datacontract-cli 1.0.14__tar.gz → 1.0.16__tar.gz

{datacontract_cli-1.0.14/datacontract_cli.egg-info → datacontract_cli-1.0.16}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: datacontract-cli
-Version: 1.0.14
+Version: 1.0.16
 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>
 License-Expression: MIT
@@ -18,7 +18,7 @@ Requires-Dist: pyyaml~=6.0.1
 Requires-Dist: ruamel.yaml<0.20,>=0.17
 Requires-Dist: requests<2.35,>=2.31
 Requires-Dist: truststore<1.0,>=0.10
-Requires-Dist: fastjsonschema<2.22.0,>=2.19.1
+Requires-Dist: fastjsonschema<2.23.0,>=2.19.1
 Requires-Dist: jsonschema<5.0.0,>=4.23.0
 Requires-Dist: pytz>=2024.1
 Requires-Dist: python-multipart<1.0.0,>=0.0.20
@@ -39,11 +39,19 @@ Provides-Extra: csv
 Requires-Dist: pandas>=2.0.0; extra == "csv"
 Provides-Extra: excel
 Requires-Dist: openpyxl<4.0.0,>=3.1.5; extra == "excel"
+Provides-Extra: dataframe
+Requires-Dist: ibis-framework[pyspark]<13.0.0,>=10.0.0; extra == "dataframe"
+Requires-Dist: pyspark<5.0.0,>=3.5.0; extra == "dataframe"
 Provides-Extra: databricks
 Requires-Dist: ibis-framework[databricks,pyspark]<13.0.0,>=10.0.0; extra == "databricks"
-Requires-Dist: databricks-sql-connector<4.4.0,>=3.7.0; extra == "databricks"
-Requires-Dist: databricks-sdk<0.122.0; extra == "databricks"
+Requires-Dist: databricks-sql-connector<4.5.0,>=3.7.0; extra == "databricks"
+Requires-Dist: databricks-sdk<0.123.0; extra == "databricks"
 Requires-Dist: pyspark<5.0.0,>=3.5.0; extra == "databricks"
+Provides-Extra: databricks-runtime
+Requires-Dist: ibis-framework[databricks]<13.0.0,>=10.0.0; extra == "databricks-runtime"
+Requires-Dist: databricks-sql-connector<4.5.0,>=3.7.0; extra == "databricks-runtime"
+Requires-Dist: databricks-sdk<0.123.0; extra == "databricks-runtime"
+Requires-Dist: packaging<27.0,>=21.3; extra == "databricks-runtime"
 Provides-Extra: iceberg
 Requires-Dist: pyiceberg==0.11.1; extra == "iceberg"
 Provides-Extra: kafka
@@ -61,12 +69,12 @@ Requires-Dist: ibis-framework[postgres]<13.0.0,>=10.0.0; extra == "redshift"
 Requires-Dist: psycopg[binary]<4.0,>=3.1; extra == "redshift"
 Requires-Dist: packaging<27.0,>=21.3; extra == "redshift"
 Provides-Extra: s3
+Requires-Dist: datacontract-cli[duckdb]; extra == "s3"
 Requires-Dist: s3fs<2027.0.0,>=2025.2.0; extra == "s3"
 Requires-Dist: duckdb-extension-httpfs<1.6.0,>=1.5.0; extra == "s3"
 Requires-Dist: duckdb-extension-aws<1.6.0,>=1.5.0; extra == "s3"
 Provides-Extra: gcs
-Requires-Dist: datacontract-cli[duckdb]; extra == "gcs"
-Requires-Dist: duckdb-extension-httpfs<1.6.0,>=1.5.0; extra == "gcs"
+Requires-Dist: datacontract-cli[s3]; extra == "gcs"
 Provides-Extra: azure
 Requires-Dist: datacontract-cli[duckdb]; extra == "azure"
 Requires-Dist: duckdb-extension-azure<1.6.0,>=1.5.0; extra == "azure"
@@ -100,12 +108,12 @@ Requires-Dist: pyarrow>=18.1.0; extra == "parquet"
 Provides-Extra: rdf
 Requires-Dist: rdflib==7.6.0; extra == "rdf"
 Provides-Extra: api
-Requires-Dist: fastapi<0.140.0,>=0.115.0; extra == "api"
+Requires-Dist: fastapi<0.141.0,>=0.115.0; extra == "api"
 Requires-Dist: uvicorn<0.52.0,>=0.44.0; extra == "api"
 Provides-Extra: protobuf
 Requires-Dist: proto-schema-parser<3.0,>=1.0; extra == "protobuf"
 Provides-Extra: all
-Requires-Dist: datacontract-cli[api,athena,azure,bigquery,csv,databricks,dbml,duckdb,excel,gcs,iceberg,impala,kafka,mysql,oracle,parquet,postgres,protobuf,rdf,redshift,s3,snowflake,sqlserver,trino]; extra == "all"
+Requires-Dist: datacontract-cli[api,athena,azure,bigquery,csv,databricks,dataframe,dbml,duckdb,excel,gcs,iceberg,impala,kafka,mysql,oracle,parquet,postgres,protobuf,rdf,redshift,s3,snowflake,sqlserver,trino]; extra == "all"
 Provides-Extra: dev
 Requires-Dist: datacontract-cli[all]; extra == "dev"
 Requires-Dist: httpx==0.28.1; extra == "dev"
@@ -119,7 +127,7 @@ Requires-Dist: pytest-xdist; extra == "dev"
 Requires-Dist: pymssql==2.3.13; extra == "dev"
 Requires-Dist: psycopg2-binary<3.0,>=2.9; extra == "dev"
 Requires-Dist: mysql-connector-python<9.7.0,>=8.0.30; extra == "dev"
-Requires-Dist: ruff==0.15.22; extra == "dev"
+Requires-Dist: ruff==0.16.0; extra == "dev"
 Requires-Dist: testcontainers[kafka,minio,mssql,mysql,postgres]==4.14.2; extra == "dev"
 Requires-Dist: trino==0.338.0; extra == "dev"
 Dynamic: license-file
@@ -131,6 +139,8 @@ Dynamic: license-file
     <img alt="Test Workflow" src="https://img.shields.io/github/actions/workflow/status/datacontract/datacontract-cli/ci.yaml?branch=main"></a>
   <a href="https://pypi.org/project/datacontract-cli/">
     <img alt="PyPI Version" src="https://img.shields.io/pypi/v/datacontract-cli" /></a>
+  <a href="https://pypistats.org/packages/datacontract-cli">
+    <img alt="PyPI Monthly Downloads" src="https://img.shields.io/badge/downloads-1.4M%2Fmonth-brightgreen" /></a>
   <a href="https://github.com/datacontract/datacontract-cli">
     <img alt="Stars" src="https://img.shields.io/github/stars/datacontract/datacontract-cli" /></a>
   <a href="https://datacontract.com/slack" rel="nofollow"><img src="https://img.shields.io/badge/slack-join_chat-white.svg?logo=slack&amp;style=social" alt="Slack Status" data-canonical-src="https://img.shields.io/badge/slack-join_chat-white.svg?logo=slack&amp;style=social" style="max-width: 100%;"></a>
@@ -145,7 +155,9 @@ It can be used as a standalone CLI tool, in a CI/CD pipeline, or directly as a P
 
