Merge pull request #1335 from bact/more-tests

Add more tests for tone and vowel analysis

Author: bact

tests/compact/testc_util.py

@@ -22,10 +22,17 @@ def test_spell_word(self):
         self.assertEqual(
             spell_word("คนดี"), ["คอ", "นอ", "คน", "ดอ", "อี", "ดี", "คนดี"]
         )
+        result = spell_word("คน")
+        self.assertIsInstance(result, list)
+        self.assertIn("คน", result)
+
         # Edge cases: None and empty string
         self.assertEqual(spell_word(None), [])
         self.assertEqual(spell_word(""), [])
 
+        # multi-syllable: last element is the full word
+        result_multi = spell_word("คนดี")
+        self.assertEqual(result_multi[-1], "คนดี")
 
 class UtilTestCaseC(unittest.TestCase):
     def test_rhyme(self):
@@ -37,6 +44,14 @@ def test_thai_word_tone_detector(self):
         self.assertEqual(
             thai_word_tone_detector("ราคา"), [("รา", "m"), ("คา", "m")]
         )
+        result = thai_word_tone_detector("คนดี")
+        self.assertIsInstance(result, list)
+        valid_tones = {"l", "m", "h", "r", "f", ""}
+        for syllable, tone in result:
+            self.assertIsInstance(syllable, str)
+            self.assertIn(tone, valid_tones)
+        self.assertIsInstance(thai_word_tone_detector("มือถือ"), list)
+
         # Edge cases: None and empty string
         self.assertEqual(thai_word_tone_detector(None), [])
         self.assertEqual(thai_word_tone_detector(""), [])
@@ -63,3 +78,5 @@ def test_check_khuap_klam(self):
 
         # Edge cases: empty string returns None
         self.assertIsNone(check_khuap_klam(""))
+        for word in ["กลม", "จริง", "ตา"]:
+            self.assertIn(check_khuap_klam(word), (True, False, None))

