Add google drive option

adjust.py

@@ -116,7 +116,7 @@ def start_adjust(input_file, to_name, start_time, end_time, progress=gr.Progress
     # Adding subtitle to video
     progress(0.8, desc=ui_lang["progress_add_subtitle"])
     add_subtitle_to_video(
-        input_file, subtitle_adjusted_file, video_subtitle_file, start_time, end_time)
+        input_file, subtitle_adjusted_file_parsed, video_subtitle_file, start_time, end_time)
 
     # Return video file link, transcribe string, transcribe.txt, subtitle.txt
     transcribe_txt_list, _ = utils.read_transcribe_subtitle_file(

app.py

@@ -12,16 +12,17 @@ with gr.Blocks() as demo:
     with gr.Column():
         with gr.Row():
             with gr.Column():
-                # input_url.render()
-                # url_download.render()
-                # url_download.click(get_video_from_url, input_url, input_video)
+                input_url.render()
+                url_download_button.render()
+                url_download_button.click(
+                    get_video_from_url, input_url, input_video)
                 input_video.render()
                 input_video.change(get_duration, input_video, [
                     start_time, end_time])
+            with gr.Column():
                 with gr.Row():
                     start_time.render()
                     end_time.render()
-            with gr.Column():
                 lang_radio.render()
                 model_dropdown.render()
                 # summary_checkbox.render()

file_name.py

@@ -6,9 +6,13 @@ current_working_directory = os.getcwd()
 download_file = f"{current_working_directory}/download_video.mp4"
 audio_file = f"{current_working_directory}/input.wav"
 transcribe_file = f"{current_working_directory}/transcribe.txt"
+
 subtitle_file = f"{current_working_directory}/subtitle.srt"
-transcribe_adjusted_file = f"{current_working_directory}/transcribe_adjusted.txt"
+subtitle_file_parsed = "subtitle.srt"
 subtitle_adjusted_file = f"{current_working_directory}/subtitle_adjusted.srt"
+subtitle_adjusted_file_parsed = "subtitle_adjusted.srt"
+
+transcribe_adjusted_file = f"{current_working_directory}/transcribe_adjusted.txt"
 video_subtitle_file = f"{current_working_directory}/output.mp4"
 sample_groups_json = f"{current_working_directory}/sample_groups.json"
 speaker_groups_json = f"{current_working_directory}/speaker_groups.json"

lang_ui.py

@@ -12,14 +12,19 @@ english_ui = {
         The following is the result of the transcribe
     """,
 
+    "get_video_start_info": "Start downloading",
+    "get_video_fail_info": "Failed to download, please make sure the link can be accessed by 'Anyone with the link'",
+    "get_video_finished_info": "Finish downloading",
+
     "input_video_warning": "Please submit a video",
     "start_time_warning": "Please provide a correct start time",
     "end_time_warning": "Please provide a correct end time",
     "lang_radio_warning": "Please choose a language for the video",
     "model_dropdown_warning": "Please choose a model size for the video",
 
-    "input_url_label": "Insert video url",
-    "download_button_value": "Download video",
+    "input_url_label": "Insert Google Drive video url (Optional)",
+    "input_url_info": "Instead of uploading your video directly, you can use Google Drive shareable link",
+    "download_button_value": "Download Google Drive Video",
     "input_video_label": "Insert video",
     "start_time_label": "Start time",
     "end_time_label": "End time",
@@ -63,14 +68,19 @@ indonesia_ui = {
         Berikut hasil akhir dari transkrip
     """,
 
+    "get_video_start_info": "Mulai mengunduh",
+    "get_video_fail_info": "Gagal mengunduh, pastikan tautan dapat diakses oleh 'Siapa saja yang memiliki link'",
+    "get_video_finished_info": "Selesai mengunduh",
+
     "input_video_warning": "Mohon masukkan video",
     "start_time_warning": "Mohon berikan waktu mulai yang sesuai",
     "end_time_warning": "Mohon berikan waktu selesai yang sesuai",
     "lang_radio_warning": "Mohon pilih bahasa yang digunakan dalam video",
     "model_dropdown_warning": "Mohon pilih ukuran model yang digunakan untuk video",
 
