Data Contract CLI
The datacontract CLI is an open-source command-line tool for working with data contracts.
It natively supports the Open Data Contract Standard to lint data contracts, connect to data sources and execute schema and quality tests, and export to different formats.
The tool is written in Python.
It can be used as a standalone CLI tool, in a CI/CD pipeline, or directly as a Python library.
Quick navigation: Documentation · Best Practices · Custom Export and Import · Development Setup
Getting started
Let’s look at this data contract: https://datacontract.com/orders-v1.odcs.yaml
We have a servers section with endpoint details to a Postgres database, schema for the structure and semantics of the data, service levels and quality attributes that describe the expected freshness and number of rows.
This data contract contains all information to connect to the database and check that the actual data meets the defined schema specification and quality expectations. We can use this information to test if the actual data product is compliant to the data contract.
Let’s use uv to install the CLI (or use the Docker image),
$ uv tool install --python python3.11 --upgrade 'datacontract-cli[all]'
Now, let’s run the tests:
$ export DATACONTRACT_POSTGRES_USERNAME=datacontract_cli.egzhawjonpfweuutedfy
$ export DATACONTRACT_POSTGRES_PASSWORD=jio10JuQfDfl9JCCPdaCCpuZ1YO
$ datacontract test https://datacontract.com/orders-v1.odcs.yaml
# returns:
Testing https://datacontract.com/orders-v1.odcs.yaml
Server: production (type=postgres, host=aws-1-eu-central-2.pooler.supabase.com, port=6543, database=postgres, schema=dp_orders_v1)
╭────────┬──────────────────────────────────────────────────────────┬─────────────────────────┬─────────╮
│ Result │ Check │ Field │ Details │
├────────┼──────────────────────────────────────────────────────────┼─────────────────────────┼─────────┤
│ passed │ Check that field 'line_item_id' is present │ line_items.line_item_id │ │
│ passed │ Check that field line_item_id has type UUID │ line_items.line_item_id │ │
│ passed │ Check that field line_item_id has no missing values │ line_items.line_item_id │ │
│ passed │ Check that field 'order_id' is present │ line_items.order_id │ │
│ passed │ Check that field order_id has type UUID │ line_items.order_id │ │
│ passed │ Check that field 'price' is present │ line_items.price │ │
│ passed │ Check that field price has type INTEGER │ line_items.price │ │
│ passed │ Check that field price has no missing values │ line_items.price │ │
│ passed │ Check that field 'sku' is present │ line_items.sku │ │
│ passed │ Check that field sku has type TEXT │ line_items.sku │ │
│ passed │ Check that field sku has no missing values │ line_items.sku │ │
│ passed │ Check that field 'customer_id' is present │ orders.customer_id │ │
│ passed │ Check that field customer_id has type TEXT │ orders.customer_id │ │
│ passed │ Check that field customer_id has no missing values │ orders.customer_id │ │
│ passed │ Check that field 'order_id' is present │ orders.order_id │ │
│ passed │ Check that field order_id has type UUID │ orders.order_id │ │
│ passed │ Check that field order_id has no missing values │ orders.order_id │ │
│ passed │ Check that unique field order_id has no duplicate values │ orders.order_id │ │
│ passed │ Check that field 'order_status' is present │ orders.order_status │ │
│ passed │ Check that field order_status has type TEXT │ orders.order_status │ │
│ passed │ Check that field 'order_timestamp' is present │ orders.order_timestamp │ │
│ passed │ Check that field order_timestamp has type TIMESTAMPTZ │ orders.order_timestamp │ │
│ passed │ Check that field 'order_total' is present │ orders.order_total │ │
│ passed │ Check that field order_total has type INTEGER │ orders.order_total │ │
│ passed │ Check that field order_total has no missing values │ orders.order_total │ │
╰────────┴──────────────────────────────────────────────────────────┴─────────────────────────┴─────────╯
🟢 data contract is valid. Run 25 checks. Took 3.938887 seconds.
Voilà, the CLI tested that the YAML itself is valid, all records comply with the schema, and all quality attributes are met.
We can also use the data contract metadata to export in many formats, e.g., to generate a SQL DDL:
$ datacontract export sql https://datacontract.com/orders-v1.odcs.yaml
# returns:
-- Data Contract: orders
-- SQL Dialect: postgres
CREATE TABLE orders (
order_id None not null primary key,
customer_id text not null,
order_total integer not null,
order_timestamp None,
order_status text
);
CREATE TABLE line_items (
line_item_id None not null primary key,
sku text not null,
price integer not null,
order_id None
);
Or generate an HTML export:
$ datacontract export html --output orders-v1.odcs.html https://datacontract.com/orders-v1.odcs.yaml
Usage
# create a new data contract from example and write it to odcs.yaml
$ datacontract init odcs.yaml
# lint the odcs.yaml and stop after the first validation error (default).
$ datacontract lint odcs.yaml
# show a changelog between two data contracts
$ datacontract changelog v1.odcs.yaml v2.odcs.yaml
# execute schema and quality checks (define credentials as environment variables)
$ datacontract test odcs.yaml
# export data contract as html (other formats: avro, dbt-models, dbt-sources, dbt-staging-sql, jsonschema, odcs, rdf, sql, sodacl, terraform, ...)
$ datacontract export html datacontract.yaml --output odcs.html
# import sql (other formats: avro, glue, bigquery, jsonschema, excel ...)
$ datacontract import sql --source my-ddl.sql --dialect postgres --output odcs.yaml
# import from Excel template
$ datacontract import excel --source odcs.xlsx --output odcs.yaml
# export to Excel template
$ datacontract export excel --output odcs.xlsx odcs.yaml
Programmatic (Python)
from datacontract.data_contract import DataContract
data_contract = DataContract(data_contract_file="odcs.yaml")
run = data_contract.test()
if not run.has_passed():
print("Data quality validation failed.")
# Abort pipeline, alert, or take corrective actions...
How to
- How to integrate Data Contract CLI in your CI/CD pipeline as a GitHub Action
- How to run the Data Contract CLI API to test data contracts with POST requests
- How to run Data Contract CLI in a Databricks pipeline
Installation
Choose the most appropriate installation method for your needs:
uv
The preferred way to install is uv:
uv tool install --python python3.11 --upgrade 'datacontract-cli[all]'
uvx
If you have uv installed, you can run datacontract-cli directly without installing:
uv run --with 'datacontract-cli[all]' datacontract --version
pip
Python 3.10, 3.11, and 3.12 are supported. We recommend using Python 3.11.
python3 -m pip install 'datacontract-cli[all]'
datacontract --version
pip with venv
Typically it is better to install the application in a virtual environment for your projects:
cd my-project
python3.11 -m venv venv
source venv/bin/activate
pip install 'datacontract-cli[all]'
datacontract --version
pipx
pipx installs into an isolated environment.
pipx install 'datacontract-cli[all]'
datacontract --version
Docker
You can also use our Docker image to run the CLI tool. It is also convenient for CI/CD pipelines.
docker pull datacontract/cli
docker run --rm -v ${PWD}:/home/datacontract datacontract/cli
You can create an alias for the Docker command to make it easier to use:
alias datacontract='docker run --rm -v "${PWD}:/home/datacontract" datacontract/cli:latest'
Note: The output of Docker command line messages is limited to 80 columns and may include line breaks. Don’t pipe docker output to files if you want to export code. Use the --output option instead.
