datacontract-cli 0.10.16__tar.gz → 0.10.18__tar.gz

{datacontract_cli-0.10.16/datacontract_cli.egg-info → datacontract_cli-0.10.18}/PKG-INFO

@@ -1,6 +1,6 @@
-Metadata-Version: 2.1
+Metadata-Version: 2.2
 Name: datacontract-cli
-Version: 0.10.16
+Version: 0.10.18
 Summary: The datacontract CLI is an open source command-line tool for working with Data Contracts. It uses data contract YAML files to lint the data contract, connect to data sources and execute schema and quality tests, detect breaking changes, 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.
 Author-email: Jochen Christ <jochen.christ@innoq.com>, Stefan Negele <stefan.negele@innoq.com>, Simon Harrer <simon.harrer@innoq.com>
 Project-URL: Homepage, https://cli.datacontract.com
@@ -11,64 +11,64 @@ Classifier: Operating System :: OS Independent
 Requires-Python: >=3.10
 Description-Content-Type: text/markdown
 License-File: LICENSE
-Requires-Dist: typer<0.15,>=0.12
+Requires-Dist: typer<0.16,>=0.15.1
 Requires-Dist: pydantic<2.11.0,>=2.8.2
 Requires-Dist: pyyaml~=6.0.1
 Requires-Dist: requests<2.33,>=2.31
 Requires-Dist: fastapi==0.115.6
-Requires-Dist: uvicorn==0.32.1
+Requires-Dist: uvicorn==0.34.0
 Requires-Dist: fastjsonschema<2.22.0,>=2.19.1
 Requires-Dist: fastparquet==2024.11.0
-Requires-Dist: python-multipart==0.0.19
+Requires-Dist: numpy<2.0.0,>=1.26.4
+Requires-Dist: python-multipart==0.0.20
 Requires-Dist: rich<13.10,>=13.7
 Requires-Dist: simple-ddl-parser==1.7.1
 Requires-Dist: duckdb==1.1.2
-Requires-Dist: soda-core-duckdb<3.5.0,>=3.3.1
+Requires-Dist: soda-core-duckdb<3.4.0,>=3.3.20
 Requires-Dist: setuptools>=60
 Requires-Dist: python-dotenv~=1.0.0
 Requires-Dist: rdflib==7.0.0
-Requires-Dist: opentelemetry-exporter-otlp-proto-grpc~=1.16
-Requires-Dist: opentelemetry-exporter-otlp-proto-http~=1.16
-Requires-Dist: boto3<1.35.69,>=1.34.41
+Requires-Dist: boto3<1.35.98,>=1.34.41
 Requires-Dist: jinja_partials>=0.2.1
 Provides-Extra: avro
 Requires-Dist: avro==1.12.0; extra == "avro"
 Provides-Extra: bigquery
-Requires-Dist: soda-core-bigquery<3.5.0,>=3.3.1; extra == "bigquery"
+Requires-Dist: soda-core-bigquery<3.4.0,>=3.3.20; extra == "bigquery"
 Provides-Extra: databricks
-Requires-Dist: soda-core-spark-df<3.5.0,>=3.3.1; extra == "databricks"
-Requires-Dist: databricks-sql-connector<3.6.0,>=3.1.2; extra == "databricks"
-Requires-Dist: databricks-sdk<0.39.0,>=0.32.0; extra == "databricks"
-Requires-Dist: soda-core-spark[databricks]<3.5.0,>=3.3.1; extra == "databricks"
+Requires-Dist: soda-core-spark-df<3.4.0,>=3.3.20; extra == "databricks"
