HowdenParser 5.2.2__tar.gz → 5.2.4__tar.gz

{howdenparser-5.2.2 → howdenparser-5.2.4}/HowdenParser/parsers/llama_parser.py

@@ -1,7 +1,10 @@
 import os
+import json
 import logging
 from pathlib import Path
 from ..parser import BaseParser
+from typing import Any
+from llama_parse import ResultType
 
 class LlamaParser(BaseParser, name="llamaparser"):
     def __init__(self, result_type: str, model: str, parse_mode: str, provider_and_model: str,
@@ -43,9 +46,15 @@ class LlamaParser(BaseParser, name="llamaparser"):
         if not api_key:
             raise EnvironmentError("Missing LLAMA-PARSER-API-TOKEN in .env file.")
 
-        rt = self.result_type
-        if rt.lower() in ("md", "markdown"):
+        rt_lower = str(self.result_type).lower()
+        if rt_lower in ("md","markdown"):
             rt = ResultType.MD
+        elif rt_lower in ("txt","text"):
+            rt = ResultType.TXT
+        elif rt_lower in ("json"):
+            rt = ResultType.JSON
+        else:
+            raise NotImplementedError(f"Result type {self.result_type} is not supported.")
 
         # Construct the parser INSIDE the worker thread
         self._parser = LlamaParse(
@@ -61,31 +70,26 @@ class LlamaParser(BaseParser, name="llamaparser"):
 
         logging.info("LlamaParser initialized inside worker thread")
 
-    def parse(self, file_path: Path, include_pagenumbers: bool = False) -> str:
+    def parse(self, file_path: Path, include_pagenumbers: bool = False) -> Any:
         # Ensure parser is created inside the thread event loop
         self._lazy_init()
+        def page_break(text: str, idx: int) -> str:
+            return f"<PAGE_NUMBER {idx}>{text}</PAGE_NUMBER {idx}>"
 
-        # Standard calls
-        if not include_pagenumbers:
-            documents = self._parser.load_data(str(file_path))
-            return "\n".join(doc.text for doc in documents)
-
-        # With page numbers
         documents = self._parser.parse(file_path)
-
-        if self.result_type.lower() in ("md", "markdown"):
-            return "\n".join(
-                f"<PAGE_NUMBER {idx}>{page.md}</PAGE_NUMBER {idx}>"
-                for idx, page in enumerate(documents.pages, start=1)
-            )
-
-        return "\n".join(
-            f"<PAGE_NUMBER {idx}>{page.text}</PAGE_NUMBER {idx}>"
-            for idx, page in enumerate(documents.pages, start=1)
-        )
+        if self.result_type in ['md','markdown']:
+            if include_pagenumbers: return "\n".join(page_break(page.md, idx) for idx, page in enumerate(documents.pages, start=1))
+            else: return "\n".join(page.md for page in documents.pages)
+        elif self.result_type in ['txt','text']:
+            if include_pagenumbers: return "\n".join(page_break(page.text, idx) for idx, page in enumerate(documents.pages, start=1))
+            else: return "\n".join(page.text for page in documents.pages)
+        elif self.result_type.lower() == 'json':
+            return documents.model_dump()
+        else:
+            raise NotImplementedError(f"Result type {self.result_type} is not supported.")
 
     def __call__(self, file_path: Path) -> str:
         return self.parse(file_path)
 
     def write_json_hyperparameter(self, folder_file_path: Path) -> None:
-        self.write_parameters(self._input_params, folder_file_path)
+        self.write_parameters(self._input_params, folder_file_path)

{howdenparser-5.2.2 → howdenparser-5.2.4}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: HowdenParser
-Version: 5.2.2
+Version: 5.2.4
 Summary: A simple configuration manager with Pydantic and JSON export.
 License: MIT
 Keywords: config,configuration,pydantic,json
@@ -16,6 +16,7 @@ Requires-Dist: llama-parse (>=0.6.58,<0.7.0)
 Requires-Dist: mistralai (>=1.9.3,<2.0.0)
 Requires-Dist: pdf2image (>=1.17.0,<2.0.0)
 Requires-Dist: pypdf2 (>=3.0.1,<4.0.0)
+Requires-Dist: pytest (>=9.0.2,<10.0.0)
 Description-Content-Type: text/markdown
 
 # OCR & LLM Parser

{howdenparser-5.2.2 → howdenparser-5.2.4}/pyproject.toml

@@ -1,41 +1,42 @@
-[project]
-name = "HowdenParser"
-version = "5.2.2"
-description = "A simple configuration manager with Pydantic and JSON export."
-readme = "README.md"
-requires-python = ">=3.12,<3.14"
-authors = [
-    { name = "JesperThoftIllemannJ", email = "jesper.jaeger@howdendanmark.dk" },
-]
-keywords = [
-    "config",
-    "configuration",
-    "pydantic",
-    "json",
-]
-homepage = "https://github.com/yourusername/config"
-repository = "https://github.com/yourusername/config"
-documentation = "https://github.com/yourusername/config"
-dependencies = [
-    "mistralai (>=1.9.3,<2.0.0)",
-    "llama-parse (>=0.6.58,<0.7.0)",
-    "pypdf2 (>=3.0.1,<4.0.0)",
-    "pdf2image (>=1.17.0,<2.0.0)",
-    "langchain (>=1.1.3,<2.0.0)",
-]
-
-[project.license]
-text = "MIT"
-
-[build-system]
-requires = [
-    "poetry-core>=2.0.0,<3.0.0",
-]
-build-backend = "poetry.core.masonry.api"
-
-[dependency-groups]
-dev = [
-    "toml (>=0.10.2,<0.11.0)",
-    "tomli-w (>=1.2.0,<2.0.0)",
-    "howdenconfig (>=1.0.6,<2.0.0)",
-]
+[project]
+name = "HowdenParser"
+version = "5.2.4"
+description = "A simple configuration manager with Pydantic and JSON export."
+readme = "README.md"
+requires-python = ">=3.12,<3.14"
+authors = [
+    { name = "JesperThoftIllemannJ", email = "jesper.jaeger@howdendanmark.dk" },
+]
+keywords = [
+    "config",
+    "configuration",
+    "pydantic",
+    "json",
+]
+homepage = "https://github.com/yourusername/config"
+repository = "https://github.com/yourusername/config"
+documentation = "https://github.com/yourusername/config"
+dependencies = [
+    "mistralai (>=1.9.3,<2.0.0)",
+    "llama-parse (>=0.6.58,<0.7.0)",
+    "pypdf2 (>=3.0.1,<4.0.0)",
+    "pdf2image (>=1.17.0,<2.0.0)",
+    "langchain (>=1.1.3,<2.0.0)",
+    "pytest (>=9.0.2,<10.0.0)",
+]
+
+[project.license]
+text = "MIT"
+
+[build-system]
+requires = [
+    "poetry-core>=2.0.0,<3.0.0",
+]
+build-backend = "poetry.core.masonry.api"
+
+[dependency-groups]
+dev = [
+    "toml (>=0.10.2,<0.11.0)",
+    "tomli-w (>=1.2.0,<2.0.0)",
+    "howdenconfig (>=1.0.6,<2.0.0)",
+]