Optional Dependencies (Extras)
The CLI tool defines several optional dependencies (also known as extras) that can be installed for using with specific servers types. With all, all server dependencies are included.
uv tool install --python python3.11 --upgrade 'datacontract-cli[all]'
A list of available extras:
| Dependency | Installation Command |
|---|---|
| Amazon Athena | pip install datacontract-cli[athena] |
| Avro Support | pip install datacontract-cli[avro] |
| Google BigQuery | pip install datacontract-cli[bigquery] |
| Databricks Integration | pip install datacontract-cli[databricks] |
| DuckDB (local/S3/GCS/Azure file testing) | pip install datacontract-cli[duckdb] |
| Iceberg | pip install datacontract-cli[iceberg] |
| Kafka Integration | pip install datacontract-cli[kafka] |
| MySQL Integration | pip install datacontract-cli[mysql] |
| PostgreSQL Integration | pip install datacontract-cli[postgres] |
| S3 Integration | pip install datacontract-cli[s3] |
| Snowflake Integration | pip install datacontract-cli[snowflake] |
| Microsoft SQL Server | pip install datacontract-cli[sqlserver] |
| Trino | pip install datacontract-cli[trino] |
| Impala | pip install datacontract-cli[impala] |
| dbt | pip install datacontract-cli[dbt] |
| DBML | pip install datacontract-cli[dbml] |
| Parquet | pip install datacontract-cli[parquet] |
| RDF | pip install datacontract-cli[rdf] |
| API (run as web server) | pip install datacontract-cli[api] |
| protobuf | pip install datacontract-cli[protobuf] |
Documentation
Commands
init
Usage: datacontract init [OPTIONS] [LOCATION]
Create an empty data contract.
╭─ Arguments ──────────────────────────────────────────────────────────────────────────────────────╮
│ location [LOCATION] The location of the data contract file to create. │
│ [default: datacontract.yaml] │
╰──────────────────────────────────────────────────────────────────────────────────────────────────╯
╭─ Options ────────────────────────────────────────────────────────────────────────────────────────╮
│ --template TEXT URL of a template or data contract │
│ --overwrite --no-overwrite Replace the existing datacontract.yaml │
│ [default: no-overwrite] │
│ --debug --no-debug Enable debug logging │
│ --help Show this message and exit. │
╰──────────────────────────────────────────────────────────────────────────────────────────────────╯
Example: datacontract init datacontract.yaml
lint
Usage: datacontract lint [OPTIONS] [LOCATION]
Validate that the datacontract.yaml is correctly formatted.
╭─ Arguments ──────────────────────────────────────────────────────────────────────────────────────╮
│ location [LOCATION] The location (url or path) of the data contract yaml. │
│ [default: datacontract.yaml] │
╰──────────────────────────────────────────────────────────────────────────────────────────────────╯
╭─ Options ────────────────────────────────────────────────────────────────────────────────────────╮
│ --json-schema TEXT The location (url or path) of the ODCS JSON │
│ Schema │
│ --output PATH Specify the file path where the test results │
│ should be written to (e.g., │
│ './test-results/TEST-datacontract.xml'). If no │
│ path is provided, the output will be printed to │
│ stdout. │
│ --output-format [json|junit] The target format for the test results. │
│ --all-errors Report all JSON Schema validation errors instead │
│ of stopping after the first one. │
│ --debug --no-debug Enable debug logging │
│ --help Show this message and exit. │
╰──────────────────────────────────────────────────────────────────────────────────────────────────╯
Example: datacontract lint datacontract.yaml
changelog
Usage: datacontract changelog [OPTIONS] V1 V2
Show a changelog between two data contracts.
╭─ Arguments ──────────────────────────────────────────────────────────────────────────────────────╮
│ * v1 TEXT The location (path) of the source (before) data contract YAML. [required] │
│ * v2 TEXT The location (path) of the target (after) data contract YAML. [required] │
╰──────────────────────────────────────────────────────────────────────────────────────────────────╯
╭─ Options ────────────────────────────────────────────────────────────────────────────────────────╮
│ --debug --no-debug Enable debug logging │
│ --help Show this message and exit. │
╰──────────────────────────────────────────────────────────────────────────────────────────────────╯
$ datacontract changelog v1.odcs.yaml v2.odcs.yaml
test
Usage: datacontract test [OPTIONS] [LOCATION]
Run schema and quality tests on configured servers.
╭─ Arguments ──────────────────────────────────────────────────────────────────────────────────────╮
│ location [LOCATION] The location (url or path) of the data contract yaml. │
│ [default: datacontract.yaml] │
╰──────────────────────────────────────────────────────────────────────────────────────────────────╯
╭─ Options ────────────────────────────────────────────────────────────────────────────────────────╮
│ --json-schema TEXT The location (url or path) │
│ of the ODCS JSON Schema │
│ --server TEXT The server configuration to │
│ run the schema and quality │
│ tests. Use the key of the │
│ server object in the data │
│ contract yaml file to refer │
│ to a server, e.g., │
│ `production`, or `all` for │
│ all servers (default). │
│ [default: all] │
│ --schema-name TEXT Which schema to test, e.g., │
│ `orders`, or `all` for all │
│ schemas (default). │
│ [default: all] │
│ --publish-test-results --no-publish-test-results Deprecated. Use publish │
│ parameter. Publish the │
│ results after the test │
│ [default: │
│ no-publish-test-results] │
│ --publish TEXT The url to publish the │
│ results after the test. │
│ --output PATH Specify the file path where │
│ the test results should be │
│ written to (e.g., │
│ './test-results/TEST-datac… │
│ --output-format [json|junit] The target format for the │
│ test results. │
│ --checks TEXT Comma-separated list of │
│ check categories to run │
│ (available: schema, │
│ quality, servicelevel, │
│ custom). Omit to enable │
│ all. │
│ --logs --no-logs Print logs │
│ [default: no-logs] │
│ --ssl-verification --no-ssl-verification SSL verification when │
│ publishing the data │
│ contract. │
│ [default: ssl-verification] │
│ --debug --no-debug Enable debug logging │
│ --help Show this message and exit. │
╰──────────────────────────────────────────────────────────────────────────────────────────────────╯
Example: datacontract test datacontract.yaml --server production
Data Contract CLI connects to a data source and runs schema and quality tests to verify that the data contract is valid.
$ datacontract test --server production datacontract.yaml
For CI/CD pipelines, see ci.
The application uses different engines, based on the server type.
Internally, it connects with DuckDB, Spark, or a native connection and executes the most tests with soda-core and fastjsonschema.
Supported Data Sources
The server block in the datacontract.yaml is used to set up the connection.
In addition, credentials, such as username and passwords, are provided with environment variables.
Feel free to create an issue, if you need support for additional types and formats.
S3
Data Contract CLI can test data that is stored in S3 buckets or any S3-compliant endpoints in various formats.
- CSV
- JSON
- Delta
- Parquet
- Iceberg (coming soon)
Examples
JSON
datacontract.yaml
servers:
production:
type: s3
endpointUrl: https://minio.example.com # not needed with AWS S3
location: s3://bucket-name/path/*/*.json
format: json
delimiter: new_line # new_line, array, or none
Delta Tables
datacontract.yaml
servers:
production:
type: s3
endpointUrl: https://minio.example.com # not needed with AWS S3
location: s3://bucket-name/path/table.delta # path to the Delta table folder containing parquet data files and the _delta_log
format: delta
Environment Variables
| Environment Variable | Example | Description |
|---|---|---|
DATACONTRACT_S3_REGION |
eu-central-1 |
Region of S3 bucket |
DATACONTRACT_S3_ACCESS_KEY_ID |
AKIAXV5Q5QABCDEFGH |
AWS Access Key ID |
DATACONTRACT_S3_SECRET_ACCESS_KEY |
93S7LRrJcqLaaaa/XXXXXXXXXXXXX |
AWS Secret Access Key |
DATACONTRACT_S3_SESSION_TOKEN |
AQoDYXdzEJr... |
AWS temporary session token (optional) |
Athena
Data Contract CLI can test data in AWS Athena stored in S3. Supports different file formats, such as Iceberg, Parquet, JSON, CSV…
Example
datacontract.yaml
servers:
athena:
type: athena
catalog: awsdatacatalog # awsdatacatalog is the default setting
schema: icebergdemodb # in Athena, this is called "database"
regionName: eu-central-1
stagingDir: s3://my-bucket/athena-results/
models:
my_table: # corresponds to a table or view name
type: table
fields:
my_column_1: # corresponds to a column
type: string
config:
physicalType: varchar
Environment Variables
| Environment Variable | Example | Description |
|---|---|---|
DATACONTRACT_S3_REGION |
eu-central-1 |
Region of Athena service |
DATACONTRACT_S3_ACCESS_KEY_ID |
AKIAXV5Q5QABCDEFGH |
AWS Access Key ID |
DATACONTRACT_S3_SECRET_ACCESS_KEY |
93S7LRrJcqLaaaa/XXXXXXXXXXXXX |
AWS Secret Access Key |
DATACONTRACT_S3_SESSION_TOKEN |
AQoDYXdzEJr... |
AWS temporary session token (optional) |
Google Cloud Storage (GCS)
The S3 integration also works with files on Google Cloud Storage through its interoperability.
