Create app.py

app.py

@@ -0,0 +1,89 @@
+import streamlit as st
+from pdf2image import convert_from_path
+import pytesseract
+import os
+from pathlib import Path
+from sentence_transformers import SentenceTransformer
+import pinecone
+import tempfile
+import shutil
+
+# Dynamically add Poppler to PATH if necessary
+os.environ["PATH"] += os.pathsep + "C:\\Program Files\\poppler-24.08.0\\bin"
+
+# Initialize Streamlit app
+st.title("PDF Image to Text/Word Converter with Search Capability")
+st.write("Upload your PDF to extract text or Word document and search content within it.")
+
+# Create a temporary directory
+temp_dir = tempfile.mkdtemp()
+
+# Upload PDF file
+uploaded_file = st.file_uploader("Choose a PDF file", type="pdf")
+
+if uploaded_file:
+    pdf_path = Path(temp_dir) / uploaded_file.name
+    with open(pdf_path, "wb") as f:
+        f.write(uploaded_file.read())
+
+    st.write("File uploaded successfully!")
+
+    # Convert PDF pages to images
+    st.write("Converting PDF to images...")
+    images = convert_from_path(pdf_path, output_folder=temp_dir)
+
+    # Extract text from images
+    st.write("Extracting text from images...")
+    extracted_text = ""
+    for idx, image in enumerate(images):
+        st.image(image, caption=f"Page {idx + 1}", use_column_width=True)
+        text = pytesseract.image_to_string(image)
+        extracted_text += text + "\n"
+
+    # Save extracted text to a .txt file
+    text_file_path = Path(temp_dir) / "extracted_text.txt"
+    with open(text_file_path, "w", encoding="utf-8") as text_file:
+        text_file.write(extracted_text)
+
+    st.success("Text extraction complete!")
+
+    # Option to download text file
+    st.download_button(
+        label="Download Extracted Text",
+        data=extracted_text,
+        file_name="extracted_text.txt",
+        mime="text/plain",
+    )
+
+    # Initialize vector model and Pinecone
+    st.write("Initializing vector search...")
+    model = SentenceTransformer('all-MiniLM-L6-v2')
+    pinecone.init(api_key="YOUR_PINECONE_API_KEY", environment="us-west1-gcp")  # Replace with your Pinecone details
+
+    # Create Pinecone index if it doesn't exist
+    index_name = "pdf-text-index"
+    if index_name not in pinecone.list_indexes():
+        pinecone.create_index(index_name, dimension=384)
+    index = pinecone.Index(index_name)
+
+    # Generate vector embeddings and upload to Pinecone
+    st.write("Generating vector embeddings...")
+    sentences = extracted_text.split("\n")
+    embeddings = model.encode(sentences)
+    for i, embedding in enumerate(embeddings):
+        index.upsert([(f"sentence-{i}", embedding, {"sentence": sentences[i]})])
+
+    # Search functionality
+    st.write("Search within the extracted text")
+    query = st.text_input("Enter your query:")
+    if query:
+        query_vector = model.encode([query])[0]
+        results = index.query(query_vector, top_k=5, include_metadata=True)
+        st.write("Top results:")
+        for match in results["matches"]:
+            st.write(f"Score: {match['score']}, Text: {match['metadata']['sentence']}")
+
+# Cleanup temporary files
+if st.button("Clean Temporary Files"):
+    shutil.rmtree(temp_dir)
+    st.success("Temporary files cleaned!")