Guide

Importing data

Synchronous, batched and job-based imports, with the limits and response shapes for each.

Three modes, chosen by file size and whether you need to wait for the result.

Mode	File size	Blocking	Use when
Synchronous	up to 20 MB	Yes	Small files where you want the row counts immediately.
Batched	any	Yes	Large CSV files; the SDK splits and uploads them for you.
Asynchronous job	any	No	Large files where you want to poll for completion.

Synchronous import#

python

bulk = ac.get_bulk_instance(ORG_ID, WORKSPACE_ID)

response = bulk.import_data(
    VIEW_ID,
    "TRUNCATEADD",      # import_type
    "csv",              # file_type
    True,               # auto_identify
    "/home/user/Sales.csv",
)
print(response)

Import types#

Value	Behaviour
`APPEND`	Add every row. Duplicates are kept.
`TRUNCATEADD`	Delete all existing rows, then add.
`UPDATEADD`	Update rows matching the configured key, add the rest.

Sample response#

json

{
  "status": "success",
  "summary": "Import data",
  "data": {
    "importSummary": {
      "totalColumnCount": 8,
      "selectedColumnCount": 8,
      "totalRowCount": 12500,
      "successRowCount": 12498,
      "warnings": 2,
      "importOperation": "updateadd"
    },
    "importErrors": "2 rows were skipped: date value did not match the column format."
  }
}

Batched import#

For files over 20 MB. The SDK splits the file, uploads each batch, and returns a job id.

python

bulk = ac.get_bulk_instance(ORG_ID, WORKSPACE_ID)

job_id = bulk.import_data_as_batches(
    VIEW_ID,
    "APPEND",
    True,
    "/home/user/large-sales.csv",
    100000,             # rows per batch
)

print(bulk.get_import_job_details(job_id))

Batch size is counted in rows. Keep each resulting batch under 100 MB.

Sample job response#

json

{
  "status": "success",
  "summary": "Get import job details",
  "data": {
    "jobId": "35130000001343088",
    "jobStatus": "JOB COMPLETED",
    "importSummary": {
      "totalRowCount": 2400000,
      "successRowCount": 2400000,
      "warnings": 0
    }
  }
}

jobStatus is one of JOB NOT STARTED, JOB IN PROGRESS, JOB COMPLETED or JOB FAILED.

Exporting data
Limits — file size ceilings and API unit costs.
Errors

Synchronous import#

Import types#

Sample response#

Batched import#

Sample job response#

Related#