tests/core/test_khavee.py

@@ -117,3 +117,144 @@ def test_check_aek_too(self):
         self.assertTrue(
             kv.check_aek_too(["หนม", "หน่ม", "หน้ม"]), [False, "aek", "too"]
         )
+
+
+class KhaveeCheckKaruLahuTestCase(unittest.TestCase):
+    """Tests for KhaveeVerifier.check_karu_lahu"""
+
+    def setUp(self):
+        self.kv = KhaveeVerifier()
+
+    def test_dead_syllable_is_karu(self):
+        self.assertEqual(self.kv.check_karu_lahu("กด"), "karu")
+
+    def test_long_live_syllable_is_karu(self):
+        self.assertEqual(self.kv.check_karu_lahu("กา"), "karu")
+
+    def test_live_syllable_with_final_consonant_is_karu(self):
+        self.assertEqual(self.kv.check_karu_lahu("กาน"), "karu")
+
+    def test_bo_mai_ek_is_lahu(self):
+        self.assertEqual(self.kv.check_karu_lahu("บ่"), "lahu")
+
+    def test_no_short_word_is_lahu(self):
+        self.assertEqual(self.kv.check_karu_lahu("ณ"), "lahu")
+
+    def test_tho_short_word_is_lahu(self):
+        self.assertEqual(self.kv.check_karu_lahu("ธ"), "lahu")
+
+    def test_ko_mai_is_lahu(self):
+        self.assertEqual(self.kv.check_karu_lahu("ก็"), "lahu")
+
+
+class KhaveeHandleKarunTestCase(unittest.TestCase):
+    """Tests for KhaveeVerifier.handle_karun_sound_silence"""
+
+    def setUp(self):
+        self.kv = KhaveeVerifier()
+
+    def test_word_without_karun_unchanged(self):
+        self.assertEqual(self.kv.handle_karun_sound_silence("คน"), "คน")
+        self.assertEqual(self.kv.handle_karun_sound_silence("กา"), "กา")
+
+    def test_word_ending_with_karun_stripped(self):
+        # เกมส์ → drop ์ and the consonant before it (ส) → เกม
+        self.assertEqual(self.kv.handle_karun_sound_silence("เกมส์"), "เกม")
+
+    def test_word_ending_with_karun_stripped_2(self):
+        # รักษ์ → drop ์ + ษ → รัก
+        self.assertEqual(self.kv.handle_karun_sound_silence("รักษ์"), "รัก")
+
+    def test_returns_string(self):
+        self.assertIsInstance(self.kv.handle_karun_sound_silence("สวัสดี"), str)
+
+
+class KhaveeCheckAekTooEdgeCasesTestCase(unittest.TestCase):
+    """Edge-case tests for KhaveeVerifier.check_aek_too"""
+
+    def setUp(self):
+        self.kv = KhaveeVerifier()
+
+    def test_non_string_raises_type_error(self):
+        with self.assertRaises(TypeError):
+            self.kv.check_aek_too(123)
+
+    def test_dead_syllable_as_aek_flag(self):
+        self.assertEqual(self.kv.check_aek_too("บท", dead_syllable_as_aek=True), "aek")
+
+    def test_dead_syllable_without_flag_returns_false(self):
+        self.assertFalse(self.kv.check_aek_too("บท", dead_syllable_as_aek=False))
+
+    def test_list_with_non_string_element_raises(self):
+        with self.assertRaises(TypeError):
+            self.kv.check_aek_too(["ไก่", 42])
+
+    def test_both_tone_marks_returns_false(self):
+        # word with both ่ and ้ should return False
+        self.assertFalse(self.kv.check_aek_too("ก่้"))
+
+
+class KhaveeCheckKlonExtendedTestCase(unittest.TestCase):
+    """Tests for check_klon k_type=8 and invalid k_type"""
+
+    def setUp(self):
+        self.kv = KhaveeVerifier()
+
+    def test_invalid_k_type_returns_error_string(self):
+        result = self.kv.check_klon("บทกวีทดสอบ", k_type=99)
+        self.assertIsInstance(result, str)
+        self.assertIn("Something went wrong", result)
+
+    def test_incomplete_klon4_poem(self):
+        result = self.kv.check_klon("ฉันชื่อหมูกรอบ", k_type=4)
+        self.assertIsInstance(result, str)
+        self.assertIn("does not have 4 complete sentences", result)
+
+    def test_incomplete_klon8_poem(self):
+        result = self.kv.check_klon("ฉันชื่อหมูกรอบ", k_type=8)
+        self.assertIsInstance(result, str)
+
+    def test_check_klon8_correct_poem(self):
+        poem = (
+            "ฉันชื่อหมูกรอบ ฉันชอบกินไก่ แล้วก็วิ่งไล่ หมาชื่อนํ้าทอง "
+            "ลคคนเก่ง เอ๋งเอ๋งคะนอง มีคนจับจอง เขาชื่อน้องเธียร"
+        )
+        self.assertIsNotNone(self.kv.check_klon(poem, k_type=8))
+
+
+class KhaveeCheckSaraEdgeCasesTestCase(unittest.TestCase):
+    """Edge-case tests for KhaveeVerifier.check_sara"""
+
+    def setUp(self):
+        self.kv = KhaveeVerifier()
+
+    def test_bo_mai_ek_returns_oo(self):
+        self.assertEqual(self.kv.check_sara("บ่"), "ออ")
+
+    def test_special_word_เออ(self):
+        self.assertEqual(self.kv.check_sara("เออ"), "เออ")
+
+    def test_special_word_เอ(self):
+        self.assertEqual(self.kv.check_sara("เอ"), "เอ")
+
+    def test_special_word_เอะ(self):
+        self.assertEqual(self.kv.check_sara("เอะ"), "เอะ")
+
+    def test_special_word_เอา(self):
+        self.assertEqual(self.kv.check_sara("เอา"), "เอา")
+
+    def test_special_word_เอาะ(self):
+        self.assertEqual(self.kv.check_sara("เอาะ"), "เอาะ")
+
+    def test_ru_sara(self):
+        self.assertEqual(self.kv.check_sara("ฤ"), "อึ")
+
+    def test_ruea_sara(self):
+        # ฤา (ฤ + sara aa U+0E32) → อือ; note: ฤๅ uses lakkhangyao, not sara aa
+        self.assertEqual(self.kv.check_sara("ฤา"), "อือ")
+
+    def test_เอือ_sara(self):
+        self.assertEqual(self.kv.check_sara("เรือ"), "เอือ")
+
+    def test_returns_string(self):
+        self.assertIsInstance(self.kv.check_sara("เริง"), str)

tests/core/test_tag.py

@@ -228,6 +228,82 @@ def test_ner_locations(self):
         )
 
