➕ Add tomli

pyproject.toml

@@ -28,7 +28,7 @@ classifiers = [
   "Programming Language :: Python :: Implementation :: CPython",
   "Programming Language :: Python :: Implementation :: PyPy",
 ]
-dynamic = ["version", "optional-dependencies"]
+dynamic = ["version", "dependencies", "optional-dependencies"]
 
 [[project.authors]]
 name = "Wu Zhenyu"
@@ -65,6 +65,9 @@ write-to = "src/translate_shell/_metainfo.py"
 [tool.setuptools-generate.metainfo-template]
 file = "templates/metainfo.py.j2"
 
+[tool.setuptools.dynamic.dependencies]
+file = "requirements.txt"
+
 # begin: scripts/update-pyproject.toml.pl
 [tool.setuptools.dynamic.optional-dependencies.color]
 file = "requirements/color.txt"

requirements.txt

@@ -0,0 +1,3 @@
+#!/usr/bin/env -S pip install -r
+
+tomli; python_version < "3.11"

src/translate_shell/__main__.py

@@ -149,6 +149,13 @@ def get_parser() -> ArgumentParser:
         default=config.source_lang,
         help="source languages. default: %(default)s",
     ).complete = LANG_COMPLETE  # type: ignore
+    parser.add_argument(
+        "--options",
+        help="advanced usage, see "
+        "https://translate-shell.readthedocs.io/en/latest/resources/config.html "
+        ". default: %(default)s",
+        action="append",
+    ).complete = LANG_COMPLETE  # type: ignore
     parser.add_argument(
         "text",
         nargs="*",

src/translate_shell/tools/po/main.py

@@ -10,6 +10,11 @@
 
 from translate_shell.translate import translate
 
+try:
+    import tomllib as tomli
+except ImportError:
+    import tomli
+
 logger = logging.getLogger(__name__)
 
 
@@ -24,10 +29,7 @@ def run(args: Namespace) -> None:
     default_target_lang = args.target_lang
     source_lang = args.source_lang
     translator = args.translator
-    option = {
-        option.partition("=")[0]: option.partition("=")[2]
-        for option in args.option
-    }
+    option = tomli.loads("\n".join(args.option))
     wrapwidth = int(args.wrapwidth)
     progress = args.progress.lower() == "true"
     verbose = args.verbose.lower() == "true"

src/translate_shell/translate.py

@@ -96,6 +96,7 @@ def translate(
         if isinstance(translator, Callable):
             translator = translator()
         true_translators += [translator]
+
     if len(translators) == 1:
         translator = true_translators[0]
         translate_once(

src/translate_shell/translators/llm/_llama_cpp.py

@@ -9,7 +9,11 @@
 from ...external.platformdirs import AppDirs
 from . import LLMTranslator
 
-MODEL = str(AppDirs("translate-shell").user_data_path / "model.bin")
+MODEL_PATH = str(AppDirs("translate-shell").user_data_path / "model.bin")
+# cache init
+old_model_path = MODEL_PATH
+old_kwargs = {}
+old_model = None
 