+Requires-Dist: soda-core-spark[databricks]<3.4.0,>=3.3.20; extra == "databricks"
+Requires-Dist: databricks-sql-connector<3.8.0,>=3.7.0; extra == "databricks"
+Requires-Dist: databricks-sdk<0.41.0; extra == "databricks"
 Provides-Extra: iceberg
 Requires-Dist: pyiceberg==0.8.1; extra == "iceberg"
 Provides-Extra: kafka
 Requires-Dist: datacontract-cli[avro]; extra == "kafka"
-Requires-Dist: soda-core-spark-df<3.5.0,>=3.3.1; extra == "kafka"
+Requires-Dist: soda-core-spark-df<3.4.0,>=3.3.20; extra == "kafka"
 Provides-Extra: postgres
-Requires-Dist: soda-core-postgres<3.5.0,>=3.3.1; extra == "postgres"
+Requires-Dist: soda-core-postgres<3.4.0,>=3.3.20; extra == "postgres"
 Provides-Extra: s3
-Requires-Dist: s3fs==2024.10.0; extra == "s3"
+Requires-Dist: s3fs==2024.12.0; extra == "s3"
+Requires-Dist: aiobotocore<2.18.0,>=2.17.0; extra == "s3"
 Provides-Extra: snowflake
 Requires-Dist: snowflake-connector-python[pandas]<3.13,>=3.6; extra == "snowflake"
-Requires-Dist: soda-core-snowflake<3.5.0,>=3.3.1; extra == "snowflake"
+Requires-Dist: soda-core-snowflake<3.4.0,>=3.3.20; extra == "snowflake"
 Provides-Extra: sqlserver
-Requires-Dist: soda-core-sqlserver<3.5.0,>=3.3.1; extra == "sqlserver"
+Requires-Dist: soda-core-sqlserver<3.4.0,>=3.3.20; extra == "sqlserver"
 Provides-Extra: trino
-Requires-Dist: soda-core-trino<3.5.0,>=3.3.1; extra == "trino"
+Requires-Dist: soda-core-trino<3.4.0,>=3.3.20; extra == "trino"
 Provides-Extra: dbt
 Requires-Dist: dbt-core>=1.8.0; extra == "dbt"
 Provides-Extra: dbml
 Requires-Dist: pydbml>=1.1.1; extra == "dbml"
 Provides-Extra: parquet
-Requires-Dist: pyarrow>=12.0.0; extra == "parquet"
+Requires-Dist: pyarrow>=18.1.0; extra == "parquet"
 Provides-Extra: all
 Requires-Dist: datacontract-cli[bigquery,databricks,dbml,dbt,iceberg,kafka,parquet,postgres,s3,snowflake,sqlserver,trino]; extra == "all"
 Provides-Extra: dev
 Requires-Dist: datacontract-cli[all]; extra == "dev"
 Requires-Dist: httpx==0.28.1; extra == "dev"
 Requires-Dist: kafka-python; extra == "dev"
-Requires-Dist: moto==5.0.22; extra == "dev"
+Requires-Dist: moto==5.0.26; extra == "dev"
 Requires-Dist: pandas>=2.1.0; extra == "dev"
 Requires-Dist: pre-commit<4.1.0,>=3.7.1; extra == "dev"
 Requires-Dist: pytest; extra == "dev"
@@ -76,7 +76,7 @@ Requires-Dist: pytest-xdist; extra == "dev"
 Requires-Dist: pymssql==2.3.2; extra == "dev"
 Requires-Dist: ruff; extra == "dev"
 Requires-Dist: testcontainers[kafka,minio,mssql,postgres]==4.9.0; extra == "dev"
-Requires-Dist: trino==0.330.0; extra == "dev"
+Requires-Dist: trino==0.331.0; extra == "dev"
 
