Improve dependency error handling and type annotations in lm module

1. Add type annotation to __all__ in pythainlp/lm/__init__.py
   - Changed from __all__ = [...] to __all__: list[str] = [...]
   - Matches codebase convention (e.g., pythainlp/chat/__init__.py)

2. Remove try/except wrapper around Qwen3 import in __init__.py
   - The wrapper was ineffective since qwen3.py uses lazy imports
   - Users would get raw ModuleNotFoundError instead of helpful message
   - Better to handle missing deps in the methods that actually need them

3. Add dependency checking in Qwen3 methods (qwen3.py)
   - load_model(): Wrap torch/transformers imports in try/except
   - generate(): Wrap torch import in try/except
   - chat(): Wrap torch import in try/except
   - All raise ImportError with helpful message: "Install them with: pip install 'pythainlp[qwen3]'"
   - Catches both ImportError and ModuleNotFoundError for robustness

These changes ensure users get clear, actionable error messages when trying to use
Qwen3 without installing the optional dependencies, rather than cryptic import errors.

Co-authored-by: bact <128572+bact@users.noreply.github.com>

Author: Copilot

pythainlp/lm/__init__.py

@@ -2,21 +2,11 @@
 # SPDX-FileType: SOURCE
 # SPDX-License-Identifier: Apache-2.0
 
-__all__ = ["calculate_ngram_counts", "remove_repeated_ngrams", "Qwen3"]
+__all__: list[str] = ["calculate_ngram_counts", "remove_repeated_ngrams", "Qwen3"]
 
 from pythainlp.lm.text_util import (
     calculate_ngram_counts,
     remove_repeated_ngrams,
 )
 
-try:
-    from pythainlp.lm.qwen3 import Qwen3
-except ImportError:
-    # If dependencies are not installed, make Qwen3 available but raise
-    # error when instantiated
-    class Qwen3:  # type: ignore
-        def __init__(self) -> None:
-            raise ImportError(
-                "Qwen3 requires additional dependencies. "
-                "Install with: pip install pythainlp[qwen3]"
-            )
+from pythainlp.lm.qwen3 import Qwen3

pythainlp/lm/qwen3.py

@@ -48,8 +48,14 @@ def load_model(
             model = Qwen3()
             model.load_model(device="cpu", torch_dtype=torch.bfloat16)
         """
-        import torch
-        from transformers import AutoModelForCausalLM, AutoTokenizer
+        try:
+            import torch
+            from transformers import AutoModelForCausalLM, AutoTokenizer
+        except (ImportError, ModuleNotFoundError) as exc:
+            raise ImportError(
+                "Qwen3 language model requires optional dependencies. "
+                "Install them with: pip install 'pythainlp[qwen3]'"
+            ) from exc
 
         # Set default torch_dtype if not provided
         if torch_dtype is None:
@@ -144,7 +150,13 @@ def generate(
                 "text parameter must be a non-empty string."
             )
 
-        import torch
+        try:
+            import torch
+        except (ImportError, ModuleNotFoundError) as exc:
+            raise ImportError(
+                "Qwen3 language model requires optional dependencies. "
+                "Install them with: pip install 'pythainlp[qwen3]'"
+            ) from exc
 
         inputs = self.tokenizer(text, return_tensors="pt")
         input_ids = inputs["input_ids"].to(self.device)
@@ -232,7 +244,13 @@ def chat(
                 lines.append(f"{role}: {content}")
             text = "\n".join(lines) + "\nassistant: "
 
-        import torch
+        try:
+            import torch
+        except (ImportError, ModuleNotFoundError) as exc:
+            raise ImportError(
+                "Qwen3 language model requires optional dependencies. "
+                "Install them with: pip install 'pythainlp[qwen3]'"
+            ) from exc
 
         inputs = self.tokenizer(text, return_tensors="pt")
         input_ids = inputs["input_ids"].to(self.device)