Skip to content

Setup and capabilities

langadapter.setup

Checks of a project's setup, reported as typed results, that build and install nothing: the common admission of a request's project and inputs, then the registered adapter's own native checks.

langadapter.capabilities

What an adapter can do for its project context, and the descriptive facts about its language.

A capability report probes the project's environment (its interpreter, local or in the container), not the calling process: each operation is available, unsupported by the language or engine, or missing setup, with what is missing and where it was looked for. Language facts are data, never instructions. VERIFY, RUN and RETRY are the plain next steps setup reports attach where a person can act on an unconfirmed item (none for unimplemented operations).

VERIFY module-attribute

VERIFY = 'run the tool check to try it'

The next step of an item only the tool check tries.

RUN module-attribute

RUN = 'start a run, which tries it for the first time'

The next step of an item only a real run tries.

RETRY module-attribute

RETRY = 'run the check again with a longer time limit'

The next step of an item whose check reached its time limit.

TRIED_BY module-attribute

TRIED_BY = ('verification', 'run')

What first tries an item or operation: the setup verification, or only a real run.

SOURCE_READS module-attribute

SOURCE_READS = ('check_structure', 'test_metrics', 'metrics', 'skeleton', 'same_skeleton', 'function_fingerprints', 'imports', 'import_path_candidates', 'duplicate_test_names', 'merge_collisions', 'is_test_file', 'validate_source', 'target_identity', 'resolve_collisions', 'analyze_validity', 'screen_source', 'normalize_bindings', 'stage_pool', 'import_statement', 'code_smells', 'project_source_files', 'call_statistics')

The source operations that read a source and change nothing (adapter.sources).

SOURCE_EDITS module-attribute

SOURCE_EDITS = ('prune_tests', 'merge_sources', 'normalize_shadowed', 'salvage_syntax', 'prune_unresolvable_imports', 'prune_collection_breakers', 'strip_module_level_skips', 'finalize_source')

The source operations that return an edited source (adapter.sources).

TEST_OPERATIONS module-attribute

TEST_OPERATIONS = ('run_suite', 'collect_suite', 'run_as_delivered', 'benchmark', 'benchmark_warmed', 'per_test_evidence', 'profile_suite')

The operations that run a suite (adapter.tests).

MUTATION_OPERATIONS module-attribute

MUTATION_OPERATIONS = ('engines', 'operators', 'open_session', 'run', 'per_suite')

The mutation operations (adapter.mutations).

OPERATION_NAMES module-attribute

OPERATION_NAMES = (*(f'sources.{name}' for name in SOURCE_READS + SOURCE_EDITS), *(f'tests.{name}' for name in TEST_OPERATIONS), *(f'mutations.{name}' for name in MUTATION_OPERATIONS))

Every operation's name as an operation report and a setup request name it, such as tests.run_suite: the four groups above with their prefixes.

OPERATION_FACETS module-attribute

OPERATION_FACETS = MappingProxyType({'tests.run_suite': ('coverage',)})

Optional measurements a caller may require of an operation: a suite run's coverage of its coverage_target.

CAPABILITIES module-attribute

CAPABILITIES = ('suite_runs', 'coverage', 'per_test_evidence', 'per_test_kill_attribution', 'mutation', 'source_merging', 'generated_suite_operations')

The capabilities a capability report covers: running a suite (suite_runs), measuring its coverage (coverage), per-test evidence (per_test_evidence) and the kills attributed to each test (per_test_kill_attribution), mutation runs (mutation), merging test sources (source_merging), and the operations that repair and screen generated tests (generated_suite_operations): salvage_syntax, prune_unresolvable_imports, prune_collection_breakers, strip_module_level_skips, analyze_validity and screen_source.