 > 📖 **Full documentation: [docs.datacontract.com](https://docs.datacontract.com)**
 >
-> Quick links: [Quickstart](https://docs.datacontract.com/quickstart) · [Commands](https://docs.datacontract.com/commands) · [Best Practices](https://docs.datacontract.com/best-practices) · [Custom Export and Import](https://docs.datacontract.com/extending) · [Development Setup](#development-setup)
+> Quick links: [Quickstart](https://docs.datacontract.com/quickstart) · [Commands](https://docs.datacontract.com/commands) · [Best Practices](https://docs.datacontract.com/best-practices) · [Custom Export and Import](https://docs.datacontract.com/extending) · [Release Notes](https://docs.datacontract.com/release-notes) · [Development Setup](#development-setup)
+>
+> For LLMs: [cli.datacontract.com/llms.txt](https://cli.datacontract.com/llms.txt) · [docs.datacontract.com/llms.txt](https://docs.datacontract.com/llms.txt) · [docs.datacontract.com/llms-full.txt](https://docs.datacontract.com/llms-full.txt)
 
 ## Getting started
 
@@ -207,6 +219,15 @@ Server: production (type=postgres, host=aws-1-eu-central-2.pooler.supabase.com,
 
 Voilà, the CLI tested that the YAML itself is valid, all records comply with the schema, and all quality attributes are met.
 