-    "input_url_label": "Masukkan link video",
-    "download_button_value": "Unduh video",
+    "input_url_label": "Masukkan tautan video Google Drive (Opsional)",
+    "input_url_info": "Jika anda terkendala untuk mengunggah video secara langsung, silahkan unggah video ke Google Drive dan gunakan tautan video terbuka dari Google Drive pada kolom berikut",
+    "download_button_value": "Unduh Video Google Drive",
     "input_video_label": "Masukkan video",
     "start_time_label": "Waktu mulai",
     "end_time_label": "Waktu selesai",

requirements.txt

@@ -1,3 +1,4 @@
+gdown
 ffmpeg-python
 pyannote-audio @ git+https://github.com/pyannote/pyannote-audio.git
 pydub

set_up.py

@@ -46,7 +46,7 @@ def prepare_input(input_file, start_time, end_time, lang, model_size, progress=g
 
     # Add subtitle to video
     progress(0.8, desc=ui_lang["progress_add_subtitle"])
-    add_subtitle_to_video(input_file, subtitle_file,
+    add_subtitle_to_video(input_file, subtitle_file_parsed,
                           video_subtitle_file, start_time, end_time)
 
     # Return video file link, transcribe string, transcribe.txt, subtitle.txt

transcribe.py

@@ -2,7 +2,7 @@ from faster_whisper import WhisperModel
 from utils import load_groups_json
 import torch
 import gc
-import file_name
+from ui import *
 import utils
 
 gc.collect()
@@ -23,7 +23,7 @@ def start_transcribe(input_file, lang_choice: int, model_size_choice: int, progr
 
     subtitle_txt_list = []
     transcribe_txt_list = []
-    for speaker, _ in zip(speaker_groups, progress.tqdm(speaker_groups, desc="Transcribing")):
+    for speaker, _ in zip(speaker_groups, progress.tqdm(speaker_groups, desc=ui_lang["progress_transcribing_audio"])):
         # Transcribe and save temp file
         audiof = f"{speaker}.wav"
         print(f"Loading {audiof}")

ui.py

@@ -1,9 +1,8 @@
 from lang_ui import get_ui_lang
 import gradio as gr
 import file_name
-import subprocess
-import re
-import urllib.request
+import ffmpeg
+import gdown
 
 # Display available langauges and set default UI language
 ui_lang_index = 1
@@ -18,18 +17,16 @@ ui_lang_radio = gr.Radio(
     available_ui_lang, type="index", value=available_ui_lang[ui_lang_index], interactive=True, show_label=False)
 top_markdown = gr.Markdown(
     ui_lang["top_markdown"])
-# TODO Add video by link
 input_url = gr.Textbox(
-    max_lines=1, label=ui_lang["input_url_label"], interactive=True)
-url_download = gr.Button(
-    ui_lang["download_button_value"], interactive=True)
+    max_lines=1, label=ui_lang["input_url_label"], info=ui_lang["input_url_info"], interactive=True)
+url_download_button = gr.Button(
+    ui_lang["download_button_value"], size='sm', interactive=True)
 input_video = gr.Video(
     label=ui_lang["input_video_label"], interactive=True)
 start_time = gr.Textbox(
     max_lines=1, placeholder="00:00:00", label=ui_lang["start_time_label"], interactive=False)
 end_time = gr.Textbox(
-    max_lines=1, placeholder="01:00:00", label=ui_lang["end_time_label"], interactive=False)
-# TODO Use custom language
+    max_lines=1, placeholder="99:99:99", label=ui_lang["end_time_label"], interactive=False)
 lang_radio = gr.Radio(
     ui_lang["lang_radio_choices"], label=ui_lang["lang_radio_label"], info=ui_lang["lang_radio_info"], type='index', interactive=True)
 model_dropdown = gr.Dropdown(
@@ -69,15 +66,17 @@ def time_str(t):
 
 
 def get_video_from_url(url):
-    try:
-        urllib.request.urlretrieve(url, file_name.download_file)
-        print("Download finished")
-        return file_name.download_file
-    except:
-        print("Download failed")
-        gr.Warning(ui_lang["input_video_warning"])
+    gr.Info(ui_lang["get_video_start_info"])
+    download_result = gdown.download(
+        url, file_name.download_file, quiet=False, fuzzy=True)
+
+    if (download_result is None):
+        gr.Warning(ui_lang["get_video_fail_info"])
         return None
 
