Create app.py

app.py

@@ -0,0 +1,106 @@
+import openai 
+openai.api_key = os.getenv("OPENAI_API_KEY")
+
+import gradio as gr 
+from youtube_transcript_api import YouTubeTranscriptApi
+from urllib.parse import urlparse, parse_qs
+from requests.structures import CaseInsensitiveDict
+
+
+def Prompt_T(context):
+  result  = """I want you to act as a content writer who is working with youtube video transcript. Summarise the following text:
+  =========
+  """+ context +"""
+  =========
+  Answer:"""
+  return result
+
+
+def split_string(string, chunk_size):
+    return [string[i:i+chunk_size] for i in range(0, len(string), chunk_size)]
+
+
+def gpt_api (input_text):
+  completion = openai.ChatCompletion.create(
+    model="gpt-3.5-turbo",
+    messages=[  {"role": "system", "content": input_text} ]
+    )
+  response = completion.choices[0].message.content
+  return response
+
+def generate(video_url, request: gr.Request):
+
+    #Если есть get переменная v, создаем video_url
+    headers =request.headers
+    url = headers.get('referer')
+    parsed_url = urlparse(url)
+    query_params = parse_qs(parsed_url.query)
+    my_dict = query_params
+    try:
+      my_v = my_dict['v'][0]
+      video_url ="https://youtube.com/watch?v="+my_v
+    except KeyError:
+      print("Ключ 'v' отсутствует в словаре.")
+      my_v = ""
+
+    #Если две переменные пустые, то показываем базовую страницу с рекламой
+    if (video_url =="") and (my_v == ""):  
+      html_embed='<div><br> An easy way to get video descriptions If you are on YouTube itself, simply add "zxc" in front of YouTube to the videos address.</div>'
+      summarize=""
+      return summarize, html_embed
+    
+    #похоже ли video_url на номальну ссылку
+    if "youtube.com/watch?v=" in video_url: x=111
+    else: return "Неверный URL", "Ошибка"
+ 
+    #Пробуем извлеч video_id пока на английском
+    video_id = video_url[-11:]
+    try:
+      t = YouTubeTranscriptApi.get_transcript(video_id,languages=["en"])
+      # do something with the transcript
+    except Exception as e:
+      return "Несмогли нати трнскрипт", "Ошибка"
+      
+    finalString = ""
+    for item in t:
+        text = item['text']
+        finalString += text + " "
+
+
+    print("Transcript:",finalString)
+    print("Transcript lenght:",len(finalString))
+    print ("===============================================")
+    input_string = finalString
+    chunk_size = 10000
+    result_list = split_string(input_string, chunk_size)
+    eng_answer=""
+    count= 0
+    for item in result_list:
+      count = count +1
+      context = item
+      eng_prompt = Prompt_T(context)
+      eng_answer = eng_answer +" \n" + gpt_api (eng_prompt)
+      print("Context:", context)
+      print(count, " - part eng_answer:", eng_answer)
+      print("==========================")
+      
+    html_embed='<iframe width="450" height="158" src="https://www.youtube.com/embed/'+ video_id +'" title="YouTube video player" frameborder="0" allow="accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture; web-share" allowfullscreen></iframe>'
+    html_content="<h6>"+"<br>"+eng_answer+"</h6>"
+    return html_content, html_embed
+title = "YouTube Summorize (only english video)"
+css="""
+footer {visibility: hidden}
+.gradio-container {padding-top: 100px}
+"""
+with gr.Blocks(css=css, title=title) as demo:
+    with gr.Row():   
+        with gr.Column():
+          input_d = gr.Textbox(label="YouTube video URL", placeholder="https://www.youtube.com/watch?v=XXXXXXXX")
+          greet_btn = gr.Button("Summarise")
+        dt_2 = gr.outputs.HTML() 
+    dt_1 = gr.outputs.HTML()
+    dt =[dt_1, dt_2 ]
+    greet_btn.click(generate, inputs=input_d, outputs=dt)
+    demo.load(generate, inputs=input_d, outputs=dt)
+
+demo.launch(share=False, debug=True )