+To test your own data, import a contract straight from an existing table (this also generates the `servers` block) and run the tests against the actual data:
+
+```bash
+$ datacontract import snowflake --source <account> --database ORDER_DB --schema PUBLIC --output datacontract.yaml
+$ datacontract test datacontract.yaml
+```
+
+Copy-paste guides with credentials setup are available for [Snowflake](https://docs.datacontract.com/testing/snowflake), [BigQuery](https://docs.datacontract.com/testing/bigquery), [Databricks](https://docs.datacontract.com/testing/databricks), and [15+ other sources](https://docs.datacontract.com/testing).
+
 We can also use the data contract metadata to export in many [formats](https://docs.datacontract.com/exports), e.g., to generate a SQL DDL:
 
 ```bash
@@ -216,17 +237,17 @@ $ datacontract export sql https://datacontract.com/orders-v1.odcs.yaml
 -- Data Contract: orders
 -- SQL Dialect: postgres
 CREATE TABLE orders (
-  order_id None not null primary key,
+  order_id UUID not null primary key,
   customer_id text not null,
   order_total integer not null,
-  order_timestamp None,
+  order_timestamp TIMESTAMPTZ,
   order_status text
 );
 CREATE TABLE line_items (
-  line_item_id None not null primary key,
+  line_item_id UUID not null primary key,
   sku text not null,
   price integer not null,
-  order_id None
+  order_id UUID
 );
 ```
 
@@ -380,6 +401,8 @@ A list of available extras:
 | Google BigQuery                          | `pip install datacontract-cli[bigquery]`   |
 | CSV                                      | `pip install datacontract-cli[csv]`        |
 | Databricks Integration                   | `pip install datacontract-cli[databricks]` |
+| Databricks Runtime                       | `pip install datacontract-cli[databricks-runtime]` (inside Databricks, where the cluster provides PySpark) |
+| DataFrame (Spark)                        | `pip install datacontract-cli[dataframe]`  |
 | DBML                                     | `pip install datacontract-cli[dbml]`       |
 | DuckDB (local/S3/GCS/Azure file testing) | `pip install datacontract-cli[duckdb]`     |
 | Excel                                    | `pip install datacontract-cli[excel]`      |
@@ -409,7 +432,7 @@ It covers everything in depth, including the complete command reference:
 
 - [Quickstart](https://docs.datacontract.com/quickstart) — install and run your first test
 - [Open Data Contract Standard](https://docs.datacontract.com/open-data-contract-standard) — the contract format
-- [Test your contract](https://docs.datacontract.com/testing) and [Connect your Data](https://docs.datacontract.com/connect) — schema & quality tests against 18+ data sources
+- [Test your Data](https://docs.datacontract.com/testing) — schema & quality tests against 18+ data sources
 - [Define your Quality Rules](https://docs.datacontract.com/quality-rules) — SQL, library, text, and custom checks
 - [Sync with dbt](https://docs.datacontract.com/dbt) · [Edit your contract](https://docs.datacontract.com/editor)
 - [Imports](https://docs.datacontract.com/imports) and [Exports](https://docs.datacontract.com/exports) — convert to/from 25+ formats

{datacontract_cli-1.0.14 → datacontract_cli-1.0.16}/README.md

@@ -5,6 +5,8 @@
     <img alt="Test Workflow" src="https://img.shields.io/github/actions/workflow/status/datacontract/datacontract-cli/ci.yaml?branch=main"></a>
   <a href="https://pypi.org/project/datacontract-cli/">
     <img alt="PyPI Version" src="https://img.shields.io/pypi/v/datacontract-cli" /></a>
+  <a href="https://pypistats.org/packages/datacontract-cli">
+    <img alt="PyPI Monthly Downloads" src="https://img.shields.io/badge/downloads-1.4M%2Fmonth-brightgreen" /></a>
   <a href="https://github.com/datacontract/datacontract-cli">
     <img alt="Stars" src="https://img.shields.io/github/stars/datacontract/datacontract-cli" /></a>
   <a href="https://datacontract.com/slack" rel="nofollow"><img src="https://img.shields.io/badge/slack-join_chat-white.svg?logo=slack&amp;style=social" alt="Slack Status" data-canonical-src="https://img.shields.io/badge/slack-join_chat-white.svg?logo=slack&amp;style=social" style="max-width: 100%;"></a>
@@ -19,7 +21,9 @@ It can be used as a standalone CLI tool, in a CI/CD pipeline, or directly as a P
 