 
 class LlamaTranslator(LLMTranslator):
@@ -30,49 +34,63 @@ def init_model(option: dict) -> Any:
         :type option: dict
         :rtype: Any
         """
-        model = option.get("model", MODEL)
-        if isinstance(model, str):
-            model = os.path.expanduser(model)
-            kwargs = {}
-            if n_ctx := option.get("n_ctx"):
-                kwargs["n_ctx"] = int(n_ctx)
-            if n_parts := option.get("n_parts"):
-                kwargs["n_parts"] = int(n_parts)
-            if n_gpu_layers := option.get("n_gpu_layers"):
-                kwargs["n_gpu_layers"] = int(n_gpu_layers)
-            if seed := option.get("seed"):
-                kwargs["seed"] = int(seed)
-            if f16_kv := option.get("f16_kv"):
-                kwargs["f16_kv"] = bool(f16_kv)
-            if logits_all := option.get("logits_all"):
-                kwargs["logits_all"] = bool(logits_all)
-            if vocab_only := option.get("vocab_only"):
-                kwargs["vocab_only"] = bool(vocab_only)
-            if use_mmap := option.get("use_mmap"):
-                kwargs["use_mmap"] = bool(use_mmap)
-            if use_mlock := option.get("use_mlock"):
-                kwargs["use_mlock"] = bool(use_mlock)
-            if embedding := option.get("embedding"):
-                kwargs["embedding"] = bool(embedding)
-            if n_threads := option.get("n_threads"):
-                kwargs["n_threads"] = n_threads
-            if n_batch := option.get("n_batch"):
-                kwargs["n_batch"] = int(n_batch)
-            if last_n_tokens_size := option.get("last_n_tokens_size"):
-                kwargs["last_n_tokens_size"] = int(last_n_tokens_size)
-            if lora_base := option.get("lora_base"):
-                kwargs["lora_base"] = lora_base
-            if lora_path := option.get("lora_path"):
-                kwargs["lora_path"] = lora_path
-            if low_vram := option.get("low_vram"):
-                kwargs["low_vram"] = bool(low_vram)
-            if tensor_split := option.get("tensor_split"):
-                kwargs["tensor_split"] = tensor_split
-            if rope_freq_base := option.get("rope_freq_base"):
-                kwargs["rope_freq_base"] = float(rope_freq_base)
-            if rope_freq_scale := option.get("rope_freq_scale"):
-                kwargs["rope_freq_scale"] = float(rope_freq_scale)
-            if verbose := option.get("verbose"):
-                kwargs["verbose"] = bool(verbose)
-            model = Llama(model, **kwargs)
+        global old_model_path, old_kwargs, old_model
+        model_path = option.get("model", MODEL_PATH)
+        if isinstance(model_path, Llama):
+            # cache clear
+            old_model_path = ""
+            old_kwargs = {}
+            old_model = model_path
+            return model_path
+        model_path = os.path.expanduser(model_path)
+        kwargs = {}
+        if n_ctx := option.get("n_ctx"):
+            kwargs["n_ctx"] = int(n_ctx)
+        if n_parts := option.get("n_parts"):
+            kwargs["n_parts"] = int(n_parts)
+        if n_gpu_layers := option.get("n_gpu_layers"):
+            kwargs["n_gpu_layers"] = int(n_gpu_layers)
+        if seed := option.get("seed"):
+            kwargs["seed"] = int(seed)
+        if f16_kv := option.get("f16_kv"):
+            kwargs["f16_kv"] = bool(f16_kv)
+        if logits_all := option.get("logits_all"):
+            kwargs["logits_all"] = bool(logits_all)
+        if vocab_only := option.get("vocab_only"):
+            kwargs["vocab_only"] = bool(vocab_only)
+        if use_mmap := option.get("use_mmap"):
+            kwargs["use_mmap"] = bool(use_mmap)
+        if use_mlock := option.get("use_mlock"):
+            kwargs["use_mlock"] = bool(use_mlock)
+        if embedding := option.get("embedding"):
+            kwargs["embedding"] = bool(embedding)
+        if n_threads := option.get("n_threads"):
+            kwargs["n_threads"] = n_threads
+        if n_batch := option.get("n_batch"):
+            kwargs["n_batch"] = int(n_batch)
+        if last_n_tokens_size := option.get("last_n_tokens_size"):
+            kwargs["last_n_tokens_size"] = int(last_n_tokens_size)
+        if lora_base := option.get("lora_base"):
+            kwargs["lora_base"] = lora_base
+        if lora_path := option.get("lora_path"):
+            kwargs["lora_path"] = lora_path
+        if low_vram := option.get("low_vram"):
+            kwargs["low_vram"] = bool(low_vram)
+        if tensor_split := option.get("tensor_split"):
+            kwargs["tensor_split"] = tensor_split
+        if rope_freq_base := option.get("rope_freq_base"):
+            kwargs["rope_freq_base"] = float(rope_freq_base)
+        if rope_freq_scale := option.get("rope_freq_scale"):
+            kwargs["rope_freq_scale"] = float(rope_freq_scale)
+        if verbose := option.get("verbose"):
+            kwargs["verbose"] = bool(verbose)
+        # cache hit
+        if kwargs == old_kwargs and model_path == old_model_path and old_model:
+            model = old_model
+        else:
+            model = Llama(model_path, **kwargs)
+            # cache reinit
+            old_model_path = model_path
+            old_kwargs = kwargs
+            old_model = model
         return model

src/translate_shell/translators/stardict/__init__.py

@@ -99,6 +99,9 @@ def get_tokens(self, text: str, tl: str, sl: str) -> tuple[list[str], str]:
                 sl = detect(text)
             except LangDetectException:
                 sl = "en"
+            # convert zh-cn to zh_CN
+            lang, _, country = sl.partition("-")
+            sl = lang + "_" + country.upper() if country else ""
 
         dictionaries = self.stardict.get(sl, self.stardict["en"]).get(tl, [])
         for directory in STARDICT_DIRS:

src/translate_shell/ui/__init__.py

@@ -18,6 +18,11 @@
 from ..translate import translate
 from ..translators import TRANSLATORS, get_dummy
 
+try:
+    import tomllib as tomli
+except ImportError:
+    import tomli
+
 with suppress(ImportError):
     from rich import traceback
     from rich.logging import RichHandler
@@ -112,6 +117,7 @@ def init(args: Namespace) -> None:
         if value is not None:
             setattr(args, attr, value)
     args.text = " ".join(args.text)
+    args.options = tomli.loads("\n".join(args.options))
     if not args.lsp:
         _readline = init_readline()
         _readline.set_completer(args.complete)
@@ -172,6 +178,7 @@ def get_processed_result_text(
         target_lang,
         source_lang,
         translators,
+        args.options,
     )
     if args.format == "json":
         rst = json.dumps(vars(translation))

tests/txt/options.txt

@@ -27,3 +27,6 @@ options:
                         target languages. default: auto
   --source-lang SOURCE_LANG
                         source languages. default: auto
+  --options OPTIONS     advanced usage, see https://translate-
+                        shell.readthedocs.io/en/latest/resources/config.html .
+                        default: None

tests/txt/usage.txt

@@ -5,4 +5,5 @@ usage: trans [-h] [-V] [--print-completion {bash,zsh,tcsh}]
              [--sleep-seconds SLEEP_SECONDS] [--config CONFIG]
              [--format {json,yaml,text}] [--translators TRANSLATORS]
              [--target-lang TARGET_LANG] [--source-lang SOURCE_LANG]
+             [--options OPTIONS]
              [text ...]