Use https://storage.googleapis.com as the endpoint URL.
Example
datacontract.yaml
servers:
production:
type: s3
endpointUrl: https://storage.googleapis.com
location: s3://bucket-name/path/*/*.json # use s3:// schema instead of gs://
format: json
delimiter: new_line # new_line, array, or none
Environment Variables
| Environment Variable | Example | Description |
|---|---|---|
DATACONTRACT_S3_ACCESS_KEY_ID |
GOOG1EZZZ... |
The GCS HMAC Key Key ID |
DATACONTRACT_S3_SECRET_ACCESS_KEY |
PDWWpb... |
The GCS HMAC Key Secret |
BigQuery
We support authentication to BigQuery using Service Account Key or Application Default Credentials (ADC). ADC supports Workload Identity Federation (WIF), GCE metadata server, and gcloud auth application-default login. The used Service Account should include the roles:
- BigQuery Job User
- BigQuery Data Viewer
When no DATACONTRACT_BIGQUERY_ACCOUNT_INFO_JSON_PATH is set, the CLI falls back to ADC/WIF automatically via Soda’s use_context_auth.
Example
datacontract.yaml
servers:
production:
type: bigquery
project: datameshexample-product
dataset: datacontract_cli_test_dataset
models:
datacontract_cli_test_table: # corresponds to a BigQuery table
type: table
fields: ...
Environment Variables
| Environment Variable | Example | Description |
|---|---|---|
DATACONTRACT_BIGQUERY_ACCOUNT_INFO_JSON_PATH |
~/service-access-key.json |
Service Account key JSON file. If not set, ADC/WIF is used automatically. |
DATACONTRACT_BIGQUERY_IMPERSONATION_ACCOUNT |
[email protected] |
Optional. Service account to impersonate. Works with both key file and ADC auth. |
Azure
Data Contract CLI can test data that is stored in Azure Blob storage or Azure Data Lake Storage (Gen2) (ADLS) in various formats.
Example
datacontract.yaml
servers:
production:
type: azure
location: abfss://datameshdatabricksdemo.dfs.core.windows.net/inventory_events/*.parquet
format: parquet
Environment Variables
Authentication works with an Azure Service Principal (SPN) aka App Registration with a secret.
| Environment Variable | Example | Description |
|---|---|---|
DATACONTRACT_AZURE_TENANT_ID |
79f5b80f-10ff-40b9-9d1f-774b42d605fc |
The Azure Tenant ID |
DATACONTRACT_AZURE_CLIENT_ID |
3cf7ce49-e2e9-4cbc-a922-4328d4a58622 |
The ApplicationID / ClientID of the app registration |
DATACONTRACT_AZURE_CLIENT_SECRET |
yZK8Q~GWO1MMXXXXXXXXXXXXX |
The Client Secret value |
SQL Server
Data Contract CLI can test data in MS SQL Server (including Azure SQL, Synapse Analytics SQL Pool, and Microsoft Fabric).
Example
datacontract.yaml
servers:
production:
type: sqlserver
host: localhost
port: 5432
database: tempdb
schema: dbo
driver: ODBC Driver 18 for SQL Server
models:
my_table_1: # corresponds to a table
type: table
fields:
my_column_1: # corresponds to a column
type: varchar
Environment Variables
| Environment Variable | Example | Description |
|---|---|---|
DATACONTRACT_SQLSERVER_AUTHENTICATION |
sql |
Supported: sql (default), cli (uses az login session), windows, ActiveDirectoryPassword, ActiveDirectoryServicePrincipal, ActiveDirectoryInteractive |
DATACONTRACT_SQLSERVER_USERNAME |
root |
Username (for sql, ActiveDirectoryPassword, ActiveDirectoryInteractive) |
DATACONTRACT_SQLSERVER_PASSWORD |
toor |
Password (for sql and ActiveDirectoryPassword) |
DATACONTRACT_SQLSERVER_CLIENT_ID |
a3cf5d29-b1a7-... |
Application/Client ID (for ActiveDirectoryServicePrincipal) |
DATACONTRACT_SQLSERVER_CLIENT_SECRET |
kX9~Qr2Lm.Tz4W... |
Client secret (for ActiveDirectoryServicePrincipal) |
DATACONTRACT_SQLSERVER_TRUST_SERVER_CERTIFICATE |
True |
Trust self-signed certificate |
DATACONTRACT_SQLSERVER_ENCRYPTED_CONNECTION |
True |
Use SSL |
DATACONTRACT_SQLSERVER_DRIVER |
ODBC Driver 18 for SQL Server |
ODBC driver name |
DATACONTRACT_SQLSERVER_TRUSTED_CONNECTION |
True |
Deprecated. Equivalent to AUTHENTICATION=windows |
Oracle
Data Contract CLI can test data in Oracle Database.
Example
datacontract.yaml
servers:
oracle:
type: oracle
host: localhost
port: 1521
service_name: ORCL
schema: ADMIN
models:
my_table_1: # corresponds to a table
type: table
fields:
my_column_1: # corresponds to a column
type: decimal
description: Decimal number
my_column_2: # corresponds to another column
type: text
description: Unicode text string
config:
oracleType: NVARCHAR2 # optional: can be used to explicitly define the type used in the database
# if not set a default mapping will be used
Environment Variables
These environment variable specify the credentials used by the datacontract tool to connect to the database.
If you’ve started the database from a container, e.g. oracle-free
this should match either system and what you specified as ORACLE_PASSWORD on the container or
alternatively what you’ve specified under APP_USER and APP_USER_PASSWORD.
If you require thick mode to connect to the database, you need to have an Oracle Instant Client
installed on the system and specify the path to the installation within the environment variable
DATACONTRACT_ORACLE_CLIENT_DIR.
| Environment Variable | Example | Description |
|---|---|---|
DATACONTRACT_ORACLE_USERNAME |
system |
Username |
DATACONTRACT_ORACLE_PASSWORD |
0x162e53 |
Password |
DATACONTRACT_ORACLE_CLIENT_DIR |
C:\oracle\client |
Path to Oracle Instant Client installation |
Databricks
Works with Unity Catalog and Hive metastore.
Needs a running SQL warehouse or compute cluster.
Example
datacontract.yaml
servers:
production:
type: databricks
catalog: acme_catalog_prod
schema: orders_latest
models:
orders: # corresponds to a table
type: table
fields: ...
Environment Variables
| Environment Variable | Example | Description |
|---|---|---|
DATACONTRACT_DATABRICKS_TOKEN |
dapia00000000000000000000000000000 |
The personal access token to authenticate |
DATACONTRACT_DATABRICKS_HTTP_PATH |
/sql/1.0/warehouses/b053a3ffffffff |
The HTTP path to the SQL warehouse or compute cluster |
DATACONTRACT_DATABRICKS_SERVER_HOSTNAME |
dbc-abcdefgh-1234.cloud.databricks.com |
The host name of the SQL warehouse or compute cluster |
Databricks (programmatic)
Works with Unity Catalog and Hive metastore.
When running in a notebook or pipeline, the provided spark session can be used.
An additional authentication is not required.
Requires a Databricks Runtime with Python >= 3.10.
Example
datacontract.yaml
servers:
production:
type: databricks
host: dbc-abcdefgh-1234.cloud.databricks.com # ignored, always use current host
catalog: acme_catalog_prod
schema: orders_latest
models:
orders: # corresponds to a table
type: table
fields: ...
Installing on Databricks Compute
Important: When using Databricks LTS ML runtimes (15.4, 16.4), installing via %pip install in notebooks can cause issues.