 > 📖 **Full documentation: [docs.datacontract.com](https://docs.datacontract.com)**
 >
-> Quick links: [Quickstart](https://docs.datacontract.com/quickstart) · [Commands](https://docs.datacontract.com/commands) · [Best Practices](https://docs.datacontract.com/best-practices) · [Custom Export and Import](https://docs.datacontract.com/extending) · [Development Setup](#development-setup)
+> Quick links: [Quickstart](https://docs.datacontract.com/quickstart) · [Commands](https://docs.datacontract.com/commands) · [Best Practices](https://docs.datacontract.com/best-practices) · [Custom Export and Import](https://docs.datacontract.com/extending) · [Release Notes](https://docs.datacontract.com/release-notes) · [Development Setup](#development-setup)
+>
+> For LLMs: [cli.datacontract.com/llms.txt](https://cli.datacontract.com/llms.txt) · [docs.datacontract.com/llms.txt](https://docs.datacontract.com/llms.txt) · [docs.datacontract.com/llms-full.txt](https://docs.datacontract.com/llms-full.txt)
 
 ## Getting started
 
@@ -81,6 +85,15 @@ Server: production (type=postgres, host=aws-1-eu-central-2.pooler.supabase.com,
 
 Voilà, the CLI tested that the YAML itself is valid, all records comply with the schema, and all quality attributes are met.
 
+To test your own data, import a contract straight from an existing table (this also generates the `servers` block) and run the tests against the actual data:
+
+```bash
+$ datacontract import snowflake --source <account> --database ORDER_DB --schema PUBLIC --output datacontract.yaml
+$ datacontract test datacontract.yaml
+```
+
+Copy-paste guides with credentials setup are available for [Snowflake](https://docs.datacontract.com/testing/snowflake), [BigQuery](https://docs.datacontract.com/testing/bigquery), [Databricks](https://docs.datacontract.com/testing/databricks), and [15+ other sources](https://docs.datacontract.com/testing).
+
 We can also use the data contract metadata to export in many [formats](https://docs.datacontract.com/exports), e.g., to generate a SQL DDL:
 
 ```bash
@@ -90,17 +103,17 @@ $ datacontract export sql https://datacontract.com/orders-v1.odcs.yaml
 -- Data Contract: orders
 -- SQL Dialect: postgres
 CREATE TABLE orders (
-  order_id None not null primary key,
+  order_id UUID not null primary key,
   customer_id text not null,
   order_total integer not null,
-  order_timestamp None,
+  order_timestamp TIMESTAMPTZ,
   order_status text
 );
 CREATE TABLE line_items (
-  line_item_id None not null primary key,
+  line_item_id UUID not null primary key,
   sku text not null,
   price integer not null,
-  order_id None
+  order_id UUID
 );
 ```
 
@@ -254,6 +267,8 @@ A list of available extras:
 | Google BigQuery                          | `pip install datacontract-cli[bigquery]`   |
 | CSV                                      | `pip install datacontract-cli[csv]`        |
 | Databricks Integration                   | `pip install datacontract-cli[databricks]` |
+| Databricks Runtime                       | `pip install datacontract-cli[databricks-runtime]` (inside Databricks, where the cluster provides PySpark) |
+| DataFrame (Spark)                        | `pip install datacontract-cli[dataframe]`  |
 | DBML                                     | `pip install datacontract-cli[dbml]`       |
 | DuckDB (local/S3/GCS/Azure file testing) | `pip install datacontract-cli[duckdb]`     |
 | Excel                                    | `pip install datacontract-cli[excel]`      |
@@ -283,7 +298,7 @@ It covers everything in depth, including the complete command reference:
 
 - [Quickstart](https://docs.datacontract.com/quickstart) — install and run your first test
 - [Open Data Contract Standard](https://docs.datacontract.com/open-data-contract-standard) — the contract format
-- [Test your contract](https://docs.datacontract.com/testing) and [Connect your Data](https://docs.datacontract.com/connect) — schema & quality tests against 18+ data sources
+- [Test your Data](https://docs.datacontract.com/testing) — schema & quality tests against 18+ data sources
 - [Define your Quality Rules](https://docs.datacontract.com/quality-rules) — SQL, library, text, and custom checks
 - [Sync with dbt](https://docs.datacontract.com/dbt) · [Edit your contract](https://docs.datacontract.com/editor)
 - [Imports](https://docs.datacontract.com/imports) and [Exports](https://docs.datacontract.com/exports) — convert to/from 25+ formats

