From 97db01770885fe9f29f7f8e815d9ac2a7c2d90cc Mon Sep 17 00:00:00 2001
From: alekszievr <44192193+alekszievr@users.noreply.github.com>
Date: Thu, 20 Feb 2025 15:16:58 +0100
Subject: [PATCH] Test: test corpus builder [cog-1234] (#564)

<!-- .github/pull_request_template.md -->

## Description
<!-- Provide a clear description of the changes in this PR -->

## DCO Affirmation
I affirm that all code in every commit of this pull request conforms to
the terms of the Topoteretes Developer Certificate of Origin


<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit

- **Chores**
- Enhanced the continuous integration workflows with updated dependency
management and environment configurations for improved test stability.

- **Tests**
- Added parameterized unit tests to verify corpus loading and structure,
ensuring more robust handling of test data.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
---
 .../eval_framework/corpus_builder_test.py     | 29 +++++++++++++++++++
 1 file changed, 29 insertions(+)
 create mode 100644 cognee/tests/unit/eval_framework/corpus_builder_test.py

diff --git a/cognee/tests/unit/eval_framework/corpus_builder_test.py b/cognee/tests/unit/eval_framework/corpus_builder_test.py
new file mode 100644
index 00000000..cf328a30
--- /dev/null
+++ b/cognee/tests/unit/eval_framework/corpus_builder_test.py
@@ -0,0 +1,29 @@
+import pytest
+from evals.eval_framework.corpus_builder.corpus_builder_executor import CorpusBuilderExecutor
+from cognee.infrastructure.databases.graph import get_graph_engine
+from unittest.mock import AsyncMock, patch
+
+benchmark_options = ["HotPotQA", "Dummy", "TwoWikiMultiHop"]
+
+
+@pytest.mark.parametrize("benchmark", benchmark_options)
+def test_corpus_builder_load_corpus(benchmark):
+    limit = 2
+    corpus_builder = CorpusBuilderExecutor(benchmark, "Default")
+    raw_corpus, questions = corpus_builder.load_corpus(limit=limit)
+    assert len(raw_corpus) > 0, f"Corpus builder loads empty corpus for {benchmark}"
+    assert len(questions) <= 2, (
+        f"Corpus builder loads {len(questions)} for {benchmark} when limit is {limit}"
+    )
+
+
+@pytest.mark.asyncio
+@pytest.mark.parametrize("benchmark", benchmark_options)
+@patch.object(CorpusBuilderExecutor, "run_cognee", new_callable=AsyncMock)
+async def test_corpus_builder_build_corpus(mock_run_cognee, benchmark):
+    limit = 2
+    corpus_builder = CorpusBuilderExecutor(benchmark, "Default")
+    questions = await corpus_builder.build_corpus(limit=limit)
+    assert len(questions) <= 2, (
+        f"Corpus builder loads {len(questions)} for {benchmark} when limit is {limit}"
+    )