 # Data Contract CLI
 
@@ -105,7 +105,7 @@ This data contract contains all information to connect to S3 and check that the
 
 Let's use [pip](https://pip.pypa.io/en/stable/getting-started/) to install the CLI (or use the [Docker image](#docker)),
 ```bash
-$ python3 -m pip install datacontract-cli[all]
+$ python3 -m pip install 'datacontract-cli[all]'
 ```
 
 
@@ -230,13 +230,13 @@ Choose the most appropriate installation method for your needs:
 Python 3.10, 3.11, and 3.12 are supported. We recommend to use Python 3.11.
 
 ```bash
-python3 -m pip install datacontract-cli[all]
+python3 -m pip install 'datacontract-cli[all]'
 ```
 
 ### pipx
 pipx installs into an isolated environment.
 ```bash
-pipx install datacontract-cli[all]
+pipx install 'datacontract-cli[all]'
 ```
 
 ### Docker
@@ -301,6 +301,7 @@ Commands
 - [diff](#diff)
 - [catalog](#catalog)
 - [publish](#publish)
+- [serve](#serve)
 
 ### init
 
@@ -365,9 +366,6 @@ Commands
 │                                                                      [default: no-examples]                                     │
 │ --publish                                                      TEXT  The url to publish the results after the test              │
 │                                                                      [default: None]                                            │
-│ --publish-to-opentelemetry    --no-publish-to-opentelemetry          Publish the results to opentelemetry. Use environment      │
-│                                                                      variables to configure the OTLP endpoint, headers, etc.    │
-│                                                                      [default: no-publish-to-opentelemetry]                     │
 │ --logs                        --no-logs                              Print logs [default: no-logs]                              │
 │ --help                                                               Show this message and exit.                                │
 ╰─────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────╯
@@ -592,7 +590,6 @@ datacontract.yaml
 servers:
   production:
     type: databricks
-    host: dbc-abcdefgh-1234.cloud.databricks.com
     catalog: acme_catalog_prod
     schema: orders_latest
 models:
@@ -603,10 +600,11 @@ models:
 
 #### 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 |
+| 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)
@@ -841,7 +839,7 @@ models:
 │                    t-staging-sql|odcs|odcs_v2|odcs_v3|rdf|avro|protobuf                                                        │
 │                    |great-expectations|terraform|avro-idl|sql|sql-query                                                        │
 │                    |html|go|bigquery|dbml|spark|sqlalchemy|data-caterer                                                        │
-│                    |dcs|iceberg]                                                                                                       │
+│                    |dcs|markdown|iceberg]                                                                                              │
 │    --output        PATH                                                  Specify the file path where the exported data will be │
 │                                                                          saved. If no path is provided, the output will be     │
 │                                                                          printed to stdout.                                    │
@@ -902,6 +900,7 @@ Available export options:
 | `sqlalchemy`         | Export to SQLAlchemy Models                             | ✅      |
 | `data-caterer`       | Export to Data Caterer in YAML format                   | ✅      |
 | `dcs`                | Export to Data Contract Specification in YAML format    | ✅      |
+| `markdown`           | Export to Markdown                                      | ✅      |
 | `iceberg`            | Export to an Iceberg JSON Schema Definition             | partial |
 | Missing something?   | Please create an issue on GitHub                        | TBD    |
 
@@ -1389,13 +1388,28 @@ datacontract catalog --files "*.odcs.yaml"
 ╰────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────╯
 ```
 
+### Serve
+
+```
+
+ Usage: datacontract serve [OPTIONS]
+
+ Start the datacontract web server.
+
+╭─ Options ──────────────────────────────────────────────────────────────────────────────────────────────────────────────────╮
+│ --port        INTEGER  Bind socket to this port. [default: 4242]                                                           │
+│ --host        TEXT     Bind socket to this host. [default: 127.0.0.1]                                                      │
+│ --help                 Show this message and exit.                                                                         │
+╰────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────╯
+
+```
+
 ## Integrations
 
 | Integration           | Option                       | Description                                                                                                   |
 |-----------------------|------------------------------|---------------------------------------------------------------------------------------------------------------|
 | Data Mesh Manager     | `--publish`                  | Push full results to the [Data Mesh Manager API](https://api.datamesh-manager.com/swagger/index.html)         |
 | Data Contract Manager | `--publish`                  | Push full results to the [Data Contract Manager API](https://api.datacontract-manager.com/swagger/index.html) |
-| OpenTelemetry         | `--publish-to-opentelemetry` | Push result as gauge metrics                                                                                  |
 
 ### Integration with Data Mesh Manager
 
@@ -1409,36 +1423,6 @@ $ datacontract test https://demo.datamesh-manager.com/demo279750347121/datacontr
  --publish https://api.datamesh-manager.com/api/test-results
 ```
 
