Update app.py

app.py

@@ -1,3 +1,5 @@
+import fitz
+import PyPDF2
 import numpy as np
 import gradio as gr
 import json
@@ -161,17 +163,38 @@ def upload_file(org, bot , files):
     file_paths = [file.name for file in files]
     allfiles=file_paths
     path = os.path.join(os.getcwd()+'\\Organizations\\', org,bot,'Inputs')
-    shutil.copy(file_paths.name, path)
+    os.makedirs(path, exist_ok=True)
+    
+    for file_path in file_paths:
+        shutil.copy(file_path, path)
     return file_paths
     
+
 def train(files):
     for file in files:
         print(file.name)
         if file.name.endswith(".pdf"):
-            loader = PyPDFLoader(file.name)
-            documents = loader.load()
+            pdf_document = fitz.open(file.name) 
+            documents = []  
+            for page_num in range(pdf_document.page_count):
+                page = pdf_document.load_page(page_num)
+                text = page.get_text()
+                documents.append(text)  
+
+            
+            output_folder = "Output"  
+            os.makedirs(output_folder, exist_ok=True)  
+            output_file_name = os.path.splitext(os.path.basename(file.name))[0]  
+            output_file_path = os.path.join(output_folder, f"{output_file_name}_page_{page_num + 1}.txt")
+
+            with open(output_file_path, "w", encoding="utf-8") as output_file:
+                for text in documents:
+                    output_file.write(text)
+
     return "Training Done"
 chat_history = []
+documents = []
+documents = construct_index(os.path.join(os.getcwd(), 'Organizations'))
 
 def construct_index(directory_path):
     file_paths=[]
@@ -179,6 +202,7 @@ def construct_index(directory_path):
         for file_name in files:
             file_path = os.path.join(root, file_name)
             file_paths.append(file_path)
+            
 
 
     for file in file_paths:        
@@ -190,6 +214,8 @@ def construct_index(directory_path):
 
 def chatbot(input_text):
     global chat_history
+    global documents
+    os.environ["OPENAI_API_KEY"] = "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"
     query = input_text
     text_splitter = CharacterTextSplitter(chunk_size=1000, chunk_overlap=0)
     texts = text_splitter.split_documents(documents)
@@ -223,6 +249,9 @@ with demo:
             with gr.Column():
                 query_input = gr.Textbox(lines=7, label="Enter your text")
                 ask_button = gr.Button("Ask")
+                text_output = gr.Text(label="Your Bot Answer")
+                ask_button.click(fn=lambda btn: [text_output.update(chatbot(query_input.value)), chat_history.append(query_input.value)], inputs=[query_input], outputs=text_output)
+                
         with gr.TabItem("Settings"):
             with gr.Tabs():