Merge pull request #152 from enoch3712/151-tesseract-only-extract-from-png-images-and-not-jpeg

151 tesseract only extract from png images and not jpeg

Author: enoch3712

extract_thinker/document_loader/document_loader_tesseract.py

@@ -15,7 +15,7 @@
 class DocumentLoaderTesseract(CachedDocumentLoader):
     """Document loader for OCR using Tesseract."""
 
-    SUPPORTED_FORMATS = ["jpeg", "png", "bmp", "tiff", "pdf"]
+    SUPPORTED_FORMATS = ["jpeg", "png", "bmp", "tiff", "pdf", "jpg"]
 
     def __init__(self, tesseract_cmd, isContainer=False, content=None, cache_ttl=300):
         super().__init__(content, cache_ttl)

pyproject.toml

@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "extract_thinker"
-version = "0.0.30"
+version = "0.0.31"
 description = "Library to extract data from files and documents agnositicaly using LLMs"
 authors = ["Júlio Almeida <[email protected]>"]
 readme = "README.md"