-### Integration with OpenTelemetry
-
-If you use OpenTelemetry, you can use the data contract URL and append the `--publish-to-opentelemetry` option to send the test results to your OLTP-compatible instance, e.g., Prometheus.
-
-The metric name is "datacontract.cli.test.result" and it uses the following encoding for the result:
-
-| datacontract.cli.test.result | Description                           |
-|------------------------------|---------------------------------------|
-| 0                            | test run passed, no warnings          |
-| 1                            | test run has warnings                 |
-| 2                            | test run failed                       |
-| 3                            | test run not possible due to an error |
-| 4                            | test status unknown                   |
-
-
-```bash
-# Fetch current data contract, execute tests on production, and publish result to open telemetry
-$ EXPORT OTEL_SERVICE_NAME=datacontract-cli
-$ EXPORT OTEL_EXPORTER_OTLP_ENDPOINT=https://YOUR_ID.apm.westeurope.azure.elastic-cloud.com:443
-$ EXPORT OTEL_EXPORTER_OTLP_HEADERS=Authorization=Bearer%20secret # Optional, when using SaaS Products
-$ EXPORT OTEL_EXPORTER_OTLP_PROTOCOL=http/protobuf # Optional, default is http/protobuf - use value grpc to use the gRPC protocol instead
-# Send to OpenTelemetry
-$ datacontract test https://demo.datamesh-manager.com/demo279750347121/datacontracts/4df9d6ee-e55d-4088-9598-b635b2fdcbbc/datacontract.yaml --server production --publish-to-opentelemetry
-```
-
-Current limitations:
-- currently, only ConsoleExporter and OTLP Exporter
-- Metrics only, no logs yet (but loosely planned)
-
-
 ## Best Practices
 
 We share best practices in using the Data Contract CLI.
@@ -1470,7 +1454,7 @@ Create a data contract based on the actual data. This is the fastest way to get
    $ datacontract lint
    ```
 
-5. Set up a CI pipeline that executes daily and reports the results to the [Data Mesh Manager](https://datamesh-manager.com). Or to some place else. You can even publish to any opentelemetry compatible system.
+5. Set up a CI pipeline that executes daily for continuous quality checks. You can also report the test results to tools like [Data Mesh Manager](https://datamesh-manager.com) 
    ```bash
    $ datacontract test --publish https://api.datamesh-manager.com/api/test-results
    ```

{datacontract_cli-0.10.16 → datacontract_cli-0.10.18}/README.md

@@ -25,7 +25,7 @@ This data contract contains all information to connect to S3 and check that the
 
 Let's use [pip](https://pip.pypa.io/en/stable/getting-started/) to install the CLI (or use the [Docker image](#docker)),
 ```bash
-$ python3 -m pip install datacontract-cli[all]
+$ python3 -m pip install 'datacontract-cli[all]'
 ```
 
 
@@ -150,13 +150,13 @@ Choose the most appropriate installation method for your needs:
 Python 3.10, 3.11, and 3.12 are supported. We recommend to use Python 3.11.
 
 ```bash
-python3 -m pip install datacontract-cli[all]
+python3 -m pip install 'datacontract-cli[all]'
 ```
 
 ### pipx
 pipx installs into an isolated environment.
 ```bash
-pipx install datacontract-cli[all]
+pipx install 'datacontract-cli[all]'
 ```
 
 ### Docker
@@ -221,6 +221,7 @@ Commands
 - [diff](#diff)
 - [catalog](#catalog)
 - [publish](#publish)
+- [serve](#serve)
 
