Update app.py

app.py

@@ -96,32 +96,32 @@ def submit_message(prompt, prompt_template, temperature, max_tokens, context_len
     #vectordb = Chroma.from_documents(split_pages, embeddings, persist_directory=persist_directory)
     #vectordb.persist()
 
-    #path = './files'
-    #pdf_files = glob.glob(os.path.join(path, "*.pdf"))
+    path = './files'
+    pdf_files = glob.glob(os.path.join(path, "*.pdf"))
     
-    #merger = PdfWriter()
+    merger = PdfWriter()
     
     # add all file in the list to the merger object
-    #for pdf in pdf_files:
-    #  merger.append(pdf)
-    #merger.write("merged-pdf.pdf")
-    #merger.close()
+    for pdf in pdf_files:
+      merger.append(pdf)
+    merger.write("merged-pdf.pdf")
+    merger.close()
     
-    #reader = PdfReader("merged-pdf.pdf")
-    #raw_text = ''
-    #for i, page in enumerate(reader.pages):
-    #  text = page.extract_text()
-    #  if text:
-     #     raw_text += text
-    #text_splitter = CharacterTextSplitter(        
-    #  separator = "\n",
-    #  chunk_size = 1000,
-    #  chunk_overlap  = 200,
-    #  length_function = len,
-    #)
-    #texts = text_splitter.split_text(raw_text)
-    #len(texts)
-    #embeddings = OpenAIEmbeddings()
+    reader = PdfReader("merged-pdf.pdf")
+    raw_text = ''
+    for i, page in enumerate(reader.pages):
+      text = page.extract_text()
+      if text:
+          raw_text += text
+    text_splitter = CharacterTextSplitter(        
+      separator = "\n",
+      chunk_size = 1000,
+      chunk_overlap  = 200,
+      length_function = len,
+    )
+    texts = text_splitter.split_text(raw_text)
+    len(texts)
+    embeddings = OpenAIEmbeddings()
         
         
     #with open("foo.pkl", 'wb') as f:
@@ -161,10 +161,10 @@ def submit_message(prompt, prompt_template, temperature, max_tokens, context_len
 
 #        completion = completion({"question": query, "chat_history": history[-context_length*2:]})
 
-        with open("foo.pkl", 'rb') as f:
-            new_docsearch = pickle.load(f)
+        #with open("foo.pkl", 'rb') as f:
+        #    new_docsearch = pickle.load(f)
         
-        docsearch = FAISS.from_texts(texts, new_docsearch)
+        docsearch = FAISS.from_texts(texts, embeddings)
         query = str(system_prompt + history[-context_length*2:] +  [prompt_msg])
         docs = docsearch.similarity_search(query)
         #print(docs[0].page_content)