datacontract-cli 1.0.5__tar.gz → 1.0.7__tar.gz

{datacontract_cli-1.0.5/datacontract_cli.egg-info → datacontract_cli-1.0.7}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: datacontract-cli
-Version: 1.0.5
+Version: 1.0.7
 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
@@ -16,6 +16,7 @@ Requires-Dist: click<9.0.0,>=8.1.0
 Requires-Dist: pydantic<2.14.0,>=2.8.2
 Requires-Dist: pyyaml~=6.0.1
 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: jsonschema<5.0.0,>=4.23.0
 Requires-Dist: pytz>=2024.1
@@ -40,7 +41,7 @@ Requires-Dist: openpyxl<4.0.0,>=3.1.5; extra == "excel"
 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<0.118.0; extra == "databricks"
+Requires-Dist: databricks-sdk<0.119.0; extra == "databricks"
 Requires-Dist: pyspark<5.0.0,>=3.5.0; extra == "databricks"
 Provides-Extra: iceberg
 Requires-Dist: pyiceberg==0.11.1; extra == "iceberg"
@@ -65,9 +66,9 @@ Requires-Dist: datacontract-cli[duckdb]; extra == "gcs"
 Requires-Dist: duckdb-extension-httpfs<1.6.0,>=1.5.0; extra == "gcs"
 Provides-Extra: azure
 Requires-Dist: datacontract-cli[duckdb]; extra == "azure"
-Requires-Dist: duckdb-extension-azure<1.1.0,>=1.0.3; (platform_system != "Linux" or platform_machine != "aarch64") and extra == "azure"
+Requires-Dist: duckdb-extension-azure<1.6.0,>=1.5.0; extra == "azure"
 Requires-Dist: azure-identity<1.18,>=1.17.1; extra == "azure"
-Requires-Dist: azure-storage-blob<12.30,>=12.24; extra == "azure"
+Requires-Dist: azure-storage-blob<12.31,>=12.24; extra == "azure"
 Provides-Extra: snowflake
 Requires-Dist: snowflake-connector-python[pandas]<4.7,>=3.6; extra == "snowflake"
 Requires-Dist: ibis-framework[snowflake]<13.0.0,>=10.0.0; extra == "snowflake"
@@ -91,7 +92,7 @@ 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<0.138.0,>=0.115.0; extra == "api"
+Requires-Dist: fastapi<0.139.0,>=0.115.0; extra == "api"
 Requires-Dist: uvicorn<0.50.0,>=0.44.0; extra == "api"
 Provides-Extra: protobuf
 Requires-Dist: proto-schema-parser<3.0,>=1.0; extra == "protobuf"
@@ -110,7 +111,7 @@ 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==0.15.17; extra == "dev"
+Requires-Dist: ruff==0.15.18; extra == "dev"
 Requires-Dist: testcontainers[kafka,minio,mssql,mysql,postgres]==4.14.2; extra == "dev"
 Requires-Dist: trino==0.337.0; extra == "dev"
 Dynamic: license-file

{datacontract_cli-1.0.5 → datacontract_cli-1.0.7}/datacontract/cli.py

@@ -111,6 +111,19 @@ def version_callback(value: bool):
         raise typer.Exit()
 
 
+def inject_system_truststore() -> None:
+    """Verify TLS using the operating system's certificate trust store instead of the
+    bundled CA certificates. This lets the CLI work behind corporate proxies or with
+    internal CAs whose root certificates are installed in the OS trust store but not in
+    the certifi bundle that requests uses by default."""
+    try:
+        import truststore
+    except ImportError:
+        console.print("[red]--system-truststore requires the 'truststore' package, which is not installed.[/red]")
+        raise typer.Exit(code=1)
+    truststore.inject_into_ssl()
+
+
 @app.callback()
 def common(
     ctx: typer.Context,
@@ -121,6 +134,13 @@ def common(
         callback=version_callback,
         is_eager=True,
     ),
+    system_truststore: bool = typer.Option(
+        False,
+        "--system-truststore",
+        help="Verify TLS using the operating system's certificate trust store "
+        "instead of the bundled CA certificates (e.g. behind a corporate proxy or internal CA).",
+        envvar="DATACONTRACT_SYSTEM_TRUSTSTORE",
+    ),
 ):
     """
     The datacontract CLI is an open source command-line tool for working with Data Contracts (https://datacontract.com).
@@ -134,6 +154,9 @@ def common(
     # Already-set environment variables take precedence.
     load_dotenv(dotenv_path=find_dotenv(usecwd=True), override=False)
 
+    if system_truststore:
+        inject_system_truststore()
+
 
 def enable_debug_logging(debug: bool, otherwise_disable_stderr: bool = False):
     if not debug and otherwise_disable_stderr:

{datacontract_cli-1.0.5 → datacontract_cli-1.0.7}/datacontract/engines/ibis/ibis_check_execute.py

@@ -849,16 +849,31 @@ def _resolve_col(columns: dict, field: str) -> str:
 
 
 def _table_database(con, server: Optional[Server]) -> Optional[str]:
-    """The schema (owner) to qualify the table with, or ``None``.
-
-    Oracle logs in as one user but the tables may be owned by a different schema
-    (``server.schema``). ibis defaults the owner to the login user during
-    introspection, so an unqualified lookup raises ``TableNotFound``. Other
-    backends already pin the schema at connect time, so only Oracle needs this.
+    """The schema to qualify the table with during introspection, or ``None``.
+
+    Two backends need the contract's ``server.schema`` passed explicitly instead
+    of relying on ibis's default:
+
+    - **Oracle** logs in as one user but the tables may be owned by a different
+      schema (``server.schema``). ibis defaults the owner to the login user, so
+      an unqualified lookup raises ``TableNotFound``.
+    - **Redshift** has no dedicated ibis backend and goes through the Postgres
+      backend (``con.name == "postgres"``). When no schema is passed, ibis's
+      Postgres introspection resolves the active schema with ``SELECT
+      current_schema`` (no parentheses) — valid on PostgreSQL but rejected by
+      Redshift with ``column "current_schema" does not exist``, since Redshift
+      only supports the parenthesized ``current_schema()``. Passing the schema
+      explicitly skips that query.
+
+    Other backends pin the schema at connect time and need no qualifier.
     """
-    if server is None:
+    if server is None or not server.schema_:
         return None
-    if getattr(con, "name", None) == "oracle" and server.schema_:
+    if getattr(con, "name", None) == "oracle":
+        return server.schema_
+    # Redshift rides the Postgres backend, so detect it by the contract's server
+    # type rather than con.name.
+    if get_server_type(server) == "redshift":
         return server.schema_
     return None
 

{datacontract_cli-1.0.5 → datacontract_cli-1.0.7/datacontract_cli.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: datacontract-cli
-Version: 1.0.5
+Version: 1.0.7
 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
@@ -16,6 +16,7 @@ Requires-Dist: click<9.0.0,>=8.1.0
 Requires-Dist: pydantic<2.14.0,>=2.8.2
 Requires-Dist: pyyaml~=6.0.1
 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: jsonschema<5.0.0,>=4.23.0
 Requires-Dist: pytz>=2024.1
@@ -40,7 +41,7 @@ Requires-Dist: openpyxl<4.0.0,>=3.1.5; extra == "excel"
 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<0.118.0; extra == "databricks"
+Requires-Dist: databricks-sdk<0.119.0; extra == "databricks"
 Requires-Dist: pyspark<5.0.0,>=3.5.0; extra == "databricks"
 Provides-Extra: iceberg
 Requires-Dist: pyiceberg==0.11.1; extra == "iceberg"
@@ -65,9 +66,9 @@ Requires-Dist: datacontract-cli[duckdb]; extra == "gcs"
 Requires-Dist: duckdb-extension-httpfs<1.6.0,>=1.5.0; extra == "gcs"
 Provides-Extra: azure
 Requires-Dist: datacontract-cli[duckdb]; extra == "azure"
-Requires-Dist: duckdb-extension-azure<1.1.0,>=1.0.3; (platform_system != "Linux" or platform_machine != "aarch64") and extra == "azure"
+Requires-Dist: duckdb-extension-azure<1.6.0,>=1.5.0; extra == "azure"
 Requires-Dist: azure-identity<1.18,>=1.17.1; extra == "azure"
-Requires-Dist: azure-storage-blob<12.30,>=12.24; extra == "azure"
+Requires-Dist: azure-storage-blob<12.31,>=12.24; extra == "azure"
 Provides-Extra: snowflake
 Requires-Dist: snowflake-connector-python[pandas]<4.7,>=3.6; extra == "snowflake"
 Requires-Dist: ibis-framework[snowflake]<13.0.0,>=10.0.0; extra == "snowflake"
@@ -91,7 +92,7 @@ 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<0.138.0,>=0.115.0; extra == "api"
+Requires-Dist: fastapi<0.139.0,>=0.115.0; extra == "api"
 Requires-Dist: uvicorn<0.50.0,>=0.44.0; extra == "api"
 Provides-Extra: protobuf
 Requires-Dist: proto-schema-parser<3.0,>=1.0; extra == "protobuf"
@@ -110,7 +111,7 @@ 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==0.15.17; extra == "dev"
+Requires-Dist: ruff==0.15.18; extra == "dev"
 Requires-Dist: testcontainers[kafka,minio,mssql,mysql,postgres]==4.14.2; extra == "dev"
 Requires-Dist: trino==0.337.0; extra == "dev"
 Dynamic: license-file