 ### init
 
@@ -285,9 +286,6 @@ Commands
 │                                                                      [default: no-examples]                                     │
 │ --publish                                                      TEXT  The url to publish the results after the test              │
 │                                                                      [default: None]                                            │
-│ --publish-to-opentelemetry    --no-publish-to-opentelemetry          Publish the results to opentelemetry. Use environment      │
-│                                                                      variables to configure the OTLP endpoint, headers, etc.    │
-│                                                                      [default: no-publish-to-opentelemetry]                     │
 │ --logs                        --no-logs                              Print logs [default: no-logs]                              │
 │ --help                                                               Show this message and exit.                                │
 ╰─────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────╯
@@ -512,7 +510,6 @@ datacontract.yaml
 servers:
   production:
     type: databricks
-    host: dbc-abcdefgh-1234.cloud.databricks.com
     catalog: acme_catalog_prod
     schema: orders_latest
 models:
@@ -523,10 +520,11 @@ models:
 
 #### 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 |
+| 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)
@@ -761,7 +759,7 @@ models:
 │                    t-staging-sql|odcs|odcs_v2|odcs_v3|rdf|avro|protobuf                                                        │
 │                    |great-expectations|terraform|avro-idl|sql|sql-query                                                        │
 │                    |html|go|bigquery|dbml|spark|sqlalchemy|data-caterer                                                        │
-│                    |dcs|iceberg]                                                                                                       │
+│                    |dcs|markdown|iceberg]                                                                                              │
 │    --output        PATH                                                  Specify the file path where the exported data will be │
 │                                                                          saved. If no path is provided, the output will be     │
 │                                                                          printed to stdout.                                    │
@@ -822,6 +820,7 @@ Available export options:
 | `sqlalchemy`         | Export to SQLAlchemy Models                             | ✅      |
 | `data-caterer`       | Export to Data Caterer in YAML format                   | ✅      |
 | `dcs`                | Export to Data Contract Specification in YAML format    | ✅      |
+| `markdown`           | Export to Markdown                                      | ✅      |
 | `iceberg`            | Export to an Iceberg JSON Schema Definition             | partial |
 | Missing something?   | Please create an issue on GitHub                        | TBD    |
 
@@ -1309,13 +1308,28 @@ datacontract catalog --files "*.odcs.yaml"
 ╰────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────╯
 ```
 
+### Serve
+
+```
+
+ Usage: datacontract serve [OPTIONS]
+
+ Start the datacontract web server.
+
+╭─ Options ──────────────────────────────────────────────────────────────────────────────────────────────────────────────────╮
+│ --port        INTEGER  Bind socket to this port. [default: 4242]                                                           │
+│ --host        TEXT     Bind socket to this host. [default: 127.0.0.1]                                                      │
+│ --help                 Show this message and exit.                                                                         │
+╰────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────╯
+
+```
+
 ## Integrations
 
 | Integration           | Option                       | Description                                                                                                   |
 |-----------------------|------------------------------|---------------------------------------------------------------------------------------------------------------|
 | Data Mesh Manager     | `--publish`                  | Push full results to the [Data Mesh Manager API](https://api.datamesh-manager.com/swagger/index.html)         |
 | Data Contract Manager | `--publish`                  | Push full results to the [Data Contract Manager API](https://api.datacontract-manager.com/swagger/index.html) |
-| OpenTelemetry         | `--publish-to-opentelemetry` | Push result as gauge metrics                                                                                  |
 
 ### Integration with Data Mesh Manager
 
@@ -1329,36 +1343,6 @@ $ datacontract test https://demo.datamesh-manager.com/demo279750347121/datacontr
  --publish https://api.datamesh-manager.com/api/test-results
 ```
 
