. Treat like a password.

tokenstring nullable

Dataset token; only on creation

formatstring
bytesinteger
rowsinteger nullable
rows_exactboolean
expires_atstring date-time nullable
unchangedboolean

Changes

No recorded changes to this operation across all 1 revision of this API.