HowdenParser 1.0.1__tar.gz → 2.0.1__tar.gz

{howdenparser-1.0.1 → howdenparser-2.0.1}/HowdenParser/parameter/llamaparser.py

@@ -2,7 +2,12 @@
 from typing import Literal
 from pydantic import BaseModel
 
+class Hest(BaseModel):
+    provider_and_model: str = "huggingface:trocr-large-printed"
+
+
 class Parameter(BaseModel):
+    model1: Hest = Hest()
     provider_and_model: str = "llamaparser:"
     result_type: Literal["md"] = "md"
     mode: bool = False

{howdenparser-1.0.1 → howdenparser-2.0.1}/HowdenParser/parser.py

@@ -5,6 +5,7 @@ from pathlib import Path
 from PyPDF2 import PdfReader
 import dotenv
 import logging
+from inspect import signature
 
 dotenv.load_dotenv()
 logger = logging.getLogger(__name__)
@@ -30,7 +31,7 @@ class Parser(BaseParser):
 
     @classmethod
     def available_parsers(cls) -> None:
-        """Return registered parsers and their init arguments."""
+        """Print registered parsers and their init arguments."""
         import inspect
         result = {}
         for name, parser_cls in BaseParser._registry.items():
@@ -41,23 +42,48 @@ class Parser(BaseParser):
             print(f"{key} with parameters: {values}")
 
     @classmethod
-    def create(cls, config: dict | None = None, **kwargs) -> BaseParser:
-        provider = kwargs["provider_and_model"].split(":")[0].lower()
-        model = kwargs["provider_and_model"].split(":")[1].lower()
+    def create(cls, config_or_dict: "Parameter | dict", **kwargs) -> BaseParser:
+        """
+        Dynamically create parser instances from a config object or dict.
+        Only passes arguments accepted by the parser constructor.
+        """
+        # Convert Parameter -> dict if needed
+        if hasattr(config_or_dict, "model_dump"):
+            config_dict = config_or_dict.model_dump()
+        elif isinstance(config_or_dict, dict):
+            config_dict = config_or_dict
+        else:
+            raise TypeError("Expected Parameter instance or dict for config_or_dict")
+
+        if "provider_and_model" not in kwargs and "provider_and_model" not in config_dict:
+            raise ValueError("provider_and_model must be specified")
+
+        # Merge dict + kwargs (kwargs take precedence)
+        merged_args = {**config_dict, **kwargs}
+
+        # Extract provider/model info
+        provider_and_model = merged_args.get("provider_and_model")
+        provider, model = provider_and_model.split(":")
+        provider = provider.lower()
+        model = model.lower()
+
         if provider not in BaseParser._registry:
-            raise ValueError(f"Unknown parser '{provider}'. "
-                             f"Available: {cls.available_parsers()}")
+            raise ValueError(f"Unknown parser '{provider}'. Available: {list(BaseParser._registry)}")
 
         parser_cls = BaseParser._registry[provider]
-        import inspect
-        merged_args = {**(config or {}), **kwargs}
-        if "model" in inspect.signature(parser_cls.__init__).parameters:
-            merged_args["model"] = model
 
-        # Remove keys not in constructor
-        sig = inspect.signature(parser_cls.__init__)
+        # Inspect constructor and only pass valid arguments
+        sig = signature(parser_cls.__init__)
         valid_args = {k: v for k, v in merged_args.items() if k in sig.parameters and k != "self"}
 
+        # Auto-insert 'model' if required
+        if "model" in sig.parameters and "model" not in valid_args:
+            valid_args["model"] = model
+
+        # Auto-insert 'provider_and_model' if required
+        if "provider_and_model" in sig.parameters and "provider_and_model" not in valid_args:
+            valid_args["provider_and_model"] = provider_and_model
+
         return parser_cls(**valid_args)
 
     @abstractmethod

{howdenparser-1.0.1 → howdenparser-2.0.1}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.3
 Name: HowdenParser
-Version: 1.0.1
+Version: 2.0.1
 Summary: A simple configuration manager with Pydantic and JSON export.
 License: MIT
 Keywords: config,configuration,pydantic,json
@@ -13,7 +13,6 @@ Classifier: Programming Language :: Python :: 3.12
 Classifier: Programming Language :: Python :: 3.13
 Requires-Dist: fitz (>=0.0.1.dev2,<0.0.2)
 Requires-Dist: hf-xet (>=1.1.7,<2.0.0)
-Requires-Dist: howdenconfig (>=0.1.13,<0.2.0)
 Requires-Dist: langchain (>=0.3.27,<0.4.0)
 Requires-Dist: llama-parse (>=0.6.58,<0.7.0)
 Requires-Dist: mistralai (>=1.9.3,<2.0.0)
@@ -78,8 +77,9 @@ print(text)
 
 if HowdenConfig package being used
 
+config: Config = Config(parameter=Parameter())
 
-parser = ParserFactory.get_parser("mistralocr:", **config.parameter.dump_model()) 
+parser = Parser.create(config.parameter)
 
 text = parser.parse(Path("document.pdf"))
 

{howdenparser-1.0.1 → howdenparser-2.0.1}/README.md

@@ -48,8 +48,9 @@ print(text)
 
 if HowdenConfig package being used
 
+config: Config = Config(parameter=Parameter())
 
-parser = ParserFactory.get_parser("mistralocr:", **config.parameter.dump_model()) 
+parser = Parser.create(config.parameter)
 
 text = parser.parse(Path("document.pdf"))
 

{howdenparser-1.0.1 → howdenparser-2.0.1}/pyproject.toml

@@ -3,7 +3,7 @@ name = "HowdenParser"
 description = ""
 readme = "README.md"
 requires-python = ">=3.12,<3.14"
-dependencies = [ "mistralai (>=1.9.3,<2.0.0)", "llama-parse (>=0.6.58,<0.7.0)", "langchain (>=0.3.27,<0.4.0)", "transformers (>=4.55.2,<5.0.0)", "fitz (>=0.0.1.dev2,<0.0.2)", "howdenconfig (>=0.1.13,<0.2.0)", "pypdf2 (>=3.0.1,<4.0.0)", "pdf2image (>=1.17.0,<2.0.0)", "torch (>=2.8.0,<3.0.0)", "torchvision (>=0.23.0,<0.24.0)", "torchaudio (>=2.8.0,<3.0.0)", "hf-xet (>=1.1.7,<2.0.0)",]
+dependencies = [ "mistralai (>=1.9.3,<2.0.0)", "llama-parse (>=0.6.58,<0.7.0)", "langchain (>=0.3.27,<0.4.0)", "transformers (>=4.55.2,<5.0.0)", "fitz (>=0.0.1.dev2,<0.0.2)", "pypdf2 (>=3.0.1,<4.0.0)", "pdf2image (>=1.17.0,<2.0.0)", "torch (>=2.8.0,<3.0.0)", "torchvision (>=0.23.0,<0.24.0)", "torchaudio (>=2.8.0,<3.0.0)", "hf-xet (>=1.1.7,<2.0.0)",]
 [[project.authors]]
 name = "JesperThoftIllemannJ"
 email = "jesper.jaeger@howdendanmark.dk"
@@ -14,7 +14,7 @@ build-backend = "poetry.core.masonry.api"
 
 [tool.poetry]
 name = "HowdenParser"
-version = "1.0.1"
+version = "2.0.1"
 description = "A simple configuration manager with Pydantic and JSON export."
 authors = [ "JesperThoftIllemannJ <jesper.jaeger@howdendanmark.dk>",]
 readme = "README.md"