-### Integration with OpenTelemetry
-
-If you use OpenTelemetry, you can use the data contract URL and append the `--publish-to-opentelemetry` option to send the test results to your OLTP-compatible instance, e.g., Prometheus.
-
-The metric name is "datacontract.cli.test.result" and it uses the following encoding for the result:
-
-| datacontract.cli.test.result | Description                           |
-|------------------------------|---------------------------------------|
-| 0                            | test run passed, no warnings          |
-| 1                            | test run has warnings                 |
-| 2                            | test run failed                       |
-| 3                            | test run not possible due to an error |
-| 4                            | test status unknown                   |
-
-
-```bash
-# Fetch current data contract, execute tests on production, and publish result to open telemetry
-$ EXPORT OTEL_SERVICE_NAME=datacontract-cli
-$ EXPORT OTEL_EXPORTER_OTLP_ENDPOINT=https://YOUR_ID.apm.westeurope.azure.elastic-cloud.com:443
-$ EXPORT OTEL_EXPORTER_OTLP_HEADERS=Authorization=Bearer%20secret # Optional, when using SaaS Products
-$ EXPORT OTEL_EXPORTER_OTLP_PROTOCOL=http/protobuf # Optional, default is http/protobuf - use value grpc to use the gRPC protocol instead
-# Send to OpenTelemetry
-$ datacontract test https://demo.datamesh-manager.com/demo279750347121/datacontracts/4df9d6ee-e55d-4088-9598-b635b2fdcbbc/datacontract.yaml --server production --publish-to-opentelemetry
-```
-
-Current limitations:
-- currently, only ConsoleExporter and OTLP Exporter
-- Metrics only, no logs yet (but loosely planned)
-
-
 ## Best Practices
 
 We share best practices in using the Data Contract CLI.
@@ -1390,7 +1374,7 @@ Create a data contract based on the actual data. This is the fastest way to get
    $ datacontract lint
    ```
 
-5. Set up a CI pipeline that executes daily and reports the results to the [Data Mesh Manager](https://datamesh-manager.com). Or to some place else. You can even publish to any opentelemetry compatible system.
+5. Set up a CI pipeline that executes daily for continuous quality checks. You can also report the test results to tools like [Data Mesh Manager](https://datamesh-manager.com) 
    ```bash
    $ datacontract test --publish https://api.datamesh-manager.com/api/test-results
    ```

{datacontract_cli-0.10.16 → datacontract_cli-0.10.18}/datacontract/breaking/breaking_rules.py

@@ -42,6 +42,10 @@ class BreakingRules:
     field_primary_removed = Severity.WARNING
     field_primary_updated = Severity.WARNING
 
+    field_primary_key_added = Severity.WARNING
+    field_primary_key_removed = Severity.WARNING
+    field_primary_key_updated = Severity.WARNING
+
     field_references_added = Severity.WARNING
     field_references_removed = Severity.WARNING
     field_references_updated = Severity.WARNING

