HowdenParser 2.0.3__tar.gz → 3.1.0__tar.gz

{howdenparser-2.0.3 → howdenparser-3.1.0}/HowdenParser/parser.py

@@ -4,8 +4,8 @@ import logging
 from pathlib import Path
 from PyPDF2 import PdfReader
 import dotenv
-import logging
-from inspect import signature
+from typing import overload, Literal
+from inspect import signature, Signature
 
 dotenv.load_dotenv()
 logger = logging.getLogger(__name__)
@@ -14,6 +14,10 @@ logger = logging.getLogger(__name__)
 class BaseParser(ABC):
     _registry: dict[str, type["BaseParser"]] = {}
 
+    def __init__(self):
+        # Every instance gets a .name attribute
+        self.name = getattr(self.__class__, "_parser_name", self.__class__.__name__)
+
     def __init_subclass__(cls, name: str | None = None, **kwargs):
         """Automatically register subclasses under a key."""
         super().__init_subclass__(**kwargs)
@@ -29,6 +33,13 @@ class BaseParser(ABC):
 class Parser(BaseParser):
     """Factory + registry interface for all parsers."""
 
+    def __new__(cls, config_or_dict: "Parameter | dict | str" = None, **kwargs) -> BaseParser:
+        """
+        Factory entrypoint.
+        Allows calling Parser(...) directly to create the correct subclass.
+        """
+        return cls._create(config_or_dict, **kwargs)
+
     @classmethod
     def available_parsers(cls) -> None:
         """Print registered parsers and their init arguments."""
@@ -37,53 +48,130 @@ class Parser(BaseParser):
         for name, parser_cls in BaseParser._registry.items():
             sig = inspect.signature(parser_cls.__init__)
             result[name] = [p for p in sig.parameters if p != "self"]
-        result.pop('', None)
+        result.pop("", None)
         for key, values in result.items():
             print(f"{key} with parameters: {values}")
 
+    # ---- Overloads for IDE autocomplete ----
+    @overload
+    @classmethod
+    def _create(
+            cls,
+            *,
+            provider_and_model: Literal["llamaparser:"],
+            result_type: str,
+            mode: bool,
+            preserve_layout_alignment_across_pages: bool,
+            merge_tables_across_pages_in_markdown: bool
+    ) -> "LlamaParser": ...
+
+    @overload
+    @classmethod
+    def _create(
+            cls,
+            *,
+            provider_and_model: Literal["mistralocr:ocr-large", "mistralocr:ocr-small"],
+    ) -> "MistralOCRParser": ...
+
+    @overload
+    @classmethod
+    def _create(
+            cls,
+            *,
+            provider_and_model: Literal["langchain:gpt-3.5-turbo", "langchain:gpt-4"],
+    ) -> "LangChainParser": ...
+
+    @overload
     @classmethod
-    def create(cls, config_or_dict: "Parameter | dict", **kwargs) -> BaseParser:
+    def _create(
+            cls,
+            *,
+            provider_and_model: Literal[
+                "huggingface:microsoft/trocr-base-handwritten",
+                "huggingface:microsoft/trocr-large-printed",
+            ],
+    ) -> "HuggingFaceParser": ...
+
+    # ---- Implementation ----
+    @classmethod
+    def _create(cls, config_or_dict: "Parameter | dict | str" = None, **kwargs) -> BaseParser:
         """
-        Dynamically create parser instances from a config object or dict.
-        Only passes arguments accepted by the parser constructor.
+        Dynamically create parser instances.
+        Supports: Parameter, dict, str (provider_and_model), or kwargs.
         """
-        # Convert Parameter -> dict if needed
-        if hasattr(config_or_dict, "model_dump"):
+
+        if config_or_dict is None:
+            config_dict = {}
+        elif hasattr(config_or_dict, "model_dump"):
             config_dict = config_or_dict.model_dump()
         elif isinstance(config_or_dict, dict):
             config_dict = config_or_dict
+        elif isinstance(config_or_dict, str):  # shorthand
+            config_dict = {"provider_and_model": 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")
+            raise TypeError("Expected Parameter instance, dict, str, or None for config_or_dict")
 
