Initial commit

app.py

@@ -0,0 +1,84 @@
+import streamlit as st
+from decimal import Decimal
+from carbon import Carbon
+import requests
+import json
+
+# Carbon API Key
+CARBON_API_KEY = "a38ee1fe5fef56fc8e1ae2afc881378804bb902882442e1554adae4f82ee23ea"
+customer_id = "Candid"
+
+# Define your main function here
+def main():
+    st.title('Google Drive Data Connector using Carbon SDK')
+    
+    st.write('### Authenticate with Carbon API')
+    carbon = Carbon(api_key=CARBON_API_KEY, customer_id=customer_id)
+    token = carbon.auth.get_access_token()
+    carbon = Carbon(access_token=token.access_token)  # authenticated object
+
+    st.write('### Connect to Google Drive')
+    get_oauth_url_response = carbon.integrations.get_oauth_url(
+        service="GOOGLE_DRIVE",
+        tags=None,
+        scope="string_example",
+        chunk_size=Decimal(1500),
+        chunk_overlap=Decimal(15),
+        skip_embedding_generation=False,
+        embedding_model="OPENAI",
+        generate_sparse_vectors=False,
+        prepend_filename_to_chunks=False,
+        max_items_per_chunk=Decimal(1),
+        salesforce_domain="string_example",
+        sync_files_on_connection=True,
+        set_page_as_boundary=False,
+        data_source_id=Decimal(1),
+        connecting_new_account=False,
+        request_id="b7620173-662c-4ae7-bb61-2e6ffd8619f5",
+        use_ocr=False,
+        parse_pdf_tables_with_ocr=False,
+        enable_file_picker=True,
+        sync_source_items=True,
+        incremental_sync=False,
+        file_sync_config={
+            "auto_synced_source_types": ["ARTICLE"],
+            "sync_attachments": False,
+            "detect_audio_language": False,
+        },
+    )
+    
+    st.write(f"OAuth URL: {get_oauth_url_response.oauth_url}")
+    
+    st.write('### Search in the Connected Data Source')
+    query = st.text_input("Enter your query:", value="Is there a web version of CNN explainer?")
+    if query:
+        url = "https://api.carbon.ai/embeddings"
+        payload = {
+            "query": query,
+            "k": 2,
+            "file_ids": [1],  # Dummy file_id for demonstration
+            "include_all_children": True,
+            "tags": {},
+            "include_tags": True,
+            "include_vectors": True,
+            "include_raw_file": True,
+            "hybrid_search": False,
+            "media_type": "TEXT",
+            "embedding_model": "OPENAI"
+        }
+        headers = {
+            "authorization": f"Bearer {CARBON_API_KEY}",
+            "customer-id": customer_id,
+            "Content-Type": "application/json"
+        }
+        response_search = requests.post(url, json=payload, headers=headers)
+        response_search_chunks = json.loads(response_search.text)
+        
+        for i, doc in enumerate(response_search_chunks['documents']):
+            st.write(f"Document {i+1}")
+            st.write(doc['content'])
+            st.write("-------------------------------------------------")
+
+# Call the main function
+if __name__ == '__main__':
+    main()

requirements.txt

@@ -0,0 +1,3 @@
+streamlit
+carbon-python-sdk
+requests