refactor(tests): reorganize noauto test cases by function; fix too-broad except

pythainlp/spell/wanchanberta_thai_grammarly.py

@@ -71,10 +71,7 @@ def align_word_ids(texts: str) -> list[int]:
         if word_idx is None:
             label_ids.append(-100)
         else:
-            try:
-                label_ids.append(2)
-            except Exception:
-                label_ids.append(-100)
+            label_ids.append(2)
 
     return label_ids
 

tests/noauto_onnx/__init__.py

@@ -21,7 +21,10 @@
 
 # Names of module to be tested
 test_packages: list[str] = [
+    "tests.noauto_onnx.testn_spell_onnx",
+    "tests.noauto_onnx.testn_tag_onnx",
     "tests.noauto_onnx.testn_tokenize_onnx",
+    "tests.noauto_onnx.testn_transliterate_onnx",
 ]
 
 

tests/noauto_onnx/testn_spell_onnx.py

@@ -0,0 +1,41 @@
+# SPDX-FileCopyrightText: 2016-2026 PyThaiNLP Project
+# SPDX-FileType: SOURCE
+# SPDX-License-Identifier: Apache-2.0
+
+# Tests for spell correction functions that require ONNX Runtime
+# These tests are NOT run in automated CI workflows due to:
+# - Large dependencies (onnxruntime)
+# - Platform-specific compatibility issues
+# - Version constraints
+
+import unittest
+
+
+class SpellONNXTestCaseN(unittest.TestCase):
+    """Tests for ONNX-based spell correction (requires onnxruntime)"""
+
+    def test_words_spelling_correction_returns_list(self):
+        from pythainlp.spell.words_spelling_correction import (
+            get_words_spell_suggestion,
+        )
+
+        result = get_words_spell_suggestion("สวัสดี")
+        self.assertIsInstance(result, list)
+
+    def test_words_spelling_correction_nonempty_input(self):
+        from pythainlp.spell.words_spelling_correction import (
+            get_words_spell_suggestion,
+        )
+
+        result = get_words_spell_suggestion("กาารเขียน")
+        self.assertIsInstance(result, list)
+        self.assertGreater(len(result), 0)
+
+    def test_words_spelling_correction_items_are_strings(self):
+        from pythainlp.spell.words_spelling_correction import (
+            get_words_spell_suggestion,
+        )
+
+        result = get_words_spell_suggestion("กาารเขียน")
+        for item in result:
+            self.assertIsInstance(item, str)

tests/noauto_onnx/testn_tag_onnx.py

@@ -0,0 +1,53 @@
+# SPDX-FileCopyrightText: 2016-2026 PyThaiNLP Project
+# SPDX-FileType: SOURCE
+# SPDX-License-Identifier: Apache-2.0
+
+# Tests for POS tagging functions that require ONNX Runtime
+# These tests are NOT run in automated CI workflows due to:
+# - Large dependencies (onnxruntime)
+# - Platform-specific compatibility issues
+# - Version constraints
+
+import unittest
+
+
+class TagONNXTestCaseN(unittest.TestCase):
+    """Tests for ONNX-based POS tagging (requires onnxruntime)"""
+
+    def test_pos_tag_wangchanberta_onnx_returns_list(self):
+        from pythainlp.tag import pos_tag
+
+        result = pos_tag(
+            ["แมว", "กิน", "ปลา"],
+            engine="wangchanberta_onnx",
+        )
+        self.assertIsInstance(result, list)
+        self.assertGreater(len(result), 0)
+
+    def test_pos_tag_wangchanberta_onnx_length_matches(self):
+        from pythainlp.tag import pos_tag
+
+        tokens = ["แมว", "กิน", "ปลา"]
+        result = pos_tag(tokens, engine="wangchanberta_onnx")
+        self.assertEqual(len(result), len(tokens))
+
+    def test_pos_tag_wangchanberta_onnx_tuple_pairs(self):
+        from pythainlp.tag import pos_tag
+
+        result = pos_tag(
+            ["แมว", "กิน", "ปลา"],
+            engine="wangchanberta_onnx",
+        )
+        for item in result:
+            self.assertIsInstance(item, tuple)
+            self.assertEqual(len(item), 2)
+            word, tag = item
+            self.assertIsInstance(word, str)
+            self.assertIsInstance(tag, str)
+
+    def test_pos_tag_wangchanberta_onnx_empty_list(self):
+        from pythainlp.tag import pos_tag
+
+        result = pos_tag([], engine="wangchanberta_onnx")
+        self.assertIsInstance(result, list)
+        self.assertEqual(len(result), 0)

tests/noauto_onnx/testn_tokenize_onnx.py

@@ -79,40 +79,3 @@ def test_sefr_cut(self):
             sefr_cut.segment("ฉันรักภาษาไทยเพราะฉันเป็นคนไทย", engine="tnhc"),
         )
 
