Update app.py

app.py

@@ -27,7 +27,6 @@ import io
 import requests
 import numpy as np
 from urllib.parse import quote
-import PyPDF2
 
 # =============================================================================
 # ───────────── EXTERNAL HELP LINKS ─────────────
@@ -117,16 +116,8 @@ def preprocess_text(text):
     return text.strip()
 
 def sanitize_json_text(text):
-    text = text.encode().decode('unicode_escape')
-    text = json.dumps(text)[1:-1]
-    return text
-
-def extract_pdf_text(pdf_file):
-    pdf_reader = PyPDF2.PdfReader(pdf_file)
-    text = ""
-    for page in pdf_reader.pages:
-        text += page.extract_text() or ""
-    return text
+    text = re.sub(r'[\x00-\x08\x0B\x0C\x0E-\x1F]', '', text)
+    return text.strip()
 
 # =============================================================================
 # ───────────── COSMOS DB FUNCTIONS ─────────────
@@ -168,13 +159,10 @@ def delete_record(container, record):
         container.delete_item(item=doc_id, partition_key=partition_key_value)
         return True, f"Record {doc_id} deleted. 🗑️"
     except exceptions.CosmosResourceNotFoundError:
-        st.write(f"Record {doc_id} not found in Cosmos DB - treating as success.")
         return True, f"Record {doc_id} not found (already deleted). 🗑️"
     except exceptions.CosmosHttpResponseError as e:
-        st.error(f"Cosmos HTTP error deleting {doc_id}: {str(e)}")
         return False, f"HTTP error deleting {doc_id}: {str(e)} 🚨"
     except Exception as e:
-        st.error(f"Unexpected error deleting {doc_id}: {str(e)}")
         return False, f"Unexpected error deleting {doc_id}: {str(e)} 😱"
 
 def save_to_cosmos_db(container, query, response1, response2):
@@ -376,38 +364,19 @@ def edit_all_documents(container, search_keyword=None):
         formatted_ts = dt.strftime("%I:%M %p %m/%d/%Y")
         header = f"{doc.get('name', 'Unnamed')} - {formatted_ts}"
         with st.expander(header):
-            col_json, col_pdf = st.columns([2, 1])
-            with col_json:
-                doc_key = f"editor_{doc['id']}"
-                initial_value = st.session_state.saved_docs.get(doc['id'], json.dumps(doc, indent=2))
-                edited_content = st.text_area("Edit JSON", value=initial_value, height=300, key=doc_key)
-            with col_pdf:
-                if 'pdf_data' in doc:
-                    st.markdown("### 📜 PDF Preview")
-                    pdf_bytes = base64.b64decode(doc['pdf_data'])
-                    st.download_button(
-                        label="⬇️ Download PDF",
-                        data=pdf_bytes,
-                        file_name=f"{doc.get('name', 'document')}.pdf",
-                        mime="application/pdf"
-                    )
-            
+            doc_key = f"editor_{doc['id']}"
+            initial_value = st.session_state.saved_docs.get(doc['id'], json.dumps(doc, indent=2))
+            edited_content = st.text_area("Edit JSON", value=initial_value, height=300, key=doc_key)
             col_save, col_delete = st.columns(2)
             with col_save:
                 if st.button("💾 Save", key=f"save_{doc['id']}"):
                     try:
-                        cleaned_content = edited_content.strip()
-                        if not (cleaned_content.startswith('{') and cleaned_content.endswith('}')):
-                            sanitized_text = sanitize_json_text(cleaned_content)
-                            updated_doc = doc.copy()
-                            updated_doc['content'] = sanitized_text
-                        else:
-                            updated_doc = json.loads(cleaned_content)
-                            updated_doc['id'] = doc['id']
-                            updated_doc['pk'] = doc.get('pk', doc['id'])
-                            for field in ['_ts', '_rid', '_self', '_etag', '_attachments']:
-                                updated_doc.pop(field, None)
-                        
+                        cleaned_content = sanitize_json_text(edited_content)
+                        updated_doc = json.loads(cleaned_content)
+                        updated_doc['id'] = doc['id']
+                        updated_doc['pk'] = doc.get('pk', doc['id'])
+                        for field in ['_ts', '_rid', '_self', '_etag', '_attachments']:
+                            updated_doc.pop(field, None)
                         success, message = update_record(container, updated_doc)
                         if success:
                             st.success(f"Saved {doc['id']}")
@@ -447,27 +416,6 @@ def new_item_default(container):
     else:
         st.error(f"Error creating new item: {message}")
 
-def new_item_from_pdf(container, pdf_file):
-    new_id = generate_unique_id()
-    pdf_bytes = pdf_file.read()
-    pdf_text = extract_pdf_text(io.BytesIO(pdf_bytes))
-    pdf_base64 = base64.b64encode(pdf_bytes).decode('utf-8')
-    default_doc = {
-        "id": new_id,
-        "pk": new_id,
-        "name": pdf_file.name,
-        "content": pdf_text[:1000],
-        "timestamp": datetime.now().isoformat(),
-        "type": "pdf_document",
-        "pdf_data": pdf_base64
-    }
-    success, message = insert_record(container, default_doc)
-    if success:
-        st.success(f"PDF document '{pdf_file.name}' created! ✨")
-        st.rerun()
-    else:
-        st.error(f"Error creating PDF item: {message}")
-
 def add_field_to_doc():
     key = st.session_state.new_field_key
     value = st.session_state.new_field_value
@@ -531,25 +479,20 @@ def vector_keyword_search(keyword, doc):
     return False
 
 def search_documents_ui(container):
-    st.sidebar.subheader("🔍 Vector Search")
     with st.sidebar.form("search_form"):
-        keyword = st.text_input("Search Documents", key="search_keyword")
+        keyword = st.text_input("Search Keyword", key="search_keyword")
         col1, col2 = st.columns(2)
         with col1:
             search_submitted = st.form_submit_button("🔍 Search")
         with col2:
             clear_submitted = st.form_submit_button("🗑️ Clear")
         if search_submitted and keyword:
-            st.session_state.active_search = keyword
+            st.session_state.active_search = keyword  # Use a separate key
             st.rerun()
         if clear_submitted:
             if 'active_search' in st.session_state:
                 del st.session_state.active_search
             st.rerun()
-    
-    uploaded_file = st.sidebar.file_uploader("Upload Document (PDF)", type=["pdf"])
-    if uploaded_file and container:
-        new_item_from_pdf(container, uploaded_file)
 
 def validate_and_preprocess_image(file_data, target_size=(576, 1024)):
     try:
@@ -668,10 +611,6 @@ def main():
     # Sidebar: Hierarchical Navigation
     st.sidebar.title("🐙 Navigator")
     
-    # Vector Search Section (Moved to Top)
-    if st.session_state.current_container:
-        search_documents_ui(st.session_state.current_container)
-    
     # Databases Section
     st.sidebar.subheader("🗃️ Databases")
     if "client" not in st.session_state:
@@ -728,6 +667,7 @@ def main():
                 new_ai_record(st.session_state.current_container)
             if st.sidebar.button("🔗 New Links Record"):
                 new_links_record(st.session_state.current_container)
+            search_documents_ui(st.session_state.current_container)
 
     # Central Area: Editable Documents with Search Filter
     if st.session_state.current_container: