Skip to main content

Documentation Index

Fetch the complete documentation index at: https://mintlify.com/obedc295/proyect_dw/llms.txt

Use this file to discover all available pages before exploring further.

ETL Dinámico ships with four pytest test modules that together cover every layer of the pipeline — from live database connectivity to full end-to-end orchestration. Two of the four tests are completely database-free (pure unit tests with mocks), one verifies incremental-load filtering logic using patched I/O, and one confirms that both the OLTP and OLAP SQL Server connections are alive. Running the full suite takes only a few seconds and gives you immediate confidence that every component is working correctly.

​
Prerequisites

Before running the test suite, make sure the following are in place:

Virtual Environment

Your project virtual environment must be activated so that all dependencies resolve from the correct location.

.env Configured

A valid .env file with OLTP_* and OLAP_* connection variables must exist in the project root for connection tests to pass.

pytest ≥ 9.1.1

pytest>=9.1.1 must be installed. It is listed in requirements.txt and is installed automatically in a fresh virtual environment.
Verify your pytest installation before running: 
pytest --version

​
Running the Tests

1

Run all tests

Execute every test module in the tests/ directory with verbose output:
pytest tests/ -v
This runs all four test functions across all four files in dependency-safe order.
2

Run a specific test file

Target a single module by passing its path directly:
pytest tests/test_transformer.py -v
Replace test_transformer.py with any other module name to isolate a different layer.
3

Run a single test function

Use the ::test_name selector to run exactly one test case:
pytest tests/test_transformer.py::test_transformaciones_pandas -v
This is useful during development when you want fast feedback on a specific assertion.

​
Test Modules at a Glance

test_connection.py

test_conexiones() — Requires a live SQL Server connection. Creates a DatabaseClient instance and executes SELECT 1 against both the OLTP and OLAP engines, asserting that each returns 1.

test_transformer.py

test_transformaciones_pandas() — Pure unit test; no database required. Exercises capitalize_transform (uppercase), concat_transform (column concatenation), and date_transform (year extraction) on a small in-memory DataFrame.

test_loader.py

test_carga_incremental_filtra_correctamente() — Mocks the database client and patches pd.read_sql to simulate an existing DW row (CustomerID=10). Asserts that to_sql is called exactly once, confirming the incremental load path was executed.

test_pipelines.py

test_flujo_completo_con_datos_reales() — Mocks the extractor and loader, then drives a full ETLPipeline.run_dynamic_etl() call with three territory rows and an upper column mapping. Verifies status, rows_extracted, and that the New_Name column contains correctly uppercased values.

​
Connection Test Warning

test_connection.py requires a real, running SQL Server instance. If your .env file is missing, contains incorrect credentials, or your SQL Server is unreachable, test_conexiones will fail with an assertion error. The remaining three tests are entirely mock-based and will pass regardless of database availability.

​
Running Without a Database

If you are working in an environment without SQL Server access, skip the connection test and run only the mock-based suite:
pytest tests/test_transformer.py tests/test_loader.py tests/test_pipelines.py -v
All three of these tests use MagicMock and patch to replace real database calls, so they pass in any environment where Python dependencies are installed.

​
Expected Output

When all four tests pass, pytest prints a report similar to the following: 
tests/test_connection.py::test_conexiones PASSED
tests/test_transformer.py::test_transformaciones_pandas PASSED
tests/test_loader.py::test_carga_incremental_filtra_correctamente PASSED
tests/test_pipelines.py::test_flujo_completo_con_datos_reales PASSED

4 passed in X.XXs
Each PASSED line confirms that the corresponding layer of the ETL stack is behaving correctly. If any line shows FAILED or ERROR, check the traceback printed below the summary — the most common cause is a misconfigured .env (connection tests) or a breaking change in a service class (unit tests).

Build docs developers (and LLMs) love

Get started for freeTalk to us