{datacontract_cli-1.0.14 → datacontract_cli-1.0.16}/datacontract/api.py

@@ -204,6 +204,25 @@ customProperties:
 contractCreatedTs: "2025-01-15T10:00:00Z"
 """
 
+
+def _require_schema_url(schema: str | None) -> str | None:
+    """Reject anything that is not an http(s) URL.
+
+    The parameter is documented as a URL, but `fetch_schema` falls through to the
+    filesystem for anything else — so without this an unauthenticated caller could
+    have the server open its own files, and tell existing paths from missing ones
+    by the error it got back.
+    """
+    if schema is None:
+        return None
+    if not schema.startswith(("http://", "https://")):
+        raise HTTPException(
+            status_code=status.HTTP_422_UNPROCESSABLE_ENTITY,
+            detail="The schema parameter must be an http:// or https:// URL.",
+        )
+    return schema
+
+
 app = FastAPI(
     docs_url="/",
     title="Data Contract CLI API",
@@ -365,7 +384,9 @@ async def lint(
         Query(description="Report all JSON Schema validation errors instead of only the first one."),
     ] = False,
 ):
-    data_contract = DataContract(data_contract_str=body, schema_location=schema, all_errors=all_errors)
+    data_contract = DataContract(
+        data_contract_str=body, schema_location=_require_schema_url(schema), all_errors=all_errors
+    )
     lint_result = data_contract.lint()
     return {"result": lint_result.result, "checks": lint_result.checks}
 

{datacontract_cli-1.0.14 → datacontract_cli-1.0.16}/datacontract/cli.py

@@ -9,6 +9,7 @@ import typer
 from click import Context
 from dotenv import find_dotenv, load_dotenv
 from rich.console import Console
+from rich.markup import escape
 from typer.core import TyperGroup
 from typing_extensions import Annotated
 
@@ -47,6 +48,10 @@ class OrderedCommands(TyperGroup):
 class OrderedCommandsWithMigrationHints(OrderedCommands):
     """Intercepts removed or renamed options on import/export and points the user to the v0.12.0 migration notes."""
 
+    # Import formats where `--schema` still means the database schema, so it must
+    # not be rewritten to the v0.12.0 `--json-schema`.
+    DATABASE_SCHEMA_IMPORTS = {"snowflake", "redshift", "postgres", "athena", "sqlserver", "oracle", "trino"}
+
     RENAMED_FLAGS = {
         "--format": None,
         "--rdf-base": "--base",
@@ -65,13 +70,17 @@ class OrderedCommandsWithMigrationHints(OrderedCommands):
         subcommand = positionals[0] if positionals else None
 
         # this function is called by both `datacontract` and `datacontract import`
-        is_import_snowflake = (positionals[:1] == ["snowflake"]) or (positionals[:2] == ["import", "snowflake"])
+        if subcommand == "import":
+            import_format = positionals[1] if len(positionals) > 1 else None
+        else:
+            import_format = subcommand
+        takes_database_schema = import_format in self.DATABASE_SCHEMA_IMPORTS
 
         rewritten_args = []
         for arg in args:
             if isinstance(arg, str) and arg.startswith("--"):
                 flag, _, value = arg.partition("=")
-                if flag == "--schema" and not is_import_snowflake:
+                if flag == "--schema" and not takes_database_schema:
                     typer.secho(
                         "Warning: --schema was replaced with --json-schema in v0.12.0 and will be removed in v0.13.0.",
                         err=True,
@@ -261,7 +270,9 @@ def main():
         from datacontract.model.exceptions import DataContractException
 
         message = e.reason if isinstance(e, DataContractException) else str(e)
-        console.print(f"[red]Error:[/red] {message}")
+        # Escape the message: bracketed text in an exception (e.g. a driver's
+        # `pip install "botocore[crt]"` hint) would otherwise be eaten as rich markup.
+        console.print(f"[red]Error:[/red] {escape(message)}")
         console.print("[dim]Pass --debug (or set DATACONTRACT_CLI_DEBUG=1) for the full traceback.[/dim]")
         sys.exit(1)
 

{datacontract_cli-1.0.14 → datacontract_cli-1.0.16}/datacontract/command_catalog.py

@@ -28,7 +28,7 @@ def catalog(
     debug: debug_option = None,
 ):
     """
-    Create a html catalog of data contracts.
+    Create an HTML catalog of data contracts.
     """
     enable_debug_logging(debug)