Skip to content

README.md

Latest commit

 

History

History
56 lines (43 loc) · 2.28 KB

README.md

File metadata and controls

56 lines (43 loc) · 2.28 KB
Raw

Tests

The test suite is Pytest only. Run it from the project root:

# via Docker
docker compose run --rm test tests/python                        # everything
docker compose run --rm test tests/python/test_device_grouping2.py  # one file

# without Docker
uv sync
uv run pytest tests/python
uv run pytest tests/python/test_device_grouping2.py

Structure

  • python/: parser, semantic-map, normalization, and device-grouping tests.
  • python/conftest.py: shared fixtures, plus the TestFileLoader that drives the real-data tests below.
  • python/test_config.yaml: maps parser formats to real export files on your machine (see below).
  • zip_data/, tmp_outputs/: git-ignored scratch directories for real archives and debug dumps.

Most tests are self-contained and use inline fixtures. The parser tests under python/extractors/ are the exception: they are parameterized over test_config.yaml and need real data.

Running the parser tests against real exports

test_config.yaml has two sections. platforms: gives a directory per platform, and test_files: lists individual files within those directories grouped by parser format, each with the record count and raw keys the parser should produce:

platforms:
  facebook:
    test_data_dir: "~/takeout-test-data/facebook"

test_files:
  json:
    - path: "facebook/security_and_login_information/account_activity.json"
      platform: "facebook"
      expected:
        num_records: 59
        raw_keys: ['action', 'timestamp', 'ip_address']

No export data ships with this repository. The paths above are placeholders, so point test_data_dir at your own unpacked exports. Any file that isn't found is skipped with a Warning: test data not found at ... line rather than failing, so a clean checkout reports these tests as passing while actually running none of them. Check the warnings if you expect real coverage.

Supported expected: keys are num_records (exact), min_records (lower bound), and raw_keys (fields that must appear in the first parsed record). A file with no expected: block only asserts that at least one record came out.

If you want data to point this at, the two synthetic exports in evaluation/efficiency/data/ unpack into the right shape. See the Sample Data section of the root README.md.