 
+class BlackboardPreProcessTestCase(unittest.TestCase):
+    """Tests for pythainlp.tag.blackboard.pre_process"""
+
+    def setUp(self):
+        from pythainlp.tag.blackboard import pre_process
+
+        self.pre_process = pre_process
+
+    def test_space_is_escaped(self):
+        self.assertEqual(self.pre_process([" "]), ["_"])
+
+    def test_regular_words_unchanged(self):
+        self.assertEqual(self.pre_process(["ผม", "รัก", "คุณ"]), ["ผม", "รัก", "คุณ"])
+
+    def test_mixed_space_and_words(self):
+        self.assertEqual(self.pre_process(["ผม", " ", "คุณ"]), ["ผม", "_", "คุณ"])
+
+    def test_multiple_spaces(self):
+        self.assertEqual(self.pre_process([" ", " "]), ["_", "_"])
+
+    def test_empty_list(self):
+        self.assertEqual(self.pre_process([]), [])
+
+    def test_non_space_special_chars_unchanged(self):
+        self.assertEqual(self.pre_process(["_"]), ["_"])  # already escaped form
+
+    def test_single_thai_word(self):
+        self.assertEqual(self.pre_process(["สวัสดี"]), ["สวัสดี"])
+
+
+class BlackboardPostProcessTestCase(unittest.TestCase):
+    """Tests for pythainlp.tag.blackboard.post_process"""
+
+    def setUp(self):
+        from pythainlp.tag.blackboard import TO_UD, post_process
+
+        self.post_process = post_process
+        self.TO_UD = TO_UD
+
+    def test_unescape_underscore_to_space(self):
+        result = self.post_process([("_", "NN"), ("คน", "NN")])
+        self.assertEqual(result, [(" ", "NN"), ("คน", "NN")])
+
+    def test_regular_words_pass_through(self):
+        result = self.post_process([("ผม", "PPRS"), ("รัก", "VACT")])
+        self.assertEqual(result, [("ผม", "PPRS"), ("รัก", "VACT")])
+
+    def test_unescape_to_space_with_ud(self):
+        result = self.post_process([("_", "NN")], to_ud=True)
+        self.assertEqual(result, [(" ", "NOUN")])
+
+    def test_regular_word_to_ud(self):
+        result = self.post_process([("คน", "NN")], to_ud=True)
+        self.assertEqual(result, [("คน", "NOUN")])
+
+    def test_all_to_ud_mappings(self):
+        for bb_tag, ud_tag in self.TO_UD.items():
+            if bb_tag == "":
+                continue  # skip empty-string sentinel
+            result = self.post_process([("word", bb_tag)], to_ud=True)
+            self.assertEqual(
+                result[0][1],
+                ud_tag,
+                f"TO_UD mapping failed for tag '{bb_tag}': "
+                f"expected '{ud_tag}', got '{result[0][1]}'",
+            )
+
+    def test_empty_list(self):
+        self.assertEqual(self.post_process([]), [])
+        self.assertEqual(self.post_process([], to_ud=True), [])
+
+    def test_no_ud_default_preserves_tag(self):
+        result = self.post_process([("คน", "VV")])
+        self.assertEqual(result[0][1], "VV")
+
+
 class TagNNERTestCase(unittest.TestCase):
     """Test pythainlp.tag.thai_nner"""
 

