Skip to main content

Configuration

Pipelines are declared in datacrafter.yml. Print the live schema (registered types included) with:

datacrafter config schema

Top-level keys

KeyRequiredNotes
versionyes"1"
project-nameyesHuman-readable name
project-idrecommendedUnique id from init
extractor or extractorsyes for run / checkSee extractors
processornoSee processors
destinationnoSee destinations

Environment interpolation: ${VAR} and ${VAR:-default} in string values.

Example

version: "1"
project-name: "my-project"
project-id: "unique-id"

extractor:
mode: "singlefile"
type: "file-csv"
method: "url"
force: true
config:
url: "https://example.com/data.csv"

processor:
config:
autoid: true
autotype: false
error_strategy: "skip"
max_retries: 3
keymap:
type: "names"
fields:
old_name: "new_name"
typemap:
field_name: "int"
custom:
type: "script"
code: "scripts/transform.py"

destination:
type: "file-jsonl"
fileprefix: "output"
compress: "gz"