-        # 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_and_model" not in merged_args:
+            raise ValueError("provider_and_model must be specified, e.g. 'llamaparser:'")
+
+        provider_and_model = str(merged_args.get("provider_and_model")).strip()
+
+        if ":" not in provider_and_model:
+            raise ValueError("provider_and_model must include a colon, e.g. 'llamaparser:'")
+
+        provider, model = provider_and_model.split(":", 1)
+        provider = (provider or "").strip().lower()
+        model = (model or "").strip().lower()
+
+        if not provider:
+            raise ValueError(f"Invalid provider_and_model '{provider_and_model}': provider part is empty.")
 
         if provider not in BaseParser._registry:
             raise ValueError(f"Unknown parser '{provider}'. Available: {list(BaseParser._registry)}")
 
         parser_cls = BaseParser._registry[provider]
 
-        # Inspect constructor and only pass valid arguments
+        # --- Filter valid constructor args ---
         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
+        # Auto-fill common args
         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
 
+        # --- Check required args ---
+        required_params = [
+            p.name
+            for p in sig.parameters.values()
+            if p.name != "self"
+               and p.default is Signature.empty
+               and p.kind in (p.POSITIONAL_OR_KEYWORD, p.KEYWORD_ONLY)
+        ]
+        missing = [name for name in required_params if name not in valid_args]
+
+        # Special rule: for llamaparser we allow empty `model`
+        if provider != "llamaparser" and not model:
+            missing.append("model")
+
+        if missing:
+            example = ""
+            if provider == "llamaparser":
+                example = "Example: Parser('llamaparser:', result_type='md', mode=True, extract_tables=True)"
+            elif provider == "mistralocr":
+                example = "Example: Parser('mistralocr:ocr-large')"
+            elif provider == "huggingface":
+                example = "Example: Parser('huggingface:microsoft/trocr-base-handwritten')"
+            elif provider == "langchain":
+                example = "Example: Parser('langchain:gpt-3.5-turbo')"
+
+            raise TypeError(
+                f"{parser_cls.__name__} cannot be created because it is missing required argument(s): {', '.join(missing)}. {example}"
+            )
+
         return parser_cls(**valid_args)
 
     @abstractmethod
@@ -93,24 +181,21 @@ class Parser(BaseParser):
 
 # --- Parsers ---
 class MistralOCRParser(BaseParser, name="mistralocr"):
-    def __init__(self,provider_and_model:str) -> None:
+    def __init__(self, provider_and_model: str) -> None:
         from mistralai import Mistral
-
         self.model = provider_and_model.split(":")[1]
         self.current_cost: float = 0.0
         self.total_cost_euro: float = 0.0
-
         api_key = os.getenv("MISTRAL-OCR-API-TOKEN")
         if not api_key:
             raise EnvironmentError("Missing MISTRAL-OCR-API-TOKEN in .env file.")
-
         self.client = Mistral(api_key=api_key)
 
     def parse(self, file_path: Path) -> str:
         def upload_pdf(filename):
             uploaded_pdf = self.client.files.upload(
                 file={"file_name": filename, "content": open(filename, "rb")},
-                purpose="ocr"
+                purpose="ocr",
             )
             signed_url = self.client.files.get_signed_url(file_id=uploaded_pdf.id)
             return signed_url.url
@@ -120,14 +205,11 @@ class MistralOCRParser(BaseParser, name="mistralocr"):
             document={"type": "document_url", "document_url": upload_pdf(file_path)},
             include_image_base64=True,
         )
-
         self.current_cost = 1 / 1000 * self._count_pages(file_path)
         self.total_cost_euro += self.current_cost
-
         return "\n".join(doc.markdown for doc in ocr_response.pages)
 
     def __call__(self, file_path: Path) -> str:
-        # When instance is called, it delegates to parse
         return self.parse(file_path)
 
