Update app1.py

app1.py

@@ -8,10 +8,42 @@ async()=>{
 }
 """
 
+def get_documents():
+    return PyPDFLoader("AI Guide for Government - AI CoE.pdf","Ethics_of_Artificial_Intelligence-2.pdf","IPOL_BRI(2016)571380_EN.pdf").load()
+    #17357182991031590738file.pdf
+
+def extract_pdfs(x, request: gr.Request, progress=gr.Progress()):
+    progress(0, desc="Test", unit = "Files")
+    print("request", request)
+
+    # Delete existing index directory and recreate the directory
+    if os.path.exists(DB_DIR):
+        import shutil
+        shutil.rmtree(DB_DIR, ignore_errors=True)
+        os.mkdir(DB_DIR)
+    
+    documents = []
+    all_text = ""
+    for num, doc in enumerate(progress.tqdm(get_documents())):
+        doc.page_content = replace_newlines_and_spaces(doc.page_content)
+        documents.append(doc)
+        all_text += doc.page_content
+        time.sleep(0.1)
+
+    return documents, all_text
+
+def replace_newlines_and_spaces(text):
+    # Replace all newline characters with spaces
+    text = text.replace("\n", " ")
+    # Replace multiple spaces with a single space
+    text = re.sub(r'\s+', ' ', text)
+    return text
+
+
 def test(x, request: gr.Request, progress=gr.Progress()):
     progress(0, desc="Test", unit = "Files")
-    print("request", request)    
-    a = "abcdefghijklmnopqrstuv"    
+    print("request", request)
+    a = "abcdefghijklmnopqrstuv"
     for letter in progress.tqdm(a, desc = "TEST", unit = "Files"):
         time.sleep(0.1)
     return a  
@@ -31,6 +63,6 @@ with gr.Blocks() as demo:
     
     b.click(test, selected, prog)
     
-    demo.load(test, inputs=None, outputs=prog) #, _js=on_load)    
+    demo.load(extract_pdfs, inputs=None, outputs=[prog, selected]) #, _js=on_load)    
     
 demo.launch()