From 6f269c718e45c46a032135b2f2cee71c207efa52 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?J=C4=99drzej=20Jamnicki?= <jedrzej.jamnicki@gmail.com>
Date: Wed, 15 May 2024 17:59:54 +0000
Subject: [PATCH] test: add unittests for SyntheticQAPairs model

---
 juddges/data/models.py               |  3 ++
 tests/test_synthetic_qa_pairs_dto.py | 79 ++++++++++++++++++++++++++++
 2 files changed, 82 insertions(+)
 create mode 100644 tests/test_synthetic_qa_pairs_dto.py

diff --git a/juddges/data/models.py b/juddges/data/models.py
index 5e84b1e..92345dd 100644
--- a/juddges/data/models.py
+++ b/juddges/data/models.py
@@ -32,6 +32,9 @@ def __len__(self) -> int:
     def test_empty(self) -> None:
         if not any(self.questions):
             raise GeneratedQAPairsEmptyError("At least one question should be generated")
+        else:
+            if not any(self.answers):
+                raise GeneratedQAPairsEmptyError("At least one answer should be generated")
 
     def test_equal_length(self) -> None:
         assertion_msg = "Number of questions and answers should be equal"
diff --git a/tests/test_synthetic_qa_pairs_dto.py b/tests/test_synthetic_qa_pairs_dto.py
new file mode 100644
index 0000000..fd32263
--- /dev/null
+++ b/tests/test_synthetic_qa_pairs_dto.py
@@ -0,0 +1,79 @@
+import unittest
+
+from juddges.data.models import LangCode, SyntheticQAPairs
+from juddges.exception import (
+    GeneratedQAPairsEmptyError,
+    GeneratedQAPairsLenghtMismatchError,
+    GeneratedQAPairsNotUniqueError,
+    LanguageMismatchError,
+)
+
+
+class TestSyntheticQAPairs(unittest.TestCase):
+    def test_len(self):
+        scenarios = [
+            {"q": [], "a": [], "expected": 0},
+            {"q": ["1"], "a": [], "expected": 1},
+            {"q": ["1"], "a": ["2"], "expected": 1},
+            {"q": ["1", "2"], "a": ["1", "2"], "expected": 2},
+            {"q": ["1"], "a": ["1", "2"], "expected": 1},
+        ]
+        for scenario in scenarios:
+            dto = SyntheticQAPairs(questions=scenario["q"], answers=scenario["a"])
+            self.assertEqual(len(dto), scenario["expected"])
+
+    def test_raises_on_empty(self):
+        scenarios = [
+            {"q": [], "a": []},
+            {"q": [], "a": [""]},
+            {"q": [""], "a": []},
+        ]
+        for scenario in scenarios:
+            dto = SyntheticQAPairs(questions=scenario["q"], answers=scenario["a"])
+            with self.assertRaises(GeneratedQAPairsEmptyError):
+                dto.test_empty()
+
+    def test_raises_on_length_mismatch(self):
+        scenarios = [
+            {"q": ["1"], "a": []},
+            {"q": [], "a": ["1"]},
+            {"q": ["1"], "a": ["1", "2"]},
+            {"q": ["1", "2"], "a": ["1"]},
+        ]
+        for scenario in scenarios:
+            dto = SyntheticQAPairs(questions=scenario["q"], answers=scenario["a"])
+            with self.assertRaises(GeneratedQAPairsLenghtMismatchError):
+                dto.test_equal_length()
+
+    def test_raises_on_not_unique(self):
+        dto = SyntheticQAPairs(questions=["1", "1"], answers=["1", "2"])
+        with self.assertRaises(GeneratedQAPairsNotUniqueError):
+            dto.test_unique_questions()
+
+    def test_raises_on_language_mismatch(self):
+        scenarios = [
+            {
+                "q": ["Jakie podstawowe prawo stosuje sąd w tej sprawie?"],
+                "a": ["The court applies the Code of Civil Procedure."],
+                "expected_lang": LangCode.POLISH,
+            },
+            {
+                "q": ["What basic law does the court apply in this case?"],
+                "a": ["Sąd stosuje Kodeks postępowania cywilnego."],
+                "expected_lang": LangCode.POLISH,
+            },
+            {
+                "q": ["What is the name of the accused?"],
+                "a": ["Jędrzej Brzęczyszczykiewicz"],
+                "expected_lang": LangCode.ENGLISH,
+            },
+            {
+                "q": ["Jakie jest imię i nazwisko oskarżonego?"],
+                "a": ["Helga Müller-Fäßler"],
+                "expected_lang": LangCode.POLISH,
+            },
+        ]
+        for scenario in scenarios:
+            dto = SyntheticQAPairs(questions=scenario["q"], answers=scenario["a"])
+            with self.assertRaises(LanguageMismatchError):
+                dto.test_language(lang=scenario["expected_lang"])