Spaces:
Runtime error
Runtime error
David Li commited on
Commit ·
4de615e
1
Parent(s): d255f96
fix: update again
Browse files
app.py
CHANGED
|
@@ -3,6 +3,7 @@ import gradio as gr
|
|
| 3 |
import ffmpeg
|
| 4 |
from yt_dlp import YoutubeDL
|
| 5 |
import os
|
|
|
|
| 6 |
|
| 7 |
youtube_livestream_codes = [
|
| 8 |
91,
|
|
@@ -23,7 +24,13 @@ youtube_mp4_codes = [
|
|
| 23 |
134
|
| 24 |
]
|
| 25 |
|
| 26 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 27 |
|
| 28 |
def get_video_metadata(video_url: str = "https://www.youtube.com/watch?v=21X5lGlDOfg&ab_channel=NASA")-> dict:
|
| 29 |
with YoutubeDL({'outtmpl': '%(id)s.%(ext)s'}) as ydl:
|
|
@@ -101,7 +108,9 @@ def get_text_from_mp3_whisper(inputType:str, mp3_file: str, url_path: str, taskN
|
|
| 101 |
return "srcLanguage is not set"
|
| 102 |
if inputType == "url":
|
| 103 |
filename = get_all_files(url_path)
|
|
|
|
| 104 |
result = model.transcribe(filename, **transcribe_options)
|
|
|
|
| 105 |
else:
|
| 106 |
result = model.transcribe(mp3_file, **transcribe_options)
|
| 107 |
# adjust for spacy mode
|
|
@@ -116,6 +125,7 @@ def get_text_from_mp3_whisper(inputType:str, mp3_file: str, url_path: str, taskN
|
|
| 116 |
lines.append(segment.get("text", "").strip())
|
| 117 |
lines.append('')
|
| 118 |
words = '\n'.join(lines)
|
|
|
|
| 119 |
input_file = filename or mp3_file
|
| 120 |
# ffmpeg -i testing.mp4 -vf subtitles=transcript.srt mysubtitledmovie.mp4
|
| 121 |
# use ffmpeg bindings to add subtitles to video
|
|
@@ -128,7 +138,7 @@ def get_text_from_mp3_whisper(inputType:str, mp3_file: str, url_path: str, taskN
|
|
| 128 |
|
| 129 |
ffmpeg.run(output_video)
|
| 130 |
# for spacy use advanced logic to extract and append to html_text using tables?
|
| 131 |
-
|
| 132 |
# get output_video as mp4
|
| 133 |
return result.get("segments"), words, "subtitled.mp4"
|
| 134 |
|
|
|
|
| 3 |
import ffmpeg
|
| 4 |
from yt_dlp import YoutubeDL
|
| 5 |
import os
|
| 6 |
+
import sys
|
| 7 |
|
| 8 |
youtube_livestream_codes = [
|
| 9 |
91,
|
|
|
|
| 24 |
134
|
| 25 |
]
|
| 26 |
|
| 27 |
+
def second_to_timecode(x: float) -> str:
|
| 28 |
+
hour, x = divmod(x, 3600)
|
| 29 |
+
minute, x = divmod(x, 60)
|
| 30 |
+
second, x = divmod(x, 1)
|
| 31 |
+
millisecond = int(x * 1000.)
|
| 32 |
+
|
| 33 |
+
return '%.2d:%.2d:%.2d,%.3d' % (hour, minute, second, millisecond)
|
| 34 |
|
| 35 |
def get_video_metadata(video_url: str = "https://www.youtube.com/watch?v=21X5lGlDOfg&ab_channel=NASA")-> dict:
|
| 36 |
with YoutubeDL({'outtmpl': '%(id)s.%(ext)s'}) as ydl:
|
|
|
|
| 108 |
return "srcLanguage is not set"
|
| 109 |
if inputType == "url":
|
| 110 |
filename = get_all_files(url_path)
|
| 111 |
+
print("Retrieved the file")
|
| 112 |
result = model.transcribe(filename, **transcribe_options)
|
| 113 |
+
print("transcribing the file")
|
| 114 |
else:
|
| 115 |
result = model.transcribe(mp3_file, **transcribe_options)
|
| 116 |
# adjust for spacy mode
|
|
|
|
| 125 |
lines.append(segment.get("text", "").strip())
|
| 126 |
lines.append('')
|
| 127 |
words = '\n'.join(lines)
|
| 128 |
+
print("done transcribing")
|
| 129 |
input_file = filename or mp3_file
|
| 130 |
# ffmpeg -i testing.mp4 -vf subtitles=transcript.srt mysubtitledmovie.mp4
|
| 131 |
# use ffmpeg bindings to add subtitles to video
|
|
|
|
| 138 |
|
| 139 |
ffmpeg.run(output_video)
|
| 140 |
# for spacy use advanced logic to extract and append to html_text using tables?
|
| 141 |
+
print(words)
|
| 142 |
# get output_video as mp4
|
| 143 |
return result.get("segments"), words, "subtitled.mp4"
|
| 144 |
|