Merge pull request #170 from enoch3712/169-add-markitdown-to-init

add markdown DL

docs/core-concepts/document-loaders/markitdown.md

@@ -1,6 +1,6 @@
 # MarkItDown Document Loader
 
-MarkItDown is a versatile document processing library that can handle multiple file formats. ExtractThinker's MarkItDown loader provides a robust interface for text extraction with optional vision mode support.
+[MarkItDown](https://github.com/microsoft/markitdown) is a versatile document processing library from Microsoft that can handle multiple file formats. ExtractThinker's MarkItDown loader provides a robust interface for text extraction with optional vision mode support.
 
 ## Basic Usage
 

extract_thinker/__init__.py

@@ -9,6 +9,7 @@
 from .document_loader.document_loader_pypdf import DocumentLoaderPyPdf
 from .document_loader.document_loader_pdfplumber import DocumentLoaderPdfPlumber
 from .document_loader.document_loader_beautiful_soup import DocumentLoaderBeautifulSoup
+from .document_loader.document_loader_markitdown import DocumentLoaderMarkItDown
 from .models.classification import Classification
 from .models.classification_response import ClassificationResponse
 from .process import Process
@@ -44,6 +45,7 @@
     'DocumentLoaderAWSTextract',
     'DocumentLoaderGoogleDocumentAI',
     'DocumentLoaderDocumentAI',
+	'DocumentLoaderMarkItDown',
     'Classification',
     'ClassificationResponse',
     'Process',

extract_thinker/document_loader/document_loader_markitdown.py

@@ -8,11 +8,6 @@
 from extract_thinker.document_loader.cached_document_loader import CachedDocumentLoader
 from extract_thinker.utils import MIME_TYPE_MAPPING
 
-try:
-    from markitdown import MarkItDown
-except ImportError:
-    raise ImportError("MarkItDown library is not installed. Please install it with 'pip install markitdown'.")
-
 class DocumentLoaderMarkItDown(CachedDocumentLoader):
     """
     Document loader that uses MarkItDown to extract content from various file formats.
@@ -26,8 +21,32 @@ class DocumentLoaderMarkItDown(CachedDocumentLoader):
     ]
 
     def __init__(self, content: Any = None, cache_ttl: int = 300, llm_client=None, llm_model=None):
+        # Check dependencies before initializing
+        self._check_dependencies()
         super().__init__(content, cache_ttl)
-        self.markitdown = MarkItDown(llm_client=llm_client, llm_model=llm_model)
+        self.markitdown = self._get_markitdown()(llm_client=llm_client, llm_model=llm_model)
+
+    @staticmethod
+    def _check_dependencies():
+        """Check if required dependencies are installed."""
+        try:
+            import markitdown
+        except ImportError:
+            raise ImportError(
+                "Could not import markitdown package. "
+                "Please install it with `pip install markitdown`."
+            )
+
+    def _get_markitdown(self):
+        """Lazy load MarkItDown."""
+        try:
+            from markitdown import MarkItDown
+            return MarkItDown
+        except ImportError:
+            raise ImportError(
+                "Could not import markitdown python package. "
+                "Please install it with `pip install markitdown`."
+            )
 
     @cachedmethod(cache=attrgetter('cache'), 
                   key=lambda self, source: hashkey(source if isinstance(source, str) else source.getvalue(), self.vision_mode))

pyproject.toml

@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "extract_thinker"
-version = "0.0.32"
+version = "0.0.33"
 description = "Library to extract data from files and documents agnositicaly using LLMs"
 authors = ["Júlio Almeida <[email protected]>"]
 readme = "README.md"