Recommended approach: Use Databricks’ native library management instead:
- Create or configure your compute cluster:
- Navigate to Compute in the Databricks workspace
- Create a new cluster or select an existing one
- Go to the Libraries tab
- Add the datacontract-cli library:
- Click Install new
- Select PyPI as the library source
- Enter package name:
datacontract-cli[databricks] - Click Install
-
Restart the cluster to apply the library installation
- Use in your notebook without additional installation:
from datacontract.data_contract import DataContract data_contract = DataContract( data_contract_file="/Volumes/acme_catalog_prod/orders_latest/datacontract/datacontract.yaml", spark=spark) run = data_contract.test() run.result
Databricks’ library management properly resolves dependencies during cluster initialization, rather than at runtime in the notebook.
Dataframe (programmatic)
Works with Spark DataFrames. DataFrames need to be created as named temporary views. Multiple temporary views are supported if your data contract contains multiple models.
Testing DataFrames is useful to test your datasets in a pipeline before writing them to a data source.
Example
datacontract.yaml
servers:
production:
type: dataframe
models:
my_table: # corresponds to a temporary view
type: table
fields: ...
Example code
from datacontract.data_contract import DataContract
df.createOrReplaceTempView("my_table")
data_contract = DataContract(
data_contract_file="datacontract.yaml",
spark=spark,
)
run = data_contract.test()
assert run.result == "passed"
Snowflake
Data Contract CLI can test data in Snowflake.
Example
datacontract.yaml
servers:
snowflake:
type: snowflake
account: abcdefg-xn12345
database: ORDER_DB
schema: ORDERS_PII_V2
models:
my_table_1: # corresponds to a table
type: table
fields:
my_column_1: # corresponds to a column
type: varchar
Environment Variables
All parameters supported by Soda, uppercased and prepended by DATACONTRACT_SNOWFLAKE_ prefix.
For example:
| Soda parameter | Environment Variable |
|---|---|
username |
DATACONTRACT_SNOWFLAKE_USERNAME |
password |
DATACONTRACT_SNOWFLAKE_PASSWORD |
warehouse |
DATACONTRACT_SNOWFLAKE_WAREHOUSE |
role |
DATACONTRACT_SNOWFLAKE_ROLE |
connection_timeout |
DATACONTRACT_SNOWFLAKE_CONNECTION_TIMEOUT |
Beware, that parameters:
accountdatabaseschema
are obtained from the servers section of the YAML-file.
E.g. from the example above:
servers:
snowflake:
account: abcdefg-xn12345
database: ORDER_DB
schema: ORDERS_PII_V2
Kafka
Kafka support is currently considered experimental.
Example
datacontract.yaml
servers:
production:
type: kafka
host: abc-12345.eu-central-1.aws.confluent.cloud:9092
topic: my-topic-name
format: json
Environment Variables
| Environment Variable | Example | Description |
|---|---|---|
DATACONTRACT_KAFKA_SASL_USERNAME |
xxx |
The SASL username (key). |
DATACONTRACT_KAFKA_SASL_PASSWORD |
xxx |
The SASL password (secret). |
DATACONTRACT_KAFKA_SASL_MECHANISM |
PLAIN |
Default PLAIN. Other supported mechanisms: SCRAM-SHA-256 and SCRAM-SHA-512 |
Postgres
Data Contract CLI can test data in Postgres or Postgres-compliant databases (e.g., RisingWave).
Example
datacontract.yaml
servers:
postgres:
type: postgres
host: localhost
port: 5432
database: postgres
schema: public
models:
my_table_1: # corresponds to a table
type: table
fields:
my_column_1: # corresponds to a column
type: varchar
Environment Variables
| Environment Variable | Example | Description |
|---|---|---|
DATACONTRACT_POSTGRES_USERNAME |
postgres |
Username |
DATACONTRACT_POSTGRES_PASSWORD |
mysecretpassword |
Password |
MySQL
Data Contract CLI can test data in MySQL or MySQL-compliant databases (e.g., MariaDB).
Example
datacontract.yaml
servers:
mysql:
type: mysql
host: localhost
port: 3306
database: mydb
models:
my_table_1: # corresponds to a table
type: table
fields:
my_column_1: # corresponds to a column
type: varchar
Environment Variables
| Environment Variable | Example | Description |
|---|---|---|
DATACONTRACT_MYSQL_USERNAME |
root |
Username |
DATACONTRACT_MYSQL_PASSWORD |
mysecretpassword |
Password |
Trino
Data Contract CLI can test data in Trino.
Example
datacontract.yaml
servers:
trino:
type: trino
host: localhost
port: 8080
catalog: my_catalog
schema: my_schema
models:
my_table_1: # corresponds to a table
type: table
fields:
my_column_1: # corresponds to a column
type: varchar
my_column_2: # corresponds to a column with custom trino type
type: object
config:
trinoType: row(en_us varchar, pt_br varchar)
Environment Variables
| Environment Variable | Example | Description |
|---|---|---|
DATACONTRACT_TRINO_USERNAME |
trino |
Username |
DATACONTRACT_TRINO_PASSWORD |
mysecretpassword |
Password |
Impala
Data Contract CLI can run Soda checks against an Apache Impala cluster.
Example
datacontract.yaml
servers:
impala:
type: impala
host: my-impala-host
port: 443
# Optional default database used for Soda scans
database: my_database
models:
my_table_1: # corresponds to a table
type: table
# fields as usual …
Environment Variables
| Environment Variable | Example | Description |
|---|---|---|
DATACONTRACT_IMPALA_USERNAME |
analytics_user |
Username used to connect to Impala |
DATACONTRACT_IMPALA_PASSWORD |
mysecretpassword |
Password for the Impala user |
DATACONTRACT_IMPALA_USE_SSL |
true |
Whether to use SSL; defaults to true if unset |
DATACONTRACT_IMPALA_AUTH_MECHANISM |
LDAP |
Authentication mechanism; defaults to LDAP |
DATACONTRACT_IMPALA_USE_HTTP_TRANSPORT |
true |
Whether to use the HTTP transport; defaults to true |
DATACONTRACT_IMPALA_HTTP_PATH |
cliservice |
HTTP path for the Impala service; defaults to cliservice |
Type-mapping note (logicalType → Impala type)
If physicalType is not specified in the schema, we recommend the following mapping from logicalType to Impala column types:
| logicalType | Recommended Impala type |
|---|---|
integer |
INT or BIGINT |
number |
DOUBLE/decimal(..) |
string |
STRING or VARCHAR |
boolean |
BOOLEAN |
date |
DATE |
datetime |
TIMESTAMP |
This keeps the Impala schema compatible with the expectations of the Soda checks generated by datacontract-cli.
API
Data Contract CLI can test APIs that return data in JSON format. Currently, only GET requests are supported.
Example
datacontract.yaml
servers:
api:
type: "api"
location: "https://api.example.com/path"
delimiter: none # new_line, array, or none (default)
models:
my_object: # corresponds to the root element of the JSON response
type: object
fields:
field1:
type: string
fields2:
type: number
Environment Variables
| Environment Variable | Example | Description |
|---|---|---|
DATACONTRACT_API_HEADER_AUTHORIZATION |
Bearer <token> |
The value for the authorization header. Optional. |
Local
Data Contract CLI can test local files in parquet, json, csv, or delta format.
Example
datacontract.yaml
servers:
local:
type: local
path: ./*.parquet
format: parquet
models:
my_table_1: # corresponds to a table
type: table
fields:
my_column_1: # corresponds to a column
type: varchar
my_column_2: # corresponds to a column
type: string
ci
Usage: datacontract ci [OPTIONS] [LOCATIONS]...
Run tests for CI/CD pipelines. Emits GitHub Actions annotations and step summary.