{datacontract_cli-1.0.5 → datacontract_cli-1.0.7}/datacontract_cli.egg-info/requires.txt

@@ -3,6 +3,7 @@ click<9.0.0,>=8.1.0
 pydantic<2.14.0,>=2.8.2
 pyyaml~=6.0.1
 requests<2.35,>=2.31
+truststore<1.0,>=0.10
 fastjsonschema<2.22.0,>=2.19.1
 jsonschema<5.0.0,>=4.23.0
 pytz>=2024.1
@@ -21,7 +22,7 @@ deepdiff<10.0.0,>=6.0.0
 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]
 
 [api]
-fastapi<0.138.0,>=0.115.0
+fastapi<0.139.0,>=0.115.0
 uvicorn<0.50.0,>=0.44.0
 
 [athena]
@@ -32,11 +33,9 @@ avro==1.12.1
 
 [azure]
 datacontract-cli[duckdb]
+duckdb-extension-azure<1.6.0,>=1.5.0
 azure-identity<1.18,>=1.17.1
-azure-storage-blob<12.30,>=12.24
-
-[azure:platform_system != "Linux" or platform_machine != "aarch64"]
-duckdb-extension-azure<1.1.0,>=1.0.3
+azure-storage-blob<12.31,>=12.24
 
 [bigquery]
 ibis-framework[bigquery]<13.0.0,>=10.0.0
@@ -47,7 +46,7 @@ pandas>=2.0.0
 [databricks]
 ibis-framework[databricks,pyspark]<13.0.0,>=10.0.0
 databricks-sql-connector<4.4.0,>=3.7.0
-databricks-sdk<0.118.0
+databricks-sdk<0.119.0
 pyspark<5.0.0,>=3.5.0
 
 [dbml]
@@ -66,7 +65,7 @@ pytest-xdist
 pymssql==2.3.13
 psycopg2-binary<3.0,>=2.9
 mysql-connector-python<9.7.0,>=8.0.30
-ruff==0.15.17
+ruff==0.15.18
 testcontainers[kafka,minio,mssql,mysql,postgres]==4.14.2
 trino==0.337.0
 

{datacontract_cli-1.0.5 → datacontract_cli-1.0.7}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "datacontract-cli"
-version = "1.0.5"
+version = "1.0.7"
 description = "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."
 license = "MIT"
 readme = "README.md"
@@ -20,6 +20,7 @@ dependencies = [
   "pydantic>=2.8.2,<2.14.0",
   "pyyaml~=6.0.1",
   "requests>=2.31,<2.35",
+  "truststore>=0.10,<1.0",
   "fastjsonschema>=2.19.1,<2.22.0",
   "jsonschema>=4.23.0,<5.0.0",
   "pytz>=2024.1",
@@ -57,7 +58,7 @@ excel = [
 databricks = [
   "ibis-framework[databricks,pyspark]>=10.0.0,<13.0.0",
   "databricks-sql-connector>=3.7.0,<4.4.0",
-  "databricks-sdk<0.118.0",
+  "databricks-sdk<0.119.0",
   # Not capped to Spark 3.5.x like older builds were: the databricks path uses a
   # caller-provided Spark session (or the databricks SQL connector) and never
   # loads the Kafka/Avro connector jars, so it isn't tied to a Scala/Spark line.
@@ -117,9 +118,9 @@ gcs = [
 
