HowdenParser 0.1.11__tar.gz → 0.1.13__tar.gz

howdenparser-0.1.13/HowdenParser/parameter/mistralocr.py

@@ -0,0 +1,5 @@
+from typing import Literal
+from pydantic import BaseModel
+
+class Parameter(BaseModel):
+    provider_and_model: Literal["mistralocr:mistral-ocr-latest"] = "mistralocr:mistral-ocr-latest"

{howdenparser-0.1.11 → howdenparser-0.1.13}/HowdenParser/parser.py

@@ -2,6 +2,8 @@ from abc import ABC, abstractmethod
 from dotenv import load_dotenv
 import os
 from pathlib import Path
+import logging
+from PyPDF2 import PdfReader
 
 load_dotenv()
 
@@ -12,14 +14,17 @@ class BaseParser(ABC):
 
 
 class MistralOCRParser(BaseParser):
-    def __init__(self, result_type: str, **kwargs) -> None:
+    def __init__(self, model: str) -> None:
         from mistralai import Mistral
+        self.model = model
+        self.current_cost: float = 0.0
+        self.total_cost_euro: float = 0.0
+
         name = "MISTRAL-OCR-API-TOKEN"
         api_key = os.getenv(name)
         if not api_key:
             raise EnvironmentError(f"Missing {name} in .env file.")
 
-        self.result_type = "md" if result_type.lower() in ("md", "markdown") else "text"
         self.client = Mistral(api_key=api_key)
 
     def parse(self, file_path: Path) -> str:
@@ -35,7 +40,7 @@ class MistralOCRParser(BaseParser):
             return signed_url.url
 
         ocr_response = self.client.ocr.process(
-            model="mistral-ocr-latest",
+            model=self.model,
             document={
             "type": "document_url",
             "document_url": upload_pdf(file_path),
@@ -43,14 +48,23 @@ class MistralOCRParser(BaseParser):
             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)
 
+    @staticmethod
+    def _count_pages(file_path: Path) -> int:
+        reader = PdfReader(str(file_path))
+        return len(reader.pages)
+
+
 
 class LangChainParser(BaseParser):
-    def __init__(self, model_name: str):
+    def __init__(self, model: str):
         from langchain.llms import OpenAI
-        self.model_name = model_name
-        self.model = OpenAI(model_name=model_name)
+        self.model_name = model
+        self.model = OpenAI(model_name=model)
 
     def parse(self, text: str) -> dict:
         response = self.model(text)
@@ -58,27 +72,37 @@ class LangChainParser(BaseParser):
 
 
 class LlamaParser(BaseParser):
-    def __init__(self, result_type: str, mode: bool, **kwargs) -> None:
+    def __init__(self, result_type: str, mode: bool) -> None:
+        logging.info("Initializing LlamaParser...")
+        logging.info("Loading LlamaParse package...")
+
         from llama_parse import LlamaParse, ResultType
+
         if result_type.lower() == "md" or result_type.lower() == "markdown":
             self.result_type = ResultType.MD
+
         name = "LLAMA-PARSER-API-TOKEN"
-        self.api_key = os.getenv(name)
-        if not self.api_key:
+        api_key = os.getenv(name)
+        if not api_key:
             raise EnvironmentError(f"Missing {name} in .env file.")
 
+        logging.info("Initializing LlamaParse parser...")
         self.parser = LlamaParse(
-            api_key=self.api_key,
+            api_key=api_key,
             result_type=self.result_type,
             premium_mode=mode
         )
+        logging.info("LlamaParser initialized successfully.")
 
     def parse(self, file_path: Path) -> str:
+        logging.info(f"Parsing file: {file_path}")
         documents = self.parser.load_data(str(file_path))
-        return "\n".join(doc.text for doc in documents)
+        text = "\n".join(doc.text for doc in documents)
+        logging.info(f"Parsing completed. Extracted {len(documents)} documents.")
+        return text
 
 class HuggingFaceParser(BaseParser):
-    def __init__(self, result_type: str, mode: bool, **kwargs) -> None:
+    def __init__(self, model: str, result_type: str) -> None:
         from transformers import pipeline
 
         if result_type.lower() in ("md", "markdown"):
@@ -91,16 +115,15 @@ class HuggingFaceParser(BaseParser):
         if not self.api_key:
             raise EnvironmentError(f"Missing {name} in .env file.")
 
