app.py

from pytubefix import YouTube
from pytubefix.cli import on_progress
import time
import math
import gradio as gr
import ffmpeg
from faster_whisper import WhisperModel
import requests
import json
import arabic_reshaper # pip install arabic-reshaper
from bidi.algorithm import get_display # pip install python-bidi
from moviepy import VideoFileClip, TextClip, CompositeVideoClip, AudioFileClip, ImageClip
import pysrt
import instaloader
import time
import concurrent.futures
import re
from io import BytesIO
from PIL import Image
import yt_dlp
api_key = "268976:66f4f58a2a905"




def fetch_data(url):
    try:
        response = requests.get(url)
        response.raise_for_status()
        return response.json()
    except requests.exceptions.RequestException as e:
        print(f"An error occurred: {e}")
        return None

def download_file(url):
    try:
        response = requests.get(url.split("#")[0], stream=True)
        response.raise_for_status()
        print(url.split("#")[1])
        with open(url.split("#")[1], 'wb') as file:
            for chunk in response.iter_content(chunk_size=8192):
                if chunk:
                    file.write(chunk)
        file.close()
        print(f"Downloaded successfully: {url.split('#')[1]}")
    except requests.exceptions.RequestException as e:
        print(f"An error occurred: {e}")

def download_chunk(url, start, end, filename, index):
    headers = {'Range': f'bytes={start}-{end}'}
    response = requests.get(url, headers=headers, stream=True)
    response.raise_for_status()
    chunk_filename = f'{filename}.part{index}'
    with open(chunk_filename, 'wb') as file:
        for chunk in response.iter_content(chunk_size=8192):
            if chunk:
                file.write(chunk)
    file.close()
    return chunk_filename

def merge_files(filename, num_parts):
    with open(filename, 'wb') as output_file:
        for i in range(num_parts):
            part_filename = f'{filename}.part{i}'
            with open(part_filename, 'rb') as part_file:
                output_file.write(part_file.read())
            # Optionally, delete the part file after merging
            # os.remove(part_filename)
            part_file.close()

def download_file_in_parallel(link, size, num_threads=4):
    url = link.split("#")[0]
    filename = link.split("#")[1]
    print(url+" filename: "+filename)
    response = requests.head(url)
    #file_size = int(response.headers['Content-Length'])
    chunk_size = size // num_threads

    ranges = [(i * chunk_size, (i + 1) * chunk_size - 1) for i in range(num_threads)]
    ranges[-1] = (ranges[-1][0], size - 1)  # Adjust the last range to the end of the file

    with concurrent.futures.ThreadPoolExecutor(max_workers=num_threads) as executor:
        futures = [
            executor.submit(download_chunk, url, start, end, filename, i)
            for i, (start, end) in enumerate(ranges)
        ]
        for future in concurrent.futures.as_completed(futures):
            future.result()  # Ensure all threads complete

    merge_files(filename, num_threads)
    print(f'Downloaded successfully: {filename}')



def one_youtube(link):
    """
    Downloads a YouTube video to the current directory.

    Args:
        link (str): The URL of the YouTube video to download.

    Returns:
        str: The file path of the downloaded video, or None if an error occurs.
    """
    ydl_opts = {
        'format': 'best',  # Downloads the best quality format with both video and audio
        'outtmpl': '/tmp/%(title)s.%(ext)s',  # Saves the file with its title and extension
    }

    try:
        with yt_dlp.YoutubeDL(ydl_opts) as ydl:
            info = ydl.extract_info(link, download=False)  # Get video info without downloading
            file_path = ydl.prepare_filename(info)  # Determine the output file path
            ydl.download([link])  # Download the video
        print(f'Video downloaded successfully to {file_path}')
        return file_path
    except Exception as e:
        raise ValueError(f"Failed to download YouTube video: {str(e)}")
        return None


# Define your functions here
def yt_download(url):
    yt = YouTube(url)
    print(yt.title)
    video_path = f"{yt.title}.mp4"
    ys = yt.streams.get_highest_resolution()
    print(ys)
    ys.download()
    return video_path, yt.title

def download_image(url, save_path='downloaded_image.jpg'): 
    response = requests.get(url) 
    image = Image.open(BytesIO(response.content)) 
    image.save(save_path) 
    return save_path
    
def insta_oneapi(url, api_key):
    shortcode = url.split("/")[-2]
    print(shortcode)
    url_one="https://api.one-api.ir/instagram/v1/post/?shortcode="+shortcode
    request_body = [{"shortcode": shortcode},]
    headers = {"one-api-token": api_key, "Content-Type": "application/json"}
    response = requests.get(url_one, headers=headers)
    print(response)
    if response.status_code == 200:

