tst1

app.py

@@ -28,6 +28,25 @@ docs = st.sidebar.file_uploader("Upload documents", accept_multiple_files=True,
 query = st.text_input("Enter search query")
 click = st.button("Search")
 
+def extract_text(doc):
+    if doc.type == 'text/plain':
+        return doc.getvalue().decode("utf-8")
+        
+    if doc.type == "application/pdf":
+        with pdfplumber.open(doc) as pdf:
+            pages = [page.extract_text() for page in pdf.pages]
+            return "\n".join(pages)
+        
+    if doc.type == "application/vnd.openxmlformats-officedocument.wordprocessingml.document":
+        return docx2txt.process(doc)
+        
+    if doc.name.endswith(".xlsx"):
+        text = textract.process(doc)
+        return text.decode("utf-8")
+
+    return None
+    
+
 if click and query:
     doc_contents = []
     
@@ -46,22 +65,5 @@ if click and query:
     for doc, score in ranked_docs:
         st.write(f"{doc.name} (score: {score:.2f})")
         
-def extract_text(doc):
-    if doc.type == 'text/plain':
-        return doc.getvalue().decode("utf-8")
-        
-    if doc.type == "application/pdf":
-        with pdfplumber.open(doc) as pdf:
-            pages = [page.extract_text() for page in pdf.pages]
-            return "\n".join(pages)
-        
-    if doc.type == "application/vnd.openxmlformats-officedocument.wordprocessingml.document":
-        return docx2txt.process(doc)
-        
-    if doc.name.endswith(".xlsx"):
-        text = textract.process(doc)
-        return text.decode("utf-8")
 
-    return None
-