from pytube import YouTube
from pydub import AudioSegment
import openai

# 設定要下載的影片 URL
url = "https://www.youtube.com/watch?v=IvOdTbIYZs4"

# 建立 YouTube 物件
yt = YouTube(url)

# 取得影片的音軌
audio_stream = yt.streams.filter(only_audio=True).first()

# 下載音軌到指定的路徑
audio_stream.download(output_path="path/to/save/directory", filename="audio")

# 讀取下載下來的音檔
audio_file = AudioSegment.from_file("path/to/save/directory/audio")

# 切割音檔成多個小檔案
chunk_size = 100 * 1000  # 100 秒
chunks = [audio_file[i:i+chunk_size] for i in range(0, len(audio_file), chunk_size)]

# 使用 OpenAI 的 Audio API 將每個小檔案轉成文字，然後合併在一起
openai.api_key = "YOUR_API_KEY"
transcript = ""
for chunk in chunks:
    with chunk.export("temp.wav", format="wav") as f:
        result = openai.Audio.transcribe("whisper-1", f)
        transcript += result["text"]

# 印出轉換後的文字
print(transcript)

import os
import openai
openai.api_key = os.getenv("OPENAI_API_KEY")

completion = openai.ChatCompletion.create(
  model="gpt-3.5-turbo",
  messages=[
    {"role": "user", "content": "Hello!"}
  ]
)

print(completion.choices[0].message)

from pytube import YouTube
from pydub import AudioSegment
import openai

# 設定要下載的影片 URL
url = "https://www.youtube.com/watch?v=IvOdTbIYZs4"

# 建立 YouTube 物件
yt = YouTube(url)

# 取得影片的音軌
audio_stream = yt.streams.filter(only_audio=True).first()

# 下載音軌到指定的路徑
audio_stream.download(output_path="path/to/save/directory", filename="audio")

# 讀取下載下來的音檔
audio_file = AudioSegment.from_file("path/to/save/directory/audio")

# 切割音檔成多個小檔案
chunk_size = 100 * 1000  # 100 秒
chunks = [audio_file[i:i+chunk_size] for i in range(0, len(audio_file), chunk_size)]

# 使用 OpenAI 的 Audio API 將每個小檔案轉成文字，然後合併在一起
openai.api_key = "YOUR_API_KEY"
transcript = ""
for chunk in chunks:
    with chunk.export("temp.wav", format="wav") as f:
        result = openai.Audio.transcribe("whisper-1", f)
        transcript += result["text"]

# 整理逐字稿
completion = openai.ChatCompletion.create(
  model="gpt-3.5-turbo",
  messages=[
    {"role": "user", "content": "使用繁體中文閱讀，幫我改錯字、加標點符號，並分段使內容更通順：\n"+transcript}
  ]
)

# 印出整理後的文字
print(completion.choices[0].message.content)

from pytube import YouTube
from pydub import AudioSegment
import openai

# 設定要下載的影片 URL
url = "https://www.youtube.com/watch?v=IvOdTbIYZs4"

# 建立 YouTube 物件
yt = YouTube(url)

# 取得影片的音軌
audio_stream = yt.streams.filter(only_audio=True).first()

# 下載音軌到指定的路徑
audio_stream.download(output_path="path/to/save/directory", filename="audio")

# 讀取下載下來的音檔
audio_file = AudioSegment.from_file("path/to/save/directory/audio")

# 切割音檔成多個小檔案
chunk_size = 100 * 1000  # 100 秒
chunks = [audio_file[i:i+chunk_size] for i in range(0, len(audio_file), chunk_size)]

# 使用 OpenAI 的 Audio API 將每個小檔案轉成文字，然後合併在一起
openai.api_key = "YOUR_API_KEY"
transcript = ""
for chunk in chunks:
    with chunk.export("temp.wav", format="wav") as f:
        result = openai.Audio.transcribe("whisper-1", f)
        transcript += result["text"]

# 依照我們指定的長度分割字串
def split_text(text, max_length):
    return [text[i:i+max_length] for i in range(0, len(text), max_length)]

# 整理逐字稿
def process_text(text):
    response = openai.ChatCompletion.create(
        model="gpt-3.5-turbo",
        messages=[
            {"role": "system", "content": "You are a helpful assistant."},
            {"role": "user", "content": "使用繁體中文閱讀，幫我改錯字、加標點符號，並分段使內容更通順：\n"+text}
        ]
    )
    return response.choices[0].message.content

# 處理長字串
def process_long_text(long_text):
    text_list = split_text(long_text, 1200)
    processed_text_list = [process_text(text) for text in text_list]
    return "".join(processed_text_list)

# 呼叫分段處理函式
processed_transcript = process_long_text(transcript)

# 印出處理後的逐字稿
print(processed_transcript)

from pytube import YouTube
from pydub import AudioSegment
import openai

# 設定要下載的影片 URL
url = "https://www.youtube.com/watch?v=IvOdTbIYZs4"

# 建立 YouTube 物件
yt = YouTube(url)

# 取得影片的音軌
audio_stream = yt.streams.filter(only_audio=True).first()

# 下載音軌到指定的路徑
audio_stream.download(output_path="path/to/save/directory", filename="audio")

# 讀取下載下來的音檔
audio_file = AudioSegment.from_file("path/to/save/directory/audio")

# 切割音檔成多個小檔案
chunk_size = 100 * 1000  # 100 秒
chunks = [audio_file[i:i+chunk_size]
          for i in range(0, len(audio_file), chunk_size)]

# 使用 OpenAI 的 Audio API 將每個小檔案轉成文字，然後合併在一起
openai.api_key = "YOUR_API_KEY"
transcript = ""
for chunk in chunks:
    with chunk.export("temp.wav", format="wav") as f:
        result = openai.Audio.transcribe("whisper-1", f)
        transcript += result["text"]


# 依照我們指定的長度分割字串
def split_text(text, max_length):
    return [text[i:i+max_length] for i in range(0, len(text), max_length)]

# 依照 type 處理文字
def process_text(text, type):
    response = openai.ChatCompletion.create(
        model="gpt-3.5-turbo",
        messages=[
            {"role": "system", "content": "You are a helpful assistant."},
            {"role": "user", "content": type + text}
        ]
    )
    return response.choices[0].message.content

# 處理長字串
def process_long_text(long_text, type):
    text_list = split_text(long_text, 1200)
    processed_text_list = [process_text(text, type) for text in text_list]
    return "".join(processed_text_list)


# 呼叫分段處理函式
processed_transcript = process_long_text(transcript, "使用繁體中文閱讀，幫我改錯字、加標點符號，並分段使內容更通順：\n")

# 呼叫取得摘要函式
processed_summary = process_long_text(processed_transcript, "閱讀以下文字，用「-」作為前綴條列出重點，用「繁體中文」呈現：\n")

with open("path/to/save/directory/text.txt", "w") as file:
    file.write("摘要：\n" + processed_summary + "\n\n\n逐字稿：\n" + processed_transcript)