╭─ Arguments ──────────────────────────────────────────────────────────────────────────────────────╮
│ locations [LOCATIONS]... The location(s) (url or path) of the data contract yaml │
│ file(s). │
╰──────────────────────────────────────────────────────────────────────────────────────────────────╯
╭─ Options ────────────────────────────────────────────────────────────────────────────────────────╮
│ --json-schema TEXT The location (url or path) │
│ of the ODCS JSON Schema │
│ --server TEXT The server configuration │
│ to run the schema and │
│ quality tests. Use the key │
│ of the server object in │
│ the data contract yaml │
│ file to refer to a server, │
│ e.g., `production`, or │
│ `all` for all servers │
│ (default). │
│ [default: all] │
│ --publish TEXT The url to publish the │
│ results after the test. │
│ --output PATH Specify the file path │
│ where the test results │
│ should be written to │
│ (e.g., │
│ './test-results/TEST-data… │
│ --output-format [json|junit] The target format for the │
│ test results. │
│ --logs --no-logs Print logs │
│ [default: no-logs] │
│ --json Print test results as JSON │
│ to stdout. │
│ --fail-on [warning|error|never] Minimum severity that │
│ causes a non-zero exit │
│ code. │
│ [default: error] │
│ --ssl-verification --no-ssl-verification SSL verification when │
│ publishing the data │
│ contract. │
│ [default: │
│ ssl-verification] │
│ --debug --no-debug Enable debug logging │
│ --help Show this message and │
│ exit. │
╰──────────────────────────────────────────────────────────────────────────────────────────────────╯
The ci command wraps test with CI/CD-specific features:
- Multiple contracts:
datacontract ci contracts/*.yaml - CI annotations: Inline annotations for failed checks (GitHub Actions and Azure DevOps)
- Markdown summary of the test results (GitHub Actions)
--json: Print test results as JSON to stdout for machine-readable output--fail-on: Control the minimum severity that causes a non-zero exit code. Default iserror; set towarningto also fail on warnings, orneverto always exit 0.
The supported server types and their configuration are equivalent to the test command.
# Single contract
$ datacontract ci datacontract.yaml
# Multiple contracts
$ datacontract ci contracts/*.yaml
# Fail on warnings too
$ datacontract ci --fail-on warning datacontract.yaml
# JSON output for scripting
$ datacontract ci --json datacontract.yaml
GitHub Actions workflow example
# .github/workflows/datacontract.yml
name: Data Contract CI
on:
push:
branches: [main]
pull_request:
jobs:
datacontract-ci:
runs-on: ubuntu-latest
steps:
- uses: actions/checkout@v4
- uses: actions/setup-python@v5
with:
python-version: "3.11"
- run: pip install datacontract-cli
# Test one or more data contracts (supports globs, e.g. contracts/*.yaml)
- run: datacontract ci datacontract.yaml
Azure DevOps pipeline example
# azure-pipelines.yml
trigger:
branches:
include:
- main
pool:
vmImage: "ubuntu-latest"
steps:
- task: UsePythonVersion@0
inputs:
versionSpec: "3.11"
- script: pip install datacontract-cli
displayName: "Install datacontract-cli"
# Test one or more data contracts (supports globs, e.g. contracts/*.yaml)
- script: datacontract ci datacontract.yaml
displayName: "Run data contract tests"
export
Usage: datacontract export [OPTIONS] COMMAND [ARGS]...
Convert a data contract to a target format.
╭─ Options ────────────────────────────────────────────────────────────────────────────────────────╮
│ --help Show this message and exit. │
╰──────────────────────────────────────────────────────────────────────────────────────────────────╯
╭─ Commands ───────────────────────────────────────────────────────────────────────────────────────╮
│ sql Export a data contract to SQL DDL. │
│ sql-query Export a data contract to a SQL query. │
│ dbt-models Export a data contract to dbt model schema YAML. │
│ dbt-sources Export a data contract to dbt sources YAML. │
│ dbt-staging-sql Export a data contract to a dbt staging SQL file. │
│ avro Export a data contract to Avro schema. │
│ avro-idl Export a data contract to Avro IDL. │
│ jsonschema Export a data contract to JSON Schema. │
│ pydantic-model Export a data contract to a Pydantic model. │
│ protobuf Export a data contract to Protobuf schema. │
│ odcs Export a data contract to ODCS format. │
│ rdf Export a data contract to RDF. │
│ html Export a data contract to HTML. │
│ markdown Export a data contract to Markdown. │
│ mermaid Export a data contract to Mermaid diagram. │
│ bigquery Export a data contract to BigQuery schema. │
│ dbml Export a data contract to DBML. │
│ go Export a data contract to Go structs. │
│ spark Export a data contract to Spark schema. │
│ sqlalchemy Export a data contract to SQLAlchemy models. │
│ iceberg Export a data contract to Iceberg schema. │
│ sodacl Export a data contract to SodaCL checks. │
│ great-expectations Export a data contract to Great Expectations suite. │
│ data-caterer Export a data contract to Data Caterer format. │
│ dcs Export a data contract to DCS format. │
│ dqx Export a data contract to DQX format. │
│ excel Export a data contract to Excel. │
│ custom Export a data contract using a custom Jinja template. │
╰──────────────────────────────────────────────────────────────────────────────────────────────────╯
Example: datacontract export html datacontract.yaml --output datacontract.html
For SQL dialects (postgres, mysql, snowflake, databricks, sqlserver, trino, oracle), use
`datacontract export sql --dialect <dialect>`.
Run datacontract export <format> --help to see the format-specific options (e.g. datacontract export sql --help). If you are missing a format, please create an issue on GitHub.
Examples
# Example export data contract as HTML
datacontract export html --output datacontract.html
SQL
The export function converts a given data contract into a SQL data definition language (DDL).
datacontract export sql datacontract.yaml --output output.sql
The SQL dialect is determined from the servers block in the data contract (e.g. type: postgres, type: snowflake). Alternatively, pass it explicitly:
datacontract export sql datacontract.yaml --dialect postgres --output output.sql
If using Databricks and an error is thrown when deploying SQL DDLs with variant columns, set the following property.
spark.conf.set(“spark.databricks.delta.schema.typeCheck.enabled”, “false”)
Great Expectations
The export function transforms a specified data contract into a comprehensive Great Expectations JSON suite.
If the contract includes multiple models, you need to specify the names of the schema/models you wish to export.
datacontract export great-expectations datacontract.yaml --schema-name orders
The export creates a list of expectations by utilizing:
- The data from the Model definition with a fixed mapping
- The expectations provided in the quality field for each model (find here the expectations gallery: Great Expectations Gallery)
Additional Arguments
To further customize the export, the following optional arguments are available:
-
suite_name: The name of the expectation suite. This suite groups all generated expectations and provides a convenient identifier within Great Expectations. If not provided, a default suite name will be generated based on the model name(s). engine: Specifies the engine used to run Great Expectations checks. Accepted values are:pandas— Use this when working with in-memory data frames through the Pandas library.spark— Use this for working with Spark dataframes.sql— Use this for working with SQL databases.
-
sql_server_type: Specifies the type of SQL server to connect with whenengineis set tosql.Providing
sql_server_typeensures that the appropriate SQL dialect and connection settings are applied during the expectation validation.
RDF
The export function converts a given data contract into a RDF representation. You have the option to
add a base_url which will be used as the default prefix to resolve relative IRIs inside the document.
datacontract export rdf --base https://www.example.com/ datacontract.yaml
The data contract is mapped onto the following concepts of a yet to be defined Data Contract Ontology named https://datacontract.com/DataContractSpecification/ :
- DataContract
- Server
- Model
Having the data contract inside an RDF Graph gives us access the following use cases:
- Interoperability with other data contract specification formats
- Store data contracts inside a knowledge graph
- Enhance a semantic search to find and retrieve data contracts
- Linking model elements to already established ontologies and knowledge
- Using full power of OWL to reason about the graph structure of data contracts
- Apply graph algorithms on multiple data contracts (Find similar data contracts, find “gatekeeper” data products, find the true domain owner of a field attribute)
DBML
The export function converts the logical data types of the datacontract into the specific ones of a concrete Database
if a server is selected via the --server option (based on the type of that server). If no server is selected, the
logical data types are exported.