{datacontract_cli-0.10.16 → datacontract_cli-0.10.18}/datacontract/cli.py

@@ -132,13 +132,11 @@ def test(
         typer.Option(help="Run the schema and quality tests on the example data within the data contract."),
     ] = None,
     publish: Annotated[str, typer.Option(help="The url to publish the results after the test")] = None,
-    publish_to_opentelemetry: Annotated[
-        bool,
-        typer.Option(
-            help="Publish the results to opentelemetry. Use environment variables to configure the OTLP endpoint, headers, etc."
-        ),
-    ] = False,
     logs: Annotated[bool, typer.Option(help="Print logs")] = False,
+    ssl_verification: Annotated[
+        bool,
+        typer.Option(help="SSL verification when publishing the test results."),
+    ] = True,
 ):
     """
     Run schema and quality tests on configured servers.
@@ -150,7 +148,6 @@ def test(
         data_contract_file=location,
         schema_location=schema,
         publish_url=publish,
-        publish_to_opentelemetry=publish_to_opentelemetry,
         server=server,
         examples=examples,
     ).test()
@@ -280,6 +277,14 @@ def import_(
         Optional[str],
         typer.Option(help="Table name to assign to the model created from the Iceberg schema."),
     ] = None,
+    template: Annotated[
+        Optional[str],
+        typer.Option(help="The location (url or path) of the Data Contract Specification Template"),
+    ] = None,
+    schema: Annotated[
+        str,
+        typer.Option(help="The location (url or path) of the Data Contract Specification JSON Schema"),
+    ] = DEFAULT_DATA_CONTRACT_SCHEMA_URL,
 ):
     """
     Create a data contract from the given source location. Saves to file specified by `output` option if present, otherwise prints to stdout.
@@ -287,6 +292,8 @@ def import_(
     result = DataContract().import_from_source(
         format=format,
         source=source,
+        template=template,
+        schema=schema,
         glue_table=glue_table,
         bigquery_table=bigquery_table,
         bigquery_project=bigquery_project,
@@ -315,6 +322,10 @@ def publish(
         str,
         typer.Option(help="The location (url or path) of the Data Contract Specification JSON Schema"),
     ] = DEFAULT_DATA_CONTRACT_SCHEMA_URL,
+    ssl_verification: Annotated[
+        bool,
+        typer.Option(help="SSL verification when publishing the data contract."),
+    ] = True,
 ):
     """
     Publish the data contract to the Data Mesh Manager.
@@ -323,6 +334,7 @@ def publish(
         data_contract_specification=DataContract(
             data_contract_file=location, schema_location=schema
         ).get_data_contract_specification(),
+        ssl_verification=ssl_verification,
     )
 
 
@@ -447,6 +459,18 @@ def _handle_result(run):
         console.print(
             f"🟢 data contract is valid. Run {len(run.checks)} checks. Took {(run.timestampEnd - run.timestampStart).total_seconds()} seconds."
         )
+    elif run.result == "warning":
+        console.print("🟠 data contract has warnings. Found the following warnings:")
+        i = 1
+        for check in run.checks:
+            if check.result != "passed":
+                field = to_field(run, check)
+                if field:
+                    field = field + " "
+                else:
+                    field = ""
+                console.print(f"{i}) {field}{check.name}: {check.reason}")
+                i += 1
     else:
         console.print("🔴 data contract is invalid, found the following errors:")
         i = 1

{datacontract_cli-0.10.16 → datacontract_cli-0.10.18}/datacontract/data_contract.py

@@ -23,7 +23,6 @@ from datacontract.export.exporter import ExportFormat
 from datacontract.export.exporter_factory import exporter_factory
 from datacontract.imports.importer_factory import importer_factory
 from datacontract.integration.datamesh_manager import publish_test_results_to_datamesh_manager
-from datacontract.integration.opentelemetry import publish_test_results_to_opentelemetry
 from datacontract.lint import resolve
 from datacontract.lint.linters.description_linter import DescriptionLinter
 from datacontract.lint.linters.example_model_linter import ExampleModelLinter
@@ -37,6 +36,8 @@ from datacontract.model.data_contract_specification import DataContractSpecifica
 from datacontract.model.exceptions import DataContractException
 from datacontract.model.run import Check, Run
 
+DEFAULT_DATA_CONTRACT_TEMPLATE_URL = "https://datacontract.com/datacontract.init.yaml"
+
 
 class DataContract:
     def __init__(
@@ -48,7 +49,6 @@ class DataContract:
         server: str = None,
         examples: bool = False,
         publish_url: str = None,
-        publish_to_opentelemetry: bool = False,
         spark: "SparkSession" = None,
         inline_definitions: bool = True,
         inline_quality: bool = True,
@@ -60,7 +60,6 @@ class DataContract:
         self._server = server
         self._examples = examples
         self._publish_url = publish_url
-        self._publish_to_opentelemetry = publish_to_opentelemetry
         self._spark = spark
         self._inline_definitions = inline_definitions
         self._inline_quality = inline_quality
@@ -75,8 +74,10 @@ class DataContract:
         }
 