-
-class TransliterateONNXTestCaseN(unittest.TestCase):
-    """Tests for ONNX-based transliteration (requires onnxruntime)"""
-
-    def test_thai2rom_onnx(self):
-        from pythainlp.transliterate.thai2rom_onnx import romanize
-
-        result = romanize("สวัสดี")
-        self.assertIsInstance(result, str)
-        self.assertGreater(len(result), 0)
-
-
-class TagONNXTestCaseN(unittest.TestCase):
-    """Tests for ONNX-based POS tagging (requires onnxruntime)"""
-
-    def test_pos_tag_wangchanberta_onnx(self):
-        from pythainlp.tag import pos_tag
-
-        result = pos_tag(
-            ["แมว", "กิน", "ปลา"],
-            engine="wangchanberta_onnx"
-        )
-        self.assertIsInstance(result, list)
-        self.assertGreater(len(result), 0)
-        self.assertEqual(len(result), 3)
-
-
-class SpellONNXTestCaseN(unittest.TestCase):
-    """Tests for ONNX-based spell correction (requires onnxruntime)"""
-
-    def test_words_spelling_correction(self):
-        from pythainlp.spell.words_spelling_correction import (
-            get_words_spell_suggestion,
-        )
-
-        result = get_words_spell_suggestion("สวัสดี")
-        self.assertIsInstance(result, list)

tests/noauto_onnx/testn_transliterate_onnx.py

@@ -0,0 +1,41 @@
+# SPDX-FileCopyrightText: 2016-2026 PyThaiNLP Project
+# SPDX-FileType: SOURCE
+# SPDX-License-Identifier: Apache-2.0
+
+# Tests for transliteration functions that require ONNX Runtime
+# These tests are NOT run in automated CI workflows due to:
+# - Large dependencies (onnxruntime)
+# - Platform-specific compatibility issues
+# - Version constraints
+
+import unittest
+
+
+class TransliterateONNXTestCaseN(unittest.TestCase):
+    """Tests for ONNX-based transliteration (requires onnxruntime)"""
+
+    def test_thai2rom_onnx_returns_string(self):
+        from pythainlp.transliterate.thai2rom_onnx import romanize
+
+        result = romanize("สวัสดี")
+        self.assertIsInstance(result, str)
+        self.assertGreater(len(result), 0)
+
+    def test_thai2rom_onnx_empty_string(self):
+        from pythainlp.transliterate.thai2rom_onnx import romanize
+
+        result = romanize("")
+        self.assertIsInstance(result, str)
+
+    def test_thai2rom_onnx_ascii_passthrough(self):
+        from pythainlp.transliterate.thai2rom_onnx import romanize
+
+        result = romanize("hello")
+        self.assertIsInstance(result, str)
+
+    def test_thai2rom_onnx_mixed_text(self):
+        from pythainlp.transliterate.thai2rom_onnx import romanize
+
+        result = romanize("ภาษาไทย")
+        self.assertIsInstance(result, str)
+        self.assertGreater(len(result), 0)

tests/noauto_torch/__init__.py

@@ -23,10 +23,14 @@
 
 # Names of module to be tested
 test_packages: list[str] = [
+    "tests.noauto_torch.testn_augment_torch",
     "tests.noauto_torch.testn_lm_torch",
+    "tests.noauto_torch.testn_parse_torch",
     "tests.noauto_torch.testn_spell_torch",
+    "tests.noauto_torch.testn_summarize_torch",
     "tests.noauto_torch.testn_tag_torch",
     "tests.noauto_torch.testn_tokenize_torch",
+    "tests.noauto_torch.testn_transliterate_torch",
 ]
 
 

tests/noauto_torch/testn_augment_torch.py

@@ -0,0 +1,75 @@
+# SPDX-FileCopyrightText: 2016-2026 PyThaiNLP Project
+# SPDX-FileType: SOURCE
+# SPDX-License-Identifier: Apache-2.0
+
+# Tests for augmentation functions that require transformers
+# These tests are NOT run in automated CI workflows due to:
+# - Large dependencies (torch, transformers)
+# - Python 3.13+ compatibility issues
+
+import unittest
+
+
+class AugmentTestCaseN(unittest.TestCase):
+    """Tests for augmentation functions (requires transformers)"""
+
+    def test_augment_wangchanberta_returns_list(self):
+        from pythainlp.augment.lm import Thai2transformersAug
+
+        augmenter = Thai2transformersAug()
+        result = augmenter.augment("แมวกิน<mask>")
+        self.assertIsInstance(result, list)
+        self.assertGreater(len(result), 0)
+
+    def test_augment_wangchanberta_items_are_strings(self):
+        from pythainlp.augment.lm import Thai2transformersAug
+
+        augmenter = Thai2transformersAug()
+        result = augmenter.augment("แมวกิน<mask>")
+        for item in result:
+            self.assertIsInstance(item, str)
+
+    def test_augment_wangchanberta_generate_returns_list(self):
+        from pythainlp.augment.lm import Thai2transformersAug
+
+        augmenter = Thai2transformersAug()
+        result = augmenter.generate("แมวกิน<mask>", num_replace_tokens=1)
+        self.assertIsInstance(result, list)
+
+    def test_augment_phayathaibert_returns_list(self):
+        from pythainlp.augment.lm import ThaiTextAugmenter
+
+        augmenter = ThaiTextAugmenter()
+        result = augmenter.augment("แมวกิน<mask>")
+        self.assertIsInstance(result, list)
+        self.assertGreater(len(result), 0)
+
+    def test_augment_phayathaibert_items_are_strings(self):
+        from pythainlp.augment.lm import ThaiTextAugmenter
+
+        augmenter = ThaiTextAugmenter()
+        result = augmenter.augment("แมวกิน<mask>")
+        for item in result:
+            self.assertIsInstance(item, str)
+
+    def test_augment_phayathaibert_num_augs_respected(self):
+        from pythainlp.augment.lm import ThaiTextAugmenter
+
+        augmenter = ThaiTextAugmenter()
+        result = augmenter.augment("แมวกิน<mask>", num_augs=2)
+        self.assertEqual(len(result), 2)
+
+    def test_augment_phayathaibert_exceeds_limit_raises(self):
+        from pythainlp.augment.lm import ThaiTextAugmenter
+
+        augmenter = ThaiTextAugmenter()
+        with self.assertRaises(ValueError):
+            augmenter.augment("แมวกิน<mask>", num_augs=10)
+
+    def test_augment_phayathaibert_adds_mask_if_missing(self):
+        from pythainlp.augment.lm import ThaiTextAugmenter
+
+        augmenter = ThaiTextAugmenter()
+        result = augmenter.augment("แมวกิน", num_augs=1)
+        self.assertIsInstance(result, list)
+        self.assertEqual(len(result), 1)

