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

[Copilot]

Test classes in testn_tokenize_onnx.py and testn_tokenize_torch.py covered non-tokenization functionality (transliteration, tagging, spell, parse, summarize, augment), making failures hard to trace and suite boundaries meaningless. Also fixes a dead try/except Exception in the spell module.

Test reorganization — ONNX (tests/noauto_onnx/)

Extracted from testn_tokenize_onnx.py into dedicated files:

New file	Class
testn_transliterate_onnx.py	TransliterateONNXTestCaseN
testn_tag_onnx.py	TagONNXTestCaseN
testn_spell_onnx.py	SpellONNXTestCaseN

Test reorganization — Torch (tests/noauto_torch/)

Extracted from testn_tokenize_torch.py into dedicated files:

New file	Class
testn_transliterate_torch.py	TransliterateTestCaseN
testn_parse_torch.py	ParseTestCaseN
testn_summarize_torch.py	SummarizeTestCaseN
testn_augment_torch.py	AugmentTestCaseN

Both __init__.py files updated to register new modules in alphabetical order.

Increased coverage

Each new file expands the original single-assertion tests with: model-loaded key checks (<start>/<end> in vocab dicts), empty-input handling, result structure validation (tuple pairs, string items), length constraints, and ValueError on out-of-range arguments.

Bug fix — too-broad try/except

pythainlp/spell/wanchanberta_thai_grammarly.py: the except Exception block wrapping label_ids.append(2) was unreachable (list.append never raises) and silently swallowed any future logic errors in that branch.

# Before — dead except, hides future bugs
try:
    label_ids.append(2)
except Exception:
    label_ids.append(-100)

# After
label_ids.append(2)

Your checklist for this pull request

• Passed code styles and structures
• Passed code linting checks and unit test

---

✨ Let Copilot coding agent set things up for you — coding agent works faster and does higher quality work when set up for your repo.

[sonarqubecloud]

Quality Gate passed

Issues
0 New issues
0 Accepted issues

Measures
0 Security Hotspots
0.0% Coverage on New Code
0.0% Duplication on New Code

See analysis details on SonarQube Cloud