DBT & DBT-SOURCES
The export function converts the datacontract to dbt models in YAML format, with support for SQL dialects.
If a server is selected via the --server option (based on the type of that server) then the DBT column data_types match the expected data types of the server.
If no server is selected, then it defaults to snowflake.
Spark
The export function converts the data contract specification into a StructType Spark schema. The returned value is a Python code picture of the model schemas. Spark DataFrame schema is defined as StructType. For more details about Spark Data Types please see the spark documentation
Avro
The export function converts the data contract specification into an avro schema. It supports specifying custom avro properties for logicalTypes and default values.
Custom Avro Properties
We support a config map on field level. A config map may include any additional key-value pairs and support multiple server type bindings.
To specify custom Avro properties in your data contract, you can define them within the config section of your field definition. Below is an example of how to structure your YAML configuration to include custom Avro properties, such as avroLogicalType and avroDefault.
NOTE: At this moment, we just support logicalType and default
Example Configuration
models:
orders:
fields:
my_field_1:
description: Example for AVRO with Timestamp (microsecond precision) https://avro.apache.org/docs/current/spec.html#Local+timestamp+%28microsecond+precision%29
type: long
example: 1672534861000000 # Equivalent to 2023-01-01 01:01:01 in microseconds
required: true
config:
avroLogicalType: local-timestamp-micros
avroDefault: 1672534861000000
Explanation
- models: The top-level key that contains different models (tables or objects) in your data contract.
- orders: A specific model name. Replace this with the name of your model.
- fields: The fields within the model. Each field can have various properties defined.
- my_field_1: The name of a specific field. Replace this with your field name.
- description: A textual description of the field.
- type: The data type of the field. In this example, it is
long. - example: An example value for the field.
- required: Is this a required field (as opposed to optional/nullable).
- config: Section to specify custom Avro properties.
- avroLogicalType: Specifies the logical type of the field in Avro. In this example, it is
local-timestamp-micros. - avroDefault: Specifies the default value for the field in Avro. In this example, it is 1672534861000000 which corresponds to ` 2023-01-01 01:01:01 UTC`.
- avroLogicalType: Specifies the logical type of the field in Avro. In this example, it is
Data Caterer
The export function converts the data contract to a data generation task in YAML format that can be ingested by Data Caterer. This gives you the ability to generate production-like data in any environment based off your data contract.
datacontract export data-caterer datacontract.yaml --schema-name orders
You can further customise the way data is generated via adding additional metadata in the YAML to suit your needs.
Iceberg
Exports to an Iceberg Table Json Schema Definition.
This export only supports a single model export at a time because Iceberg’s schema definition is for a single table and the exporter maps 1 model to 1 table, use the --schema-name flag
to limit your contract export to a single model.
$ datacontract export iceberg --schema-name orders https://datacontract.com/examples/orders-latest/datacontract.yaml --output /tmp/orders_iceberg.json
$ cat /tmp/orders_iceberg.json | jq '.'
{
"type": "struct",
"fields": [
{
"id": 1,
"name": "order_id",
"type": "string",
"required": true
},
{
"id": 2,
"name": "order_timestamp",
"type": "timestamptz",
"required": true
},
{
"id": 3,
"name": "order_total",
"type": "long",
"required": true
},
{
"id": 4,
"name": "customer_id",
"type": "string",
"required": false
},
{
"id": 5,
"name": "customer_email_address",
"type": "string",
"required": true
},
{
"id": 6,
"name": "processed_timestamp",
"type": "timestamptz",
"required": true
}
],
"schema-id": 0,
"identifier-field-ids": [
1
]
}
Custom
The export function converts the data contract specification into the custom format with Jinja. You can specify the path to a Jinja template with the --template argument, allowing you to output files in any format.
datacontract export custom --template template.txt datacontract.yaml
Jinja templates & variables
You can directly use the ODCS (Open Data Contract Standard) object as a template variable. data_contract is an OpenDataContractStandard instance; top-level fields include name, id, version, schema_ (the list of schemas), servers, team, etc.
$ cat template.txt
title: {{ data_contract.name }}
schemas:
{%- for schema in data_contract.schema_ %}
- name: {{ schema.name }}
{%- endfor %}
$ datacontract export custom --template template.txt datacontract.yaml
title: Orders Latest
schemas:
- name: orders
Example Jinja Templates for a customized dbt model
You can export a given dbt model containing any logic by adding the schema-name filter/parameter (in ODCS, “schemas” are the equivalent of “models” in dbt).
It adds jinja variable passed to your template.file:
schema_name: strschema: SchemaObject from ODCS
Below is an example of a dbt staging layer that converts a field of type: timestamp to a DATETIME type with time zone conversion.
-
template.sqlSELECT {%- for field in schema.properties %} {%- if field.physicalType == "timestamp" %} DATETIME({{ field.name }}, "Asia/Tokyo") AS {{ field.name }}, {%- else %} {{ field.name }} AS {{ field.name }}, {%- endif %} {%- endfor %} FROM {{ "{{" }} ref('{{ schema_name }}') {{ "}}" }} - export command
datacontract export custom datacontract.odcs.yaml --template template.sql --schema-name orders -
output.sqlSELECT order_id AS order_id, DATETIME(order_timestamp, "Asia/Tokyo") AS order_timestamp, order_total AS order_total, customer_id AS customer_id, customer_email_address AS customer_email_address, DATETIME(processed_timestamp, "Asia/Tokyo") AS processed_timestamp, FROM {{ ref('orders') }}
ODCS Excel Template
The export function converts a data contract into an ODCS (Open Data Contract Standard) Excel template. This creates a user-friendly Excel spreadsheet that can be used for authoring, sharing, and managing data contracts using the familiar Excel interface.
datacontract export excel --output datacontract.xlsx datacontract.yaml
The Excel format enables:
- User-friendly authoring: Create and edit data contracts in Excel’s familiar interface
- Easy sharing: Distribute data contracts as standard Excel files
- Collaboration: Enable non-technical stakeholders to contribute to data contract definitions
- Round-trip conversion: Import Excel templates back to YAML data contracts
For more information about the Excel template structure, visit the ODCS Excel Template repository.
import
Usage: datacontract import [OPTIONS] COMMAND [ARGS]...
Create a data contract from a source format.
╭─ Options ────────────────────────────────────────────────────────────────────────────────────────╮
│ --help Show this message and exit. │
╰──────────────────────────────────────────────────────────────────────────────────────────────────╯
╭─ Commands ───────────────────────────────────────────────────────────────────────────────────────╮
│ sql Import a data contract from a SQL DDL file. │
│ avro Import a data contract from an Avro schema file. │
│ dbt Import a data contract from a dbt manifest file. │
│ dbml Import a data contract from a DBML file. │
│ glue Import a data contract from AWS Glue. │
│ bigquery Import a data contract from BigQuery. │
│ unity Import a data contract from Databricks Unity Catalog. │
│ jsonschema Import a data contract from a JSON Schema file. │
│ json Import a data contract from a JSON file. │
│ odcs Import a data contract from an ODCS file. │
│ parquet Import a data contract from a Parquet file. │
│ csv Import a data contract from a CSV file. │
│ protobuf Import a data contract from a Protobuf schema file. │
│ spark Import a data contract from a Spark schema. │
│ iceberg Import a data contract from an Iceberg schema. │
│ excel Import a data contract from an Excel file. │
╰──────────────────────────────────────────────────────────────────────────────────────────────────╯
Example: datacontract import sql --source ddl.sql --dialect postgres --output datacontract.yaml
Run datacontract import <format> --help to see the format-specific options (e.g. datacontract import sql --help). If you are missing a format, please create an issue on GitHub.
Examples
# Example import from SQL DDL
datacontract import sql --source my_ddl.sql --dialect postgres
# To save to file
datacontract import sql --source my_ddl.sql --dialect postgres --output datacontract.yaml
BigQuery
BigQuery data can either be imported off of JSON Files generated from the table descriptions or directly from the Bigquery API. In case you want to use JSON Files, specify the source parameter with a path to the JSON File.