tests/noauto_torch/testn_parse_torch.py

@@ -0,0 +1,34 @@
+# SPDX-FileCopyrightText: 2016-2026 PyThaiNLP Project
+# SPDX-FileType: SOURCE
+# SPDX-License-Identifier: Apache-2.0
+
+# Tests for parsing functions that require torch and transformers
+# These tests are NOT run in automated CI workflows due to:
+# - Large dependencies (torch, transformers)
+# - Python 3.13+ compatibility issues
+
+import unittest
+
+
+class ParseTestCaseN(unittest.TestCase):
+    """Tests for parsing functions (requires torch and transformers)"""
+
+    def test_dependency_parsing_returns_list(self):
+        from pythainlp.parse import dependency_parsing
+
+        result = dependency_parsing("แมวกินปลา")
+        self.assertIsInstance(result, list)
+        self.assertGreater(len(result), 0)
+
+    def test_dependency_parsing_result_structure(self):
+        from pythainlp.parse import dependency_parsing
+
+        result = dependency_parsing("แมวกินปลา")
+        for item in result:
+            self.assertIsInstance(item, dict)
+
+    def test_dependency_parsing_empty_string(self):
+        from pythainlp.parse import dependency_parsing
+
+        result = dependency_parsing("")
+        self.assertIsInstance(result, list)

tests/noauto_torch/testn_summarize_torch.py

@@ -0,0 +1,47 @@
+# SPDX-FileCopyrightText: 2016-2026 PyThaiNLP Project
+# SPDX-FileType: SOURCE
+# SPDX-License-Identifier: Apache-2.0
+
+# Tests for summarization functions that require transformers
+# These tests are NOT run in automated CI workflows due to:
+# - Large dependencies (torch, transformers)
+# - Python 3.13+ compatibility issues
+
+import unittest
+
+
+class SummarizeTestCaseN(unittest.TestCase):
+    """Tests for summarization functions (requires transformers)"""
+
+    def test_summarize_keybert_returns_list(self):
+        from pythainlp.summarize.keybert import KeyBERT
+
+        text = "แมวเป็นสัตว์เลี้ยงที่น่ารัก แมวชอบกินปลา แมวชอบนอนหลับ"
+        keybert = KeyBERT()
+        result = keybert.extract_keywords(text, max_keywords=2)
+        self.assertIsInstance(result, list)
+
+    def test_summarize_keybert_max_keywords_respected(self):
+        from pythainlp.summarize.keybert import KeyBERT
+
+        text = "แมวเป็นสัตว์เลี้ยงที่น่ารัก แมวชอบกินปลา แมวชอบนอนหลับ"
+        keybert = KeyBERT()
+        result = keybert.extract_keywords(text, max_keywords=2)
+        self.assertLessEqual(len(result), 2)
+
+    def test_summarize_mt5_returns_list(self):
+        from pythainlp.summarize.mt5 import mT5Summarizer
+
+        text = "แมวเป็นสัตว์เลี้ยงที่น่ารัก แมวชอบกินปลา แมวชอบนอนหลับ"
+        summarizer = mT5Summarizer()
+        result = summarizer.summarize(text)
+        self.assertIsInstance(result, list)
+
+    def test_summarize_mt5_result_items_are_strings(self):
+        from pythainlp.summarize.mt5 import mT5Summarizer
+
+        text = "แมวเป็นสัตว์เลี้ยงที่น่ารัก แมวชอบกินปลา แมวชอบนอนหลับ"
+        summarizer = mT5Summarizer()
+        result = summarizer.summarize(text)
+        for item in result:
+            self.assertIsInstance(item, str)