        result = response.json()
        try:
            time.sleep(10)
            response = requests.get(result["result"]['media'][0]["url"], stream=True)
            response.raise_for_status()
            with open("video.mp4", 'wb') as file:
                for chunk in response.iter_content(chunk_size=8192):
                    if chunk:
                        file.write(chunk)
            file.close()
            print(f"Downloaded successfully")
            image_url = result["result"]['media'][0]["cover"] 
            image_file_path = download_image(image_url)
            return "video.mp4", image_file_path
        except requests.exceptions.RequestException as e:
            print(f"An error occurred: {e}")
    else:
        print(f"Error: {response.status_code}, {response.text}")
        return None


def insta_download(permalink):
    # Create an instance of Instaloader
    L = instaloader.Instaloader()

    try:
        # Extract the shortcode from the permalink
        if "instagram.com/reel/" in permalink:
            shortcode = permalink.split("instagram.com/reel/")[-1].split("/")[0]
        elif "instagram.com/p/" in permalink:
            shortcode = permalink.split("instagram.com/p/")[-1].split("/")[0]
        else:
            raise ValueError("Invalid permalink format")

        # Load the post using the shortcode
        post = instaloader.Post.from_shortcode(L.context, shortcode)

        # Check if the post is a video
        if not post.is_video:
            raise ValueError("The provided permalink is not a video.")

        # Get the video URL
        video_url = post.video_url

        # Extract the filename from the URL
        filename = video_url.split("/")[-1]
        # Remove query parameters
        filename = filename.split("?")[0]

        # Download the video using requests
        response = requests.get(video_url, stream=True)
        response.raise_for_status()  # Raise an error for bad responses

        # Save the content to a file
        with open(filename, 'wb') as file:
            for chunk in response.iter_content(chunk_size=8192):
                file.write(chunk)

        print(f"Downloaded video {filename} successfully.")
        return filename
    except Exception as e:
        print(f"Failed to download video from {permalink}: {e}")

def extract_audio(input_video_name):
    # Define the input video file and output audio file
    mp3_file = "audio.mp3"

    # Load the video clip
    video_clip = VideoFileClip(input_video_name)

    # Extract the audio from the video clip
    audio_clip = video_clip.audio

    # Write the audio to a separate file
    audio_clip.write_audiofile(mp3_file)

    # Close the video and audio clips
    audio_clip.close()
    video_clip.close()

    print("Audio extraction successful!")
    return mp3_file

def transcribe(audio):
    model = WhisperModel("tiny")
    segments, info = model.transcribe(audio)
    segments = list(segments)
    for segment in segments:
        print("[%.2fs -> %.2fs] %s" % (segment.start, segment.end, segment.text))
    return segments

def format_time(seconds):
    hours = math.floor(seconds / 3600)
    seconds %= 3600
    minutes = math.floor(seconds / 60)
    seconds %= 60
    milliseconds = round((seconds - math.floor(seconds)) * 1000)
    seconds = math.floor(seconds)
    formatted_time = f"{hours:02d}:{minutes:02d}:{seconds:01d},{milliseconds:03d}"
    return formatted_time

def generate_subtitle_file(language, segments, input_video_name):
    subtitle_file = f"sub-{input_video_name}.{language}.srt"
    text = ""
    for index, segment in enumerate(segments):
        segment_start = format_time(segment.start)
        segment_end = format_time(segment.end)
        text += f"{str(index+1)} \n"
        text += f"{segment_start} --> {segment_end} \n"
        text += f"{segment.text} \n"
        text += "\n"
    f = open(subtitle_file, "w", encoding='utf8')
    f.write(text)
    f.close()
    return subtitle_file

def read_srt_file(file_path):
    try:
        with open(file_path, 'r', encoding='utf-8') as file:
            srt_content = file.read()
        file.close()
        return srt_content
    except FileNotFoundError:
        print(f"The file {file_path} was not found.")
    except Exception as e:
        print(f"An error occurred: {e}")

def clean_text(text):
    # Remove 'srt ' from the start of each line
    # Remove ''' from the start and end
    text = re.sub(r"^```|```$", '', text)
    text = re.sub(r'^srt', '', text, flags=re.MULTILINE)
    return text

def enhance_text(api_key, text, google):
    url = "https://api.one-api.ir/chatbot/v1/gpt4o/"

    # Prepare the request body
    request_body = [{
        "role": "user",
        "content": f"{text} ROLE: Expert translator for Persian motivational anime subtitles. TASK: Translate input English SRT (`[X.XXs -> Y.YYs] English text`) to Persian SRT. REQUIREMENTS: 1. STRICT FORMAT: Output MUST be `[X.XXs -> Y.YYs] Persian Text` per line. 2. TIMESTAMPS: CRITICAL - NEVER change timestamps; copy EXACTLY. 3. TRANSLATION: Create motivational, touching, natural, understandable Persian; capture spirit, avoid literalism. 4. LENGTH: Ensure Persian text is concise & readable in time. AVOID: Format errors, timestamp changes, unnatural translation, overly long text.  in respose dont add any thing exept for the srt formated translation.Now translate:"
    },]