To import from the Bigquery API, you have to omit source and instead need to provide --project and --dataset. Additionally you may specify --table to enumerate the tables that should be imported. If no tables are given, all available tables of the dataset will be imported.
For providing authentication to the Client, please see the google documentation or the one about authorizing client libraries.
Examples:
# Example import from Bigquery JSON
datacontract import bigquery --source my_bigquery_table.json
# Example import from Bigquery API with specifying the tables to import
datacontract import bigquery --project <project_id> --dataset <dataset_id> --table <tableid_1> --table <tableid_2> --table <tableid_3>
# Example import from Bigquery API importing all tables in the dataset
datacontract import bigquery --project <project_id> --dataset <dataset_id>
Unity Catalog
# Example import from a Unity Catalog JSON file
datacontract import unity --source my_unity_table.json
# Example import single table from Unity Catalog via HTTP endpoint using PAT
export DATACONTRACT_DATABRICKS_SERVER_HOSTNAME="https://xyz.cloud.databricks.com"
export DATACONTRACT_DATABRICKS_TOKEN=<token>
datacontract import unity --table <table_full_name>
Please refer to Databricks documentation on how to set up a profile
# Example import single table from Unity Catalog via HTTP endpoint using Profile
export DATACONTRACT_DATABRICKS_PROFILE="my-profile"
datacontract import unity --table <table_full_name>
dbt
Importing from dbt manifest file.
You may give the --model parameter to enumerate the tables that should be imported. If no tables are given, all available tables of the database will be imported.
Examples:
# Example import from dbt manifest with specifying the tables to import
datacontract import dbt --source <manifest_path> --model <model_name_1> --model <model_name_2> --model <model_name_3>
# Example import from dbt manifest importing all tables in the database
datacontract import dbt --source <manifest_path>
Excel
Importing from ODCS Excel Template.
Examples:
# Example import from ODCS Excel Template
datacontract import excel --source odcs.xlsx
Glue
Importing from Glue reads the necessary Data directly off of the AWS API.
You may give the --table parameter to enumerate the tables that should be imported. If no tables are given, all available tables of the database will be imported.
Examples:
# Example import from AWS Glue with specifying the tables to import
datacontract import glue --database <database_name> --table <table_name_1> --table <table_name_2> --table <table_name_3>
# Example import from AWS Glue importing all tables in the database
datacontract import glue --database <database_name>
Spark
Importing from Spark table or view these must be created or accessible in the Spark context. Specify tables list in the --tables option. If the tables are registered in Databricks and have a table-level description, it will also be added to the Data Contract Specification.
# Example: Import Spark table(s) from Spark context
datacontract import spark --tables "users,orders"
# Example: Import Spark table
DataContract.import_from_source("spark", "users")
DataContract.import_from_source(format = "spark", source = "users")
# Example: Import Spark dataframe
DataContract.import_from_source("spark", "users", dataframe = df_user)
DataContract.import_from_source(format = "spark", source = "users", dataframe = df_user)
# Example: Import Spark table + table description
DataContract.import_from_source("spark", "users", description = "description")
DataContract.import_from_source(format = "spark", source = "users", description = "description")
# Example: Import Spark dataframe + table description
DataContract.import_from_source("spark", "users", dataframe = df_user, description = "description")
DataContract.import_from_source(format = "spark", source = "users", dataframe = df_user, description = "description")
DBML
Importing from DBML Documents. NOTE: Since DBML does not have strict requirements on the types of columns, this import may create non-valid datacontracts, as not all types of fields can be properly mapped. In this case you will have to adapt the generated document manually. We also assume, that the description for models and fields is stored in a Note within the DBML model.
You may give the --table or --schema parameter to enumerate the tables or schemas that should be imported.
If no tables are given, all available tables of the source will be imported. Likewise, if no schema is given, all schemas are imported.
Examples:
# Example import from DBML file, importing everything
datacontract import dbml --source <file_path>
# Example import from DBML file, filtering for tables from specific schemas
datacontract import dbml --source <file_path> --schema <schema_1> --schema <schema_2>
# Example import from DBML file, filtering for tables with specific names
datacontract import dbml --source <file_path> --table <table_name_1> --table <table_name_2>
# Example import from DBML file, filtering for tables with specific names from a specific schema
datacontract import dbml --source <file_path> --table <table_name_1> --schema <schema_1>
Iceberg
Importing from an Iceberg Table Json Schema Definition. Specify location of json files using the source parameter.
Examples:
datacontract import iceberg --source ./tests/fixtures/iceberg/simple_schema.json --table test-table
CSV
Importing from CSV File. Specify file in source parameter. It does autodetection for encoding and csv dialect
Example:
datacontract import csv --source "test.csv"
protobuf
Importing from protobuf File. Specify file in source parameter.
Example:
datacontract import protobuf --source "test.proto"
catalog
Usage: datacontract catalog [OPTIONS]
Create a html catalog of data contracts.
╭─ Options ────────────────────────────────────────────────────────────────────────────────────────╮
│ --files TEXT Glob pattern for the data contract files to include in the │
│ catalog. Applies recursively to any subfolders. │
│ [default: *.yaml] │
│ --output TEXT Output directory for the catalog html files. │
│ [default: catalog/] │
│ --json-schema TEXT The location (url or path) of the ODCS JSON Schema │
│ --debug --no-debug Enable debug logging │
│ --help Show this message and exit. │
╰──────────────────────────────────────────────────────────────────────────────────────────────────╯
Example: datacontract catalog --files "**/*.yaml" --output catalog/
Examples:
# create a catalog right in the current folder
datacontract catalog --output "."
# Create a catalog based on a filename convention
datacontract catalog --files "*.odcs.yaml"
publish
Usage: datacontract publish [OPTIONS] [LOCATION]
Publish the data contract to the Entropy Data.
╭─ Arguments ──────────────────────────────────────────────────────────────────────────────────────╮
│ location [LOCATION] The location (url or path) of the data contract yaml. │
│ [default: datacontract.yaml] │
╰──────────────────────────────────────────────────────────────────────────────────────────────────╯
╭─ Options ────────────────────────────────────────────────────────────────────────────────────────╮
│ --json-schema TEXT The location (url or path) of the ODCS JSON │
│ Schema │
│ --ssl-verification --no-ssl-verification SSL verification when publishing the data │
│ contract. │
│ [default: ssl-verification] │
│ --debug --no-debug Enable debug logging │
│ --help Show this message and exit. │
╰──────────────────────────────────────────────────────────────────────────────────────────────────╯
Example: datacontract publish datacontract.yaml
api
Usage: datacontract api [OPTIONS]
Start the datacontract CLI as server application with REST API.
The OpenAPI documentation as Swagger UI is available on http://localhost:4242.
You can execute the commands directly from the Swagger UI.
To protect the API, you can set the environment variable DATACONTRACT_CLI_API_KEY to a secret API
key.
To authenticate, requests must include the header 'x-api-key' with the correct API key.
This is highly recommended, as data contract tests may be subject to SQL injections or leak
sensitive information.
To connect to servers (such as a Snowflake data source), set the credentials as environment
variables as documented in
https://cli.datacontract.com/#test
It is possible to run the API with extra arguments for `uvicorn.run()` as keyword arguments, e.g.:
`datacontract api --port 1234 --root_path /datacontract`.
╭─ Options ────────────────────────────────────────────────────────────────────────────────────────╮
│ --port INTEGER Bind socket to this port. [default: 4242] │
│ --host TEXT Bind socket to this host. Hint: For running in docker, set it │
│ to 0.0.0.0 │
│ [default: 127.0.0.1] │
│ --debug --no-debug Enable debug logging │
│ --help Show this message and exit. │
╰──────────────────────────────────────────────────────────────────────────────────────────────────╯
Example: datacontract api --port 4242 --host 0.0.0.0
Integrations
| Integration | Option | Description |
|---|---|---|
| Entropy Data | --publish |
Push full results to the Entropy Data API |
Integration with Entropy Data
If you use Entropy Data, you can use the data contract URL to reference to the contract and append the --publish option to send and display the test results. Set an environment variable for your API key.