     @classmethod
-    def init(cls, template: str = "https://datacontract.com/datacontract.init.yaml") -> DataContractSpecification:
-        return resolve.resolve_data_contract(data_contract_location=template)
+    def init(
+        cls, template: str = DEFAULT_DATA_CONTRACT_TEMPLATE_URL, schema: typing.Optional[str] = None
+    ) -> DataContractSpecification:
+        return resolve.resolve_data_contract(data_contract_location=template, schema_location=schema)
 
     def lint(self, enabled_linters: typing.Union[str, set[str]] = "all") -> Run:
         """Lint the data contract by deserializing the contract and checking the schema, as well as calling the configured linters.
@@ -232,9 +233,6 @@ class DataContract:
         if self._publish_url is not None:
             publish_test_results_to_datamesh_manager(run, self._publish_url)
 
-        if self._publish_to_opentelemetry:
-            publish_test_results_to_opentelemetry(run)
-
         return run
 
     def _get_examples_server(self, data_contract, run, tmp_dir):
@@ -347,9 +345,15 @@ class DataContract:
         )
 
     def import_from_source(
-        self, format: str, source: typing.Optional[str] = None, **kwargs
+        self,
+        format: str,
+        source: typing.Optional[str] = None,
+        template: typing.Optional[str] = None,
+        schema: typing.Optional[str] = None,
+        **kwargs,
     ) -> DataContractSpecification:
-        data_contract_specification_initial = DataContract.init()
+        template = DEFAULT_DATA_CONTRACT_TEMPLATE_URL if template is None else template
+        data_contract_specification_initial = DataContract.init(template=template, schema=schema)
 
         return importer_factory.create(format).import_source(
             data_contract_specification=data_contract_specification_initial, source=source, import_args=kwargs

{datacontract_cli-0.10.16 → datacontract_cli-0.10.18}/datacontract/engines/fastjsonschema/check_jsonschema.py

@@ -11,7 +11,7 @@ from datacontract.engines.fastjsonschema.s3.s3_read_files import yield_s3_files
 from datacontract.export.jsonschema_converter import to_jsonschema
 from datacontract.model.data_contract_specification import DataContractSpecification, Server
 from datacontract.model.exceptions import DataContractException
-from datacontract.model.run import Check, Run
+from datacontract.model.run import Check, ResultEnum, Run
 
 # Thread-safe cache for primaryKey fields.
 _primary_key_cache = {}
@@ -256,18 +256,29 @@ def check_jsonschema(run: Run, data_contract: DataContractSpecification, server:
                     type="schema",
                     name="Check that JSON has valid schema",
                     model=model_name,
-                    result="info",
+                    result=ResultEnum.info,
                     reason="JSON Schema check skipped for GCS, as GCS is currently not supported",
                     engine="jsonschema",
                 )
             )
+        elif server.type == "azure":
+            run.checks.append(
+                Check(
+                    type="schema",
+                    name="Check that JSON has valid schema",
+                    model=model_name,
+                    result=ResultEnum.info,
+                    reason="JSON Schema check skipped for azure, as azure is currently not supported",
+                    engine="jsonschema",
+                )
+            )
         else:
             run.checks.append(
                 Check(
                     type="schema",
                     name="Check that JSON has valid schema",
                     model=model_name,
-                    result="warn",
+                    result=ResultEnum.warning,
                     reason=f"Server type {server.type} not supported",
                     engine="jsonschema",
                 )
@@ -279,7 +290,7 @@ def check_jsonschema(run: Run, data_contract: DataContractSpecification, server:
                 type="schema",
                 name="Check that JSON has valid schema",
                 model=model_name,
-                result="passed",
+                result=ResultEnum.passed,
                 reason="All JSON entries are valid.",
                 engine="jsonschema",
             )