+    gr.Info(ui_lang["get_video_finished_info"])
+    return file_name.download_file
+
 
 def get_duration(input_file):
     print("Checking file")
@@ -88,17 +87,13 @@ def get_duration(input_file):
             end_time.update(None, interactive=False)
         ]
     print("Getting duration")
-    result = subprocess.run(['ffprobe', '-v', 'error', '-show_entries', 'format=duration', '-of',
-                            'default=noprint_wrappers=1:nokey=1', input_file], stdout=subprocess.PIPE, stderr=subprocess.PIPE, text=True)
-    duration = float(result.stdout.strip())
-    print(f"Result: {duration}")
-    # info_json = ffmpeg.probe(input_file)
+    info_json = ffmpeg.probe(input_file)
     print("Probing finished")
-    # duration_seconds = float(info_json['format']['duration'])
-    # duration_formatted = time_str(duration_seconds)
+    duration_seconds = float(info_json['format']['duration'])
+    duration_formatted = time_str(duration_seconds)
     return [
         start_time.update("00:00:00", interactive=True),
-        end_time.update(duration, interactive=True)
+        end_time.update(duration_formatted, interactive=True)
     ]
 
 
@@ -113,6 +108,10 @@ def change_lang(input):
         # Top
         top_markdown.update(
             ui_lang["top_markdown"]),
+        input_url.update(
+            label=ui_lang["input_url_label"], info=ui_lang["input_url_info"]),
+        url_download_button.update(
+            ui_lang["download_button_value"]),
         input_video.update(
             label=ui_lang["input_video_label"]),
         start_time.update(
@@ -123,15 +122,14 @@ def change_lang(input):
             choices=ui_lang["lang_radio_choices"], value=None, label=ui_lang["lang_radio_label"], info=ui_lang["lang_radio_info"],),
         model_dropdown.update(
             choices=ui_lang["model_dropdown_choices"], value=None, label=ui_lang["model_dropdown_label"], info=ui_lang["model_dropdown_info"]),
-        summary_checkbox.update(
-            label=ui_lang["summary_checkbox_label"], info=ui_lang["summary_checkbox_info"]),
         start_button.update(
             ui_lang["start_button_value"]),
 
         # Middle
         middle_markdown.update(
             ui_lang["middle_markdown"]),
-        adjust_speaker.update(label=ui_lang["adjust_speaker_value"]),
+        adjust_speaker.update(
+            label=ui_lang["adjust_speaker_value"]),
         prev_button.update(
             ui_lang["prev_button_value"]),
         next_button.update(
@@ -142,11 +140,13 @@ def change_lang(input):
         # Bottom
         bottom_markdown.update(
             ui_lang["bottom_markdown"]),
-        output_video.update(label=ui_lang["output_video_label"]),
-        output_transcribe.update(label=ui_lang["output_transcribe_label"]),
+        output_video.update(
+            label=ui_lang["output_video_label"]),
+        output_transcribe.update(
+            label=ui_lang["output_transcribe_label"]),
     ]
 
 
 # comp_to_update and change_lang return must always be in equal number
 comp_to_update = [
-    top_markdown, input_video, start_time, end_time, lang_radio, model_dropdown, summary_checkbox, start_button, middle_markdown, adjust_speaker, prev_button, next_button, adjust_button, bottom_markdown, output_video, output_transcribe]
+    top_markdown, input_url, url_download_button, input_video, start_time, end_time, lang_radio, model_dropdown, start_button, middle_markdown, adjust_speaker, prev_button, next_button, adjust_button, bottom_markdown, output_video, output_transcribe]