 azure = [
   "datacontract-cli[duckdb]",
-  'duckdb-extension-azure>=1.0.3,<1.1.0; platform_system != "Linux" or platform_machine != "aarch64"',
+  "duckdb-extension-azure>=1.5.0,<1.6.0",
   "azure-identity>=1.17.1,<1.18",
-  "azure-storage-blob>=12.24,<12.30",
+  "azure-storage-blob>=12.24,<12.31",
 ]
 
 snowflake = [
@@ -168,7 +169,7 @@ rdf = [
 ]
 
 api = [
-  "fastapi>=0.115.0,<0.138.0",
+  "fastapi>=0.115.0,<0.139.0",
   "uvicorn>=0.44.0,<0.50.0",
 ]
 
@@ -198,7 +199,7 @@ dev = [
   # psycopg v3 and DuckDB's mysql extension respectively).
   "psycopg2-binary>=2.9,<3.0",
   "mysql-connector-python>=8.0.30,<9.7.0",
-  "ruff==0.15.17",
+  "ruff==0.15.18",
   "testcontainers[minio,postgres,kafka,mssql,mysql]==4.14.2",
   "trino==0.337.0",
 ]

{datacontract_cli-1.0.5 → datacontract_cli-1.0.7}/tests/test_cli.py

@@ -73,6 +73,39 @@ def test_env_file_does_not_override_existing_environment_variables(tmp_path, mon
     assert os.environ.get("DATACONTRACT_TEST_DOTENV_VAR") == "from-environment"
 
 
+def test_system_truststore_option_in_help():
+    result = runner.invoke(app, ["--help"], env={"COLUMNS": "200"})
+    assert result.exit_code == 0
+    plain_output = re.sub(r"\x1b\[[0-9;]*m", "", result.stdout)
+    assert "--system-truststore" in plain_output
+
+
+def test_system_truststore_flag_injects(monkeypatch):
+    calls = []
+    monkeypatch.setattr("datacontract.cli.inject_system_truststore", lambda: calls.append(True))
+    result = runner.invoke(app, ["--system-truststore", "lint", "unknown.yaml"])
+    assert result.exit_code == 1  # file does not exist, but the app callback has run
+    assert calls == [True]
+
+
+def test_system_truststore_env_var_injects(monkeypatch):
+    calls = []
+    monkeypatch.setattr("datacontract.cli.inject_system_truststore", lambda: calls.append(True))
+    monkeypatch.setenv("DATACONTRACT_SYSTEM_TRUSTSTORE", "1")
+    result = runner.invoke(app, ["lint", "unknown.yaml"])
+    assert result.exit_code == 1
+    assert calls == [True]
+
+
+def test_system_truststore_not_injected_by_default(monkeypatch):
+    calls = []
+    monkeypatch.setattr("datacontract.cli.inject_system_truststore", lambda: calls.append(True))
+    monkeypatch.delenv("DATACONTRACT_SYSTEM_TRUSTSTORE", raising=False)
+    result = runner.invoke(app, ["lint", "unknown.yaml"])
+    assert result.exit_code == 1
+    assert calls == []
+
+
 def test_changelog_help():
     result = runner.invoke(app, ["changelog", "--help"])
     assert result.exit_code == 0

{datacontract_cli-1.0.5 → datacontract_cli-1.0.7}/tests/test_ibis_oracle_schema.py

@@ -51,6 +51,22 @@ def test_table_database_none_for_non_oracle_backend():
     assert _table_database(con, server) is None
 
 
+def test_table_database_uses_server_schema_for_redshift():
+    # Redshift rides the Postgres ibis backend (con.name == "postgres"). Without
+    # an explicit schema, ibis introspects via `SELECT current_schema` (no
+    # parens), which Redshift rejects ("column current_schema does not exist").
+    # Passing the schema explicitly avoids that query.
+    server = Server(type="redshift", schema="dwh_academic")
+    con = _FakeBackend("postgres", {})
+    assert _table_database(con, server) == "dwh_academic"
+
+
+def test_table_database_none_for_redshift_without_schema():
+    server = Server(type="redshift")
+    con = _FakeBackend("postgres", {})
+    assert _table_database(con, server) is None
+
+
 def test_table_database_none_without_server():
     con = _FakeBackend("oracle", {})
     assert _table_database(con, None) is None