Fix false-positive in EnrichmentPipelineTest by introducing mock Config with DummyModule

[davide954]

Problem

The existing EnrichmentPipelineTest used new Config() directly,
which returns an empty module list in the test environment.

As a result, the enrichment verification loop never executed, making
the test pass silently regardless of whether the pipeline logic was
correct — a false positive flagged by @tejas-2232.

Solution

Following the approach agreed with @jnioche discussed in #142 (mock Config
rather than expanding the schema with all production columns):

• DummyModule: a minimal EnrichmentModule that declares a
  single dummy_enriched column and sets it to 42.0 in process().
• createMockConfig(): builds a Config pre-loaded with the
  DummyModule and injects the matching config entry via reflection,
  following the same pattern used in ConfigTest.
• assertFalse(modules.isEmpty()): guard assertion that fails
  fast if modules are not loaded, preventing silent false positives.
• Dynamic indices: all field lookups use schema.fieldIndex(...)
  instead of hardcoded positions, making the test robust against
  schema changes.

What this does NOT change

• No changes to production code.
• No new test dependencies (no Mockito).
• The test scope remains focused on the Usage-filter logic of
  EnrichmentPipeline (Usage → enriched, Tax/Fee → skipped).

Verification

The parameterized test now actually exercises the enrichment path:

• Usage → dummy_enriched is populated (42.0)
• Tax → dummy_enriched remains null
• Fee → dummy_enriched remains null

[davide954]

@jnioche I watch out your PR #149 and probably this tests will not succeed correctly with your new implementations, so I won't do anything till you merge that PR, I understand the sensitivity of this code change. I'll keep you posted.

[jnioche]

Hi @davide954, am away this week. Will comment and review when I'm back

[jnioche]

@jnioche I watch out your PR #149 and probably this tests will not succeed correctly with your new implementations, so I won't do anything till you merge that PR, I understand the sensitivity of this code change. I'll keep you posted.

I merged that PR. It has no impact on your PR

[jnioche]

thanks @davide954

[jnioche]

@jnioche I watch out your PR #149 and probably this tests will not succeed correctly with your new implementations, so I won't do anything till you merge that PR, I understand the sensitivity of this code change. I'll keep you posted.

I merged that PR. It has no impact on your PR

actually it does, sorry! Just fixed it

[davide954]

ok now I see it works, thanks @jnioche.