# Fetch current data contract, execute tests on production, and publish result to entropy data
$ EXPORT ENTROPY_DATA_API_KEY=xxx
$ datacontract test https://demo.entropy-data.com/demo279750347121/datacontracts/4df9d6ee-e55d-4088-9598-b635b2fdcbbc/datacontract.yaml \
--server production \
--publish https://api.entropy-data.com/api/test-results
Best Practices
We share best practices in using the Data Contract CLI.
Data-first Approach
Create a data contract based on the actual data. This is the fastest way to get started and to get feedback from the data consumers.
- Use an existing physical schema (e.g., SQL DDL) as a starting point to define your logical data model in the contract. Double check right after the import whether the actual data meets the imported logical data model. Just to be sure.
$ datacontract import sql --source ddl.sql $ datacontract test - Add quality checks and additional type constraints one by one to the contract and make sure the
data still adheres to the contract.
$ datacontract test - Validate that the
datacontract.yamlis correctly formatted and adheres to the Data Contract Specification.$ datacontract lint - Set up a CI pipeline that executes daily for continuous quality checks. Use the
cicommand for CI-optimized output (GitHub Actions annotations and step summary, Azure DevOps annotations). You can also report the test results to tools like Data Mesh Manager.$ datacontract ci --publish https://api.datamesh-manager.com/api/test-results
Contract-First
Create a data contract based on the requirements from use cases.
- Start with a
datacontract.yamltemplate.$ datacontract init - Create the model and quality guarantees based on your business requirements. Fill in the terms,
descriptions, etc. Validate that your
datacontract.yamlis correctly formatted.$ datacontract lint - Use the export function to start building the providing data product as well as the integration
into the consuming data products.
# data provider $ datacontract export dbt-models # data consumer $ datacontract export dbt-sources $ datacontract export dbt-staging-sql - Test that your data product implementation adheres to the contract.
$ datacontract test
Customizing Exporters and Importers
Custom Exporter
Using the exporter factory to add a new custom exporter
from datacontract.data_contract import DataContract
from datacontract.export.exporter import Exporter
from datacontract.export.exporter_factory import exporter_factory
# Create a custom class that implements export method
class CustomExporter(Exporter):
def export(self, data_contract, model, server, sql_server_type, export_args) -> dict:
result = {
"title": data_contract.info.title,
"version": data_contract.info.version,
"description": data_contract.info.description,
"email": data_contract.info.contact.email,
"url": data_contract.info.contact.url,
"model": model,
"model_columns": ", ".join(list(data_contract.models.get(model).fields.keys())),
"export_args": export_args,
"custom_args": export_args.get("custom_arg", ""),
}
return result
# Register the new custom class into factory
exporter_factory.register_exporter("custom_exporter", CustomExporter)
if __name__ == "__main__":
# Create a DataContract instance
data_contract = DataContract(
data_contract_file="/path/datacontract.yaml"
)
# Call export
result = data_contract.export(
export_format="custom_exporter", model="orders", server="production", custom_arg="my_custom_arg"
)
print(result)
Output
{
'title': 'Orders Unit Test',
'version': '1.0.0',
'description': 'The orders data contract',
'email': '[email protected]',
'url': 'https://wiki.example.com/teams/checkout',
'model': 'orders',
'model_columns': 'order_id, order_total, order_status',
'export_args': {'server': 'production', 'custom_arg': 'my_custom_arg'},
'custom_args': 'my_custom_arg'
}
Custom Importer
Using the importer factory to add a new custom importer
from datacontract.model.data_contract_specification import DataContractSpecification, Field, Model
from datacontract.data_contract import DataContract
from datacontract.imports.importer import Importer
from datacontract.imports.importer_factory import importer_factory
import json
# Create a custom class that implements import_source method
class CustomImporter(Importer):
def import_source(
self, data_contract_specification: DataContractSpecification, source: str, import_args: dict
) -> dict:
source_dict = json.loads(source)
data_contract_specification.id = source_dict.get("id_custom")
data_contract_specification.info.title = source_dict.get("title")
data_contract_specification.info.version = source_dict.get("version")
data_contract_specification.info.description = source_dict.get("description_from_app")
for model in source_dict.get("models", []):
fields = {}
for column in model.get('columns'):
field = Field(
description=column.get('column_description'),
type=column.get('type')
)
fields[column.get('name')] = field
dc_model = Model(
description=model.get('description'),
fields= fields
)
data_contract_specification.models[model.get('name')] = dc_model
return data_contract_specification
# Register the new custom class into factory
importer_factory.register_importer("custom_company_importer", CustomImporter)
if __name__ == "__main__":
# Get a custom data from other app
json_from_custom_app = '''
{
"id_custom": "uuid-custom",
"version": "0.0.2",
"title": "my_custom_imported_data",
"description_from_app": "Custom contract description",
"models": [
{
"name": "model1",
"description": "model description from app",
"columns": [
{
"name": "columnA",
"type": "varchar",
"column_description": "my_column description"
},
{
"name": "columnB",
"type": "varchar",
"column_description": "my_columnB description"
}
]
}
]
}
'''
# Create a DataContract instance
data_contract = DataContract()
# Call import_from_source
result = data_contract.import_from_source(
format="custom_company_importer",
data_contract_specification=DataContract.init(),
source=json_from_custom_app
)
print(result.to_yaml() )
Output
dataContractSpecification: 1.2.1
id: uuid-custom
info:
title: my_custom_imported_data
version: 0.0.2
description: Custom contract description
models:
model1:
fields:
columnA:
type: varchar
description: my_column description
columnB:
type: varchar
description: my_columnB description
Development Setup
- Install uv
- Python base interpreter should be 3.11.x.
- Docker engine must be running to execute the tests.
# make sure uv is installed
uv python pin 3.11
uv venv
uv pip install -e '.[dev]'
uv run ruff check
uv run pytest
Contribution
We are happy to receive your contributions. Propose your change in an issue or directly create a pull request with your improvements.
Before creating a pull request, please make sure that all tests are passing (uv run pytest) and
your code is properly formatted (ruff format). Create a changelog entry and reference fixed
issues (if any).
Troubleshooting
Windows: Some tests fail
Run in WSL. (We need to fix the paths in the tests so that normal Windows will work, contributions are appreciated)
PyCharm does not pick up the .venv
This uv issue might be relevant.
Try to sync all groups:
uv sync --all-groups --all-extras
Errors in tests that use PySpark (e.g. test_test_kafka.py)
Ensure you have a JDK 17 or 21 installed. Java 25 causes issues.
java --version
Docker Build
docker build -t datacontract/cli .
docker run --rm -v ${PWD}:/home/datacontract datacontract/cli
Docker compose integration
We’ve included a docker-compose.yml configuration to simplify the build, test, and deployment of the image.
Building the Image with Docker Compose
To build the Docker image using Docker Compose, run the following command:
docker compose build
This command utilizes the docker-compose.yml to build the image, leveraging predefined settings such as the build context and Dockerfile location. This approach streamlines the image creation process, avoiding the need for manual build specifications each time.
Testing the Image
After building the image, you can test it directly with Docker Compose:
docker compose run --rm datacontract --version
This command runs the container momentarily to check the version of the datacontract CLI. The --rm flag ensures that the container is automatically removed after the command executes, keeping your environment clean.
Release Steps
- Update the version in
pyproject.toml - Have a look at the
CHANGELOG.md - Create release commit manually
- Execute
./release - Wait until GitHub Release is created
- Add the release notes to the GitHub Release
Companies using this tool
- Entropy Data
- INNOQ
- Data Catering
- Oliver Wyman
- And many more. To add your company, please create a pull request.
Related Tools
- Entropy Data is a commercial tool to manage data contracts. It contains a web UI, access management, and data governance for a data product marketplace based on data contracts.
- Data Contract Editor is an editor for Data Contracts, including a live html preview.
- Data Contract Playground allows you to validate and export your data contract to different formats within your browser.
License
Credits
Created by Stefan Negele, Jochen Christ, and Simon Harrer.
