From b6161eb20eb62817eab8756ea4b39c4e1c3d18ec Mon Sep 17 00:00:00 2001
From: "Andreas Koepf (aider)" <andreas.koepf@aleph-alpha.com>
Date: Thu, 23 Jan 2025 20:26:12 +0100
Subject: [PATCH] test: Add comprehensive unit tests for WordReversalDataset

---
 tests/test_word_reversal.py | 84 +++++++++++++++++++++++++++++++++++++
 1 file changed, 84 insertions(+)
 create mode 100644 tests/test_word_reversal.py

diff --git a/tests/test_word_reversal.py b/tests/test_word_reversal.py
new file mode 100644
index 00000000..69b27f08
--- /dev/null
+++ b/tests/test_word_reversal.py
@@ -0,0 +1,84 @@
+"""Tests for word reversal task generation"""
+import pytest
+
+from reasoning_gym.algorithmic.word_reversal import (
+    WordReversalConfig,
+    WordReversalDataset,
+)
+
+
+def test_word_reversal_config_validation():
+    """Test that invalid configs raise appropriate errors"""
+    with pytest.raises(AssertionError):
+        config = WordReversalConfig(min_words=0)
+        config.validate()
+
+    with pytest.raises(AssertionError):
+        config = WordReversalConfig(min_words=10, max_words=5)
+        config.validate()
+
+
+def test_word_reversal_dataset_deterministic():
+    """Test that dataset generates same items with same seed"""
+    config = WordReversalConfig(seed=42, size=10)
+    dataset1 = WordReversalDataset(config)
+    dataset2 = WordReversalDataset(config)
+
+    for i in range(len(dataset1)):
+        assert dataset1[i] == dataset2[i]
+
+
+def test_word_reversal_dataset_items():
+    """Test basic properties of generated items"""
+    config = WordReversalConfig(
+        min_words=3,
+        max_words=6,
+        size=10,
+        seed=42
+    )
+    dataset = WordReversalDataset(config)
+
+    for i in range(len(dataset)):
+        item = dataset[i]
+        # Check item structure
+        assert isinstance(item, dict)
+        assert "question" in item
+        assert "answer" in item
+        assert "metadata" in item
+        
+        # Check metadata
+        assert "num_words" in item["metadata"]
+        assert "words" in item["metadata"]
+        
+        # Verify word count constraints
+        words = item["metadata"]["words"]
+        assert len(words) >= config.min_words
+        assert len(words) <= config.max_words
+        
+        # Verify reversal is correct
+        question_words = [w.strip() for w in item["question"].split(":")[1].strip().split(",")]
+        answer_words = item["answer"].split(", ")
+        assert answer_words == list(reversed(question_words))
+
+
+def test_word_reversal_dataset_iteration():
+    """Test that iteration respects dataset size"""
+    config = WordReversalConfig(size=5, seed=42)
+    dataset = WordReversalDataset(config)
+
+    items = list(dataset)
+    assert len(items) == config.size
+
+    # Test multiple iterations yield same items
+    assert items == list(dataset)
+
+
+def test_word_reversal_text_preprocessing():
+    """Test that text preprocessing handles edge cases"""
+    config = WordReversalConfig(size=1, seed=42)
+    dataset = WordReversalDataset(config)
+    
+    # Verify words were extracted from text
+    assert len(dataset.words) > 0
+    # Verify words contain only alphanumeric characters
+    assert all(word.isalnum() for word in dataset.words)