Code cleanup

app.py

@@ -51,16 +51,14 @@ def extract_youtube_video_id(url):
 
 
 def load_whisper():
-    # if not model:
+
     model = whisper.load_model("small")
     print('Loaded Whisper Medium model')
-    # else:
-    #     print('Already downloaded Whisper model')
-    print('Transcribing with Whisper model')
-    result = model.transcribe("audio.mp3")
-    st.session_state[WHISPER] = result["text"]
-    write_file(result["text"], "transcript.txt")
-AUDIO_FILE = "audio.mp3"
+    if st.session_state[AUDIO_EXISTS]:
+        print('Transcribing with Whisper model')
+        result = model.transcribe("audio.mp3")
+        st.session_state[WHISPER] = result["text"]
+        write_file(result["text"], "transcript.txt")
 
 def load_audio():
     if os.path.exists(AUDIO_FILE):
@@ -75,7 +73,6 @@ def load_audio():
 def display():
     container = st.container()
     text_container = st.container()
-    # whisper_container = st.container()
     load_audio()
 
     #Download Button section
@@ -93,64 +90,24 @@ def display():
             input_submit_button = st.form_submit_button(label='Send')
 
     if input_submit_button and user_input:
-
-        st.write("You entered... " + st.session_state[URL])
-        # transcribe()
-        # download()
-        # download_audio()
+        download()
         load_whisper()
 
     
     with text_container:
        st.text_area(label="Youtube Transcript:",
                     height=200,
-                    value=st.session_state[TEXT])
-    # with whisper_container:
-    #    st.text_area(label="Whisper Transcript:",
-    #                 height=200,
-    #                 value=st.session_state[WHISPER])
-
-
-def download_audio():
-    if st.session_state[URL]:
-        print("Downloading....")
-        yt = pt.YouTube(st.session_state[URL])
-        stream = yt.streams.filter(only_audio=True)[0]
-        stream.download(filename="audio.mp3")
-        print("Downloaded Audio file....")
-
+                    value=st.session_state[WHISPER])
+       
 def download():
   id = extract_youtube_video_id(st.session_state[URL])
   command = [f"yt-dlp --no-config -v --extract-audio --audio-format mp3 {st.session_state[URL]} -o audio.mp3"]
   print(command)
   out = subprocess.run(command, shell=True)
-  print('Download with YT-DLP done!!')
-
-
-def transcribe():
-    loader = YoutubeLoader.from_youtube_url(
-        st.session_state[URL], add_video_info=True)
-    splitter = RecursiveCharacterTextSplitter(chunk_size=2000,chunk_overlap=500)
-    docs = loader.load_and_split(splitter)
-    length = len(docs)
-    index = int(length/3+1)
-    print(f"Loaded {length} documents, Displaying {index}-th document")
-    # st.session_state[TEXT] = docs[index].page_content
-    st.session_state[TEXT] = write_chunks(docs,"transcript.txt")
-
-
-def write_chunks(docs, filename):
-    full_doc = ''
-    for doc in docs:
-       full_doc = full_doc + doc.page_content + "\n"
-    with open(filename, "w") as f:
-        f.write(full_doc)
-        return full_doc
 
 def write_file(text, filename):
     with open(filename, "w") as f:
         f.write(text)
-        # return full_doc
 
 def main():
     init_state()