    # Add the API key to the request
    headers = {
        "one-api-token": api_key,
        "Content-Type": "application/json"
    }

    # Make the POST request
    response = requests.post(url, headers=headers, json=request_body)

    # Check the response status
    if response.status_code == 200:
        result = response.json()
        clean_text(result["result"][0])
        last = clean_text(result["result"][0])
        print("result: ")
        print(last)
        return last
    else:
        print(f"Error: {response.status_code}, {response.text}")
        return None

def translate_text(api_key, source_lang, target_lang, text):
    url = "https://api.one-api.ir/translate/v1/google/"
    request_body = {"source": source_lang, "target": target_lang, "text": text}
    headers = {"one-api-token": api_key, "Content-Type": "application/json"}
    response = requests.post(url, headers=headers, json=request_body)
    if response.status_code == 200:
        result = response.json()
        enhanced_text = enhance_text(api_key, text, result['result'])
        return enhanced_text
    else:
        print(f"Error: {response.status_code}, {response.text}")
        return None

def write_google(google_translate):
    google = "google_translate.srt"
    with open(google, 'w', encoding="utf-8") as f:
        f.write(google_translate)
    f.close()

def time_to_seconds(time_obj):
    return time_obj.hours * 3600 + time_obj.minutes * 60 + time_obj.seconds + time_obj.milliseconds / 1000

def create_subtitle_clips(subtitles, videosize, fontsize, font, color, debug):
    subtitle_clips = []
    for subtitle in subtitles:
        start_time = time_to_seconds(subtitle.start) # Add 2 seconds offset
        end_time = time_to_seconds(subtitle.end)
        duration = end_time - start_time
        video_width, video_height = videosize
        max_width = video_width * 0.8 
        max_height = video_height * 0.2
        #reshaped_text = arabic_reshaper.reshape(subtitle.text)
        #bidi_text = get_display(reshaped_text)
        text_clip = TextClip(font, subtitle.text, font_size=fontsize, size=(int(video_width * 0.8), int(video_height * 0.2)) ,text_align="center" ,color=color, method='caption').with_start(start_time).with_duration(duration)
        subtitle_x_position = 'center'
        subtitle_y_position = video_height * 0.68
        text_position = (subtitle_x_position, subtitle_y_position)
        subtitle_clips.append(text_clip.with_position(text_position))
    return subtitle_clips



def process_video(url, type):


    if type=="insta":
        input_video, image_path=insta_oneapi(url, api_key)
        input_video_name = input_video.replace(".mp4", "")
        video = VideoFileClip(input_video)
        image_clip = ImageClip(image_path).with_duration(1)
        # Set the position and size of the image (optional) 
        image_clip = image_clip.with_position(("center", "center")).resized(height=video.size[1])
        first_video = CompositeVideoClip([video.with_start(1), image_clip])
        input_video = input_video_name+"_cover.mp4"
        input_video_name = input_video.replace(".mp4", "")
        first_video.write_videofile(input_video, codec="libx264", audio_codec="aac", logger=None)
        input_audio = extract_audio(input_video)
        # **********************************************
    elif type == "youtube":
        input_video = one_youtube(url)
        if input_video is None:
            raise ValueError("Failed to download YouTube video")
        input_audio = extract_audio(input_video)
        input_video_name = os.path.splitext(input_video)[0]
    # Get the current local time
    t = time.localtime()
    # Format the time as a string
    current_time = time.strftime("%H:%M:%S", t)
    print("Current Time =", current_time)
    segments = transcribe(audio=input_audio)
    language = "fa"
    subtitle_file = generate_subtitle_file(language=language, segments=segments, input_video_name=input_video_name)
    source_language = "en"
    target_language = "fa"
    srt_string = read_srt_file(subtitle_file)
    google_translate = translate_text(api_key, source_language, target_language, srt_string)
    write_google(google_translate)
    video = VideoFileClip(input_video)
    audio = AudioFileClip(input_audio)
    video = video.with_audio(audio)
    print(video)
    subtitles = pysrt.open("google_translate.srt", encoding="utf-8")
    output_video_file = input_video_name + '_subtitled' + ".mp4"
    subtitle_clips = create_subtitle_clips(subtitles, video.size, 24, 'arial.ttf', 'yellow', False)
    final_video = CompositeVideoClip([video] + subtitle_clips)
    final_video.write_videofile(output_video_file, codec="libx264", audio_codec="aac", logger=None)
    video.close()
    audio.close()
    print('final')
    # Get the current local time
    t = time.localtime()
    
    # Format the time as a string
    current_time = time.strftime("%H:%M:%S", t)
    print("Current Time =", current_time)
    
    # Generate the URL for the file
    return output_video_file

def download_file(file_path):
    return gr.File.update(file_path)

iface = gr.Interface(fn=process_video, inputs=[gr.Text(),gr.Dropdown(["insta","youtube"])], outputs="file")

iface.launch(debug=True)