tests/core/test_util.py

@@ -40,18 +40,23 @@
     reign_year_to_ad,
     remove_dangling,
     remove_dup_spaces,
+    remove_repeat_vowels,
+    remove_spaces_before_marks,
     remove_tone_ipa,
     remove_tonemark,
     remove_trailing_repeat_consonants,
     remove_zw,
+    reorder_vowels,
     sound_syllable,
+    spell_syllable,
     spelling,
     syllable_length,
     syllable_open_close_detector,
     text_to_arabic_digit,
     text_to_num,
     text_to_thai_digit,
     th_zodiac,
+    thai_consonant_to_spelling,
     thai_digit_to_arabic_digit,
     thai_keyboard_dist,
     thai_strftime,
@@ -65,6 +70,7 @@
     to_idna,
     to_lunar_date,
     tone_detector,
+    tone_to_spelling,
     words_to_num,
 )
 from pythainlp.util.morse import morse_decode, morse_encode
@@ -1048,3 +1054,66 @@ def test_analyze_thai_text(self):
             analyze_thai_text("เล่น"),
             {'สระ เอ': 1, 'ล': 1, 'ไม้เอก': 1, 'น': 1}
         )
+
+    # ### pythainlp.util.pronounce
+
+    def test_thai_consonant_to_spelling(self):
+        self.assertEqual(thai_consonant_to_spelling("ก"), "กอ")
+        self.assertEqual(thai_consonant_to_spelling("ข"), "ขอ")
+        self.assertEqual(thai_consonant_to_spelling("น"), "นอ")
+        # multi-char strings and non-consonants pass through unchanged
+        self.assertEqual(thai_consonant_to_spelling("กา"), "กา")
+        self.assertEqual(thai_consonant_to_spelling("า"), "า")
+        self.assertEqual(thai_consonant_to_spelling("A"), "A")
+        self.assertEqual(thai_consonant_to_spelling(""), "")
+
+    def test_tone_to_spelling(self):
+        self.assertEqual(tone_to_spelling("่"), "ไม้เอก")
+        self.assertEqual(tone_to_spelling("้"), "ไม้โท")
+        self.assertEqual(tone_to_spelling("๊"), "ไม้ตรี")
+        self.assertEqual(tone_to_spelling("๋"), "ไม้จัตวา")
+        # non-tone-mark characters pass through unchanged
+        self.assertEqual(tone_to_spelling("ก"), "ก")
+        self.assertEqual(tone_to_spelling(""), "")
+
+    # ### pythainlp.util.spell_words
+
+    def test_spell_syllable(self):
+        result = spell_syllable("แมว")
+        self.assertIsInstance(result, list)
+        self.assertEqual(result[-1], "แมว")
+        result_kon = spell_syllable("คน")
+        self.assertGreater(len(result_kon), 0)
+        self.assertIn("คน", result_kon)
+
+    # ### pythainlp.util.normalize – remove_repeat_vowels,
+    #     remove_spaces_before_marks, reorder_vowels
+
+    def test_remove_repeat_vowels(self):
+        self.assertEqual(remove_repeat_vowels(""), "")
+        self.assertEqual(remove_repeat_vowels("สวัสดี"), "สวัสดี")
+        self.assertEqual(remove_repeat_vowels("นานาา"), "นานา")
+        self.assertEqual(remove_repeat_vowels("ดีีีี"), "ดี")
+        # double sara E is reordered to sara Ae before repeat-removal
+        self.assertEqual(remove_repeat_vowels("เเปลก"), "แปลก")
+
+    def test_remove_spaces_before_marks(self):
+        self.assertEqual(remove_spaces_before_marks(""), "")
+        self.assertEqual(remove_spaces_before_marks("กิน"), "กิน")
+        self.assertEqual(remove_spaces_before_marks("ก ิ"), "กิ")
+        self.assertEqual(remove_spaces_before_marks("ก ุ"), "กุ")
+        self.assertEqual(remove_spaces_before_marks("ก ่า"), "ก่า")
+        # spaces between regular consonants are preserved
+        self.assertIn(" ", remove_spaces_before_marks("ก ข"))
+
+    def test_reorder_vowels(self):
+        self.assertEqual(reorder_vowels(""), "")
+        self.assertEqual(reorder_vowels("สวัสดี"), "สวัสดี")
+        # two sara E → sara Ae
+        self.assertEqual(reorder_vowels("เเปลก"), "แปลก")
+        # nikhahit (ํ) + sara aa (า) → sara am (ำ)
+        self.assertEqual(reorder_vowels("\u0e01\u0e4d\u0e32"), "\u0e01\u0e33")
+        # tone mark reorder: both characters still present after reorder
+        result = reorder_vowels("\u0e01\u0e32\u0e48")
+        self.assertIn("\u0e48", result)
+        self.assertIn("\u0e32", result)