Upload folder using huggingface_hub

app.py

@@ -1,7 +1,6 @@
 import gradio as gr
 from gradio_client import Client, handle_file
 import re
-from thefuzz import fuzz
 
 # hugging face clients for both OCR options
 surya_ocr_client = Client("artificialguybr/Surya-OCR")
@@ -30,7 +29,7 @@ def ocr_extraction(image, ocr_model):
             result = client.predict(
                 image=handle_file(image),
                 got_mode="plain texts OCR",
-				fine_grained_mode="box",
+                fine_grained_mode="box",
                 ocr_color="red",
                 ocr_box="Hello!!",
                 api_name="/run_GOT"
@@ -43,29 +42,19 @@ def ocr_extraction(image, ocr_model):
     except Exception as e:
         return f"An error occurred: {str(e)}"
 
-def search_keyword(keyword, search_type):
+def search_keyword(keyword):
     global extracted_text
     if not extracted_text:
         return "No OCR text found. Please extract text from an image first."
     if not keyword:
         return extracted_text
     
-    if search_type == "Direct Search":
-        highlighted_text = re.sub(f"({re.escape(keyword)})", r'<span style="background-color: yellow;">\1</span>', extracted_text, flags=re.IGNORECASE)
-    else:  # Nearest Search
-        words = extracted_text.split()
-        highlighted_words = []
-        for word in words:
-            if fuzz.ratio(word.lower(), keyword.lower()) >= 80:  # Adjust threshold as needed
-                highlighted_words.append(f'<span style="background-color: yellow;">{word}</span>')
-            else:
-                highlighted_words.append(word)
-        highlighted_text = " ".join(highlighted_words)
+    highlighted_text = re.sub(f"({re.escape(keyword)})", r'<span style="background-color: yellow;">\1</span>', extracted_text, flags=re.IGNORECASE)
     
     return highlighted_text
 
 with gr.Blocks(theme=gr.themes.Soft()) as gr_interface:
-    gr.Markdown("# 📷 OCR Text Extraction and Advanced Keyword Search 🔍")
+    gr.Markdown("# 📷 OCR Text Extraction and Direct Keyword Search 🔍")
     
     with gr.Row():
         with gr.Column(scale=1):
@@ -87,7 +76,6 @@ with gr.Blocks(theme=gr.themes.Soft()) as gr_interface:
     with gr.Row():
         with gr.Column(scale=1):
             keyword_input = gr.Textbox(label="Enter keyword to search")
-            search_type = gr.Radio(["Direct Search", "Nearest Search"], label="Search Type", value="Direct Search")
             search_button = gr.Button("Search Keyword", variant="secondary")
         
         with gr.Column(scale=2):
@@ -101,8 +89,8 @@ with gr.Blocks(theme=gr.themes.Soft()) as gr_interface:
     
     search_button.click(
         fn=search_keyword,
-        inputs=[keyword_input, search_type],
+        inputs=[keyword_input],
         outputs=highlighted_output
     )
 
-gr_interface.launch(share=True)
+gr_interface.launch(share=True)

requirements.txt

@@ -0,0 +1 @@
+thefuzz