-        # OCR + text extraction pipeline
         # Example model: microsoft/layoutlmv3-base-finetuned-docvqa
         self.parser = pipeline(
             task="document-question-answering",
-            model="impira/layoutlm-document-qa",
+            model=model,
             use_auth_token=self.api_key
         )
 
     def parse(self, file_path: Path) -> str:
-        import fitz  # PyMuPDF for PDF → images
+        import fitz
 
         pdf_doc = fitz.open(file_path)
         output_parts = []
@@ -113,7 +136,6 @@ class HuggingFaceParser(BaseParser):
                 output_parts.append(response[0]["answer"])
 
         if self.result_type == "markdown":
-            # Here you could add rules to format into markdown
             return "\n\n".join(output_parts)
         else:
             return " ".join(output_parts)
@@ -121,17 +143,17 @@ class HuggingFaceParser(BaseParser):
 # === Step 3: Dynamic factory using string input ===
 class ParserFactory:
     @staticmethod
-    def get_parser(parser_type: str, **kwargs) -> BaseParser:
-        provider = parser_type.partition(":")[0].lower()
-        model = parser_type.partition(":")[2]
+    def get_parser(provider_model: str, **kwargs) -> BaseParser:
+        provider = provider_model.partition(":")[0].lower()
+        model = provider_model.partition(":")[2]
         if provider == "langchain":
-            return LangChainParser(model_name=model)
+            return LangChainParser(model=model)
         elif provider == "llamaparser":
-            return LlamaParser(**kwargs)
+            return LlamaParser(kwargs["result_type"], kwargs["mode"])
         elif provider == "huggingface":
-            return HuggingFaceParser(model_name=model, **kwargs)
+            return HuggingFaceParser(model=model, **kwargs)
         elif provider == "mistralocr":
-            return MistralOCRParser(**kwargs)
+            return MistralOCRParser(model)
         else:
-            raise ValueError(f"Unknown parser type: {parser_type}")
+            raise ValueError(f"Unknown parser type: {provider_model}")
 

{howdenparser-0.1.11 → howdenparser-0.1.13}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.3
 Name: HowdenParser
-Version: 0.1.11
+Version: 0.1.13
 Summary: A simple configuration manager with Pydantic and JSON export.
 License: MIT
 Keywords: config,configuration,pydantic,json
@@ -12,9 +12,11 @@ Classifier: Programming Language :: Python :: 3
 Classifier: Programming Language :: Python :: 3.12
 Classifier: Programming Language :: Python :: 3.13
 Requires-Dist: fitz (>=0.0.1.dev2,<0.0.2)
+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)
+Requires-Dist: pypdf2 (>=3.0.1,<4.0.0)
 Requires-Dist: transformers (>=4.55.2,<5.0.0)
 Project-URL: Documentation, https://github.com/yourusername/config
 Project-URL: Homepage, https://github.com/yourusername/config

{howdenparser-0.1.11 → howdenparser-0.1.13}/pyproject.toml

@@ -3,7 +3,7 @@ name = "HowdenParser"
 description = ""
 readme = "README.md"
 requires-python = ">=3.12,<4.0"
-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)",]
+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)",]
 [[project.authors]]
 name = "JesperThoftIllemannJ"
 email = "jesper.jaeger@howdendanmark.dk"
@@ -14,7 +14,7 @@ build-backend = "poetry.core.masonry.api"
 
 [tool.poetry]
 name = "HowdenParser"
-version = "0.1.11"
+version = "0.1.13"
 description = "A simple configuration manager with Pydantic and JSON export."
 authors = [ "JesperThoftIllemannJ <jesper.jaeger@howdendanmark.dk>",]
 readme = "README.md"

howdenparser-0.1.11/HowdenParser/parameter/mistralocr.py

@@ -1,7 +0,0 @@
-from typing import Literal
-from pydantic import BaseModel
-
-class Parameter(BaseModel):
-    model: str = "mistralocr:"
-    result_type: Literal["md"] = "md"
-    mode: bool = False