     @staticmethod
@@ -147,52 +229,52 @@ class LangChainParser(BaseParser, name="langchain"):
         return {"source": "LangChain", "output": response}
 
     def __call__(self, file_path: Path) -> str:
-        # When instance is called, it delegates to parse
         return self.parse(file_path)
 
 
 class LlamaParser(BaseParser, name="llamaparser"):
-    def __init__(self, result_type: str, mode: bool, provider_and_model) -> None:
+    def __init__(self, result_type: str,
+                 mode: bool,
+                 provider_and_model: str,
+                 merge_tables_across_pages_in_markdown: bool,
+                 preserve_layout_alignment_across_pages: bool) -> None:
         logging.info("Initializing LlamaParser...")
-
         self.result_type = result_type
         self.mode = mode
         self.provider_and_model = provider_and_model
+        self.merge_tables_across_pages_in_markdown = merge_tables_across_pages_in_markdown
+        self.preserve_layout_alignment_across_pages = preserve_layout_alignment_across_pages
 
         from llama_parse import LlamaParse, ResultType
-
         if result_type.lower() in ("md", "markdown"):
             result_type = ResultType.MD
-
         api_key = os.getenv("LLAMA-PARSER-API-TOKEN")
         if not api_key:
             raise EnvironmentError("Missing LLAMA-PARSER-API-TOKEN in .env file.")
-
-        self._parser = LlamaParse(api_key=api_key, result_type=result_type, premium_mode=self.mode)
+        self._parser = LlamaParse(api_key=api_key,
+                                  result_type=result_type,
+                                  premium_mode=self.mode,
+                                  merge_tables_across_pages_in_markdown=self.merge_tables_across_pages_in_markdown,
+                                  preserve_layout_alignment_across_pages=self.preserve_layout_alignment_across_pages)
 
     def parse(self, file_path: Path) -> str:
         documents = self._parser.load_data(str(file_path))
         return "\n".join(doc.text for doc in documents)
 
     def __call__(self, file_path: Path) -> str:
-        # When instance is called, it delegates to parse
         return self.parse(file_path)
 
 
 class HuggingFaceParser(BaseParser, name="huggingface"):
     def __init__(self, provider_and_model: str) -> None:
         from transformers import TrOCRProcessor, VisionEncoderDecoderModel
-
         api_key = os.getenv("HF-API-TOKEN")
         if not api_key:
             raise EnvironmentError("Missing HF-API-TOKEN in .env file.")
-
         model_name = provider_and_model.split(":")[1]
         logger.info(f"Loading Hugging Face OCR model: {model_name}")
-
         self.processor = TrOCRProcessor.from_pretrained(model_name, token=api_key)
         self.model = VisionEncoderDecoderModel.from_pretrained(model_name, token=api_key)
-
         logger.info("Model and processor loaded successfully.")
 
     def parse(self, file_path: Path) -> str:
@@ -208,9 +290,6 @@ class HuggingFaceParser(BaseParser, name="huggingface"):
             generated_ids = self.model.generate(pixel_values)
             text = self.processor.batch_decode(generated_ids, skip_special_tokens=True)[0]
             all_text += text + "\n"
-            logger.debug(f"OCR text (page {i}): {text[:100]}...")  # preview first 100 chars
-
+            logger.debug(f"OCR text (page {i}): {text[:100]}...")
         logger.info("OCR completed for all pages.")
         return all_text
-
-

{howdenparser-2.0.3 → howdenparser-3.1.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.3
 Name: HowdenParser
-Version: 2.0.3
+Version: 3.1.0
 Summary: A simple configuration manager with Pydantic and JSON export.
 License: MIT
 Keywords: config,configuration,pydantic,json

{howdenparser-2.0.3 → howdenparser-3.1.0}/pyproject.toml

@@ -14,7 +14,7 @@ build-backend = "poetry.core.masonry.api"
 
 [tool.poetry]
 name = "HowdenParser"
-version = "2.0.3"
+version = "3.1.0"
 description = "A simple configuration manager with Pydantic and JSON export."
 authors = [ "JesperThoftIllemannJ <jesper.jaeger@howdendanmark.dk>",]
 readme = "README.md"