hueso/Alpaca
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Rewrote system for getting youtube transcripts

Browse Source
This commit is contained in:
jeffser
2024-10-16 14:47:48 -06:00
parent 134a907eff
commit 218c10f4ad
4 changed files with 73 additions and 24 deletions
Download Patch File Download Diff File Expand all files Collapse all files

26
src/generic_actions.py
View File

@@ -4,7 +4,7 @@ Working on organizing the code
"""
import os, requests
from pytube import YouTube
from youtube_transcript_api import YouTubeTranscriptApi
from html2text import html2text
from .internal import cache_dir
@@ -18,25 +18,35 @@ def connect_remote(remote_url:str, bearer_token:str):
window.model_manager.update_local_list()
window.save_server_config()
def attach_youtube(video_url:str, caption_name:str):
def attach_youtube(video_title:str, video_author:str, watch_url:str, video_url:str, video_id:str, caption_name:str):
buffer = window.message_text_view.get_buffer()
text = buffer.get_text(buffer.get_start_iter(), buffer.get_end_iter(), False).replace(video_url, "")
buffer.delete(buffer.get_start_iter(), buffer.get_end_iter())
buffer.insert(buffer.get_start_iter(), text, len(text))
yt = YouTube(video_url)
text = "{}\n{}\n{}\n\n".format(yt.title, yt.author, yt.watch_url)
result_text = "{}\n{}\n{}\n\n".format(video_title, video_author, watch_url)
caption_name = caption_name.split(' (')[-1][:-1]
if caption_name.startswith('Translate:'):
original_caption_name = get_youtube_transcripts(video_id)[0].split(' (')[-1][:-1]
transcript = YouTubeTranscriptApi.list_transcripts(video_id).find_transcript([original_caption_name]).translate(caption_name.split(':')[-1]).fetch()
result_text += '(Auto translated from Japanese)\n'
else:
transcript = YouTubeTranscriptApi.get_transcript(video_id, languages=[caption_name])
result_text += '\n'.join([t['text'] for t in transcript])
for event in yt.captions[caption_name.split('(')[-1][:-1]].json_captions['events']:
text += "{}\n".format(event['segs'][0]['utf8'].replace('\n', '\\n'))
if not os.path.exists(os.path.join(cache_dir, 'tmp/youtube')):
os.makedirs(os.path.join(cache_dir, 'tmp/youtube'))
file_path = os.path.join(os.path.join(cache_dir, 'tmp/youtube'), f'{yt.title} ({caption_name.split(" (")[0]})')
file_path = os.path.join(os.path.join(cache_dir, 'tmp/youtube'), '{} ({})'.format(video_title.replace('/', ' '), caption_name))
with open(file_path, 'w+', encoding="utf-8") as f:
f.write(text)
f.write(result_text)
window.attach_file(file_path, 'youtube')
def get_youtube_transcripts(video_id:str):
return ['{} ({})'.format(t.language, t.language_code) for t in YouTubeTranscriptApi.list_transcripts(video_id)]
def attach_website(url:str):
response = requests.get(url)
if response.status_code == 200:

27
src/window.py
View File

@@ -69,7 +69,7 @@ class AlpacaWindow(Adw.ApplicationWindow):
preferences_dialog = Gtk.Template.Child()
shortcut_window : Gtk.ShortcutsWindow = Gtk.Template.Child()
file_preview_dialog = Gtk.Template.Child()
file_preview_text_view = Gtk.Template.Child()
file_preview_text_label = Gtk.Template.Child()
file_preview_image = Gtk.Template.Child()
welcome_dialog = Gtk.Template.Child()
welcome_carousel = Gtk.Template.Child()
@@ -405,7 +405,7 @@ class AlpacaWindow(Adw.ApplicationWindow):
if content:
if file_type == 'image':
self.file_preview_image.set_visible(True)
self.file_preview_text_view.set_visible(False)
self.file_preview_text_label.set_visible(False)
image_data = base64.b64decode(content)
loader = GdkPixbuf.PixbufLoader.new()
loader.write(image_data)
@@ -418,10 +418,8 @@ class AlpacaWindow(Adw.ApplicationWindow):
self.file_preview_open_button.set_name(file_path)
else:
self.file_preview_image.set_visible(False)
self.file_preview_text_view.set_visible(True)
buffer = self.file_preview_text_view.get_buffer()
buffer.delete(buffer.get_start_iter(), buffer.get_end_iter())
buffer.insert(buffer.get_start_iter(), content, len(content.encode('utf-8')))
self.file_preview_text_label.set_visible(True)
buffer = self.file_preview_text_label.set_label(content)
if file_type == 'youtube':
self.file_preview_dialog.set_title(content.split('\n')[0])
self.file_preview_open_button.set_name(content.split('\n')[2])
@@ -760,20 +758,23 @@ Generate a title following these rules:
if youtube_regex.match(text):
try:
yt = YouTube(text)
captions = yt.captions
if len(captions) == 0:
transcriptions = generic_actions.get_youtube_transcripts(yt.video_id)
if len(transcriptions) == 0:
self.show_toast(_("This video does not have any transcriptions"), self.main_overlay)
return
video_title = yt.title
if not any(filter(lambda x: '(en' in x, transcriptions)):
transcriptions.insert(0, 'English (Translate:en)')
dialog_widget.simple_dropdown(
_('Attach YouTube Video?'),
_('{}\n\nPlease select a transcript to include').format(video_title),
lambda caption_name, video_url=text: generic_actions.attach_youtube(video_url, caption_name),
["{} ({})".format(caption.name.title(), caption.code) for caption in captions]
_('{}\n\nPlease select a transcript to include').format(yt.streams[0].title),
lambda caption_name, yt=yt, video_url=text: generic_actions.attach_youtube(yt.streams[0].title, yt.author, yt.watch_url, video_url, yt.video_id, caption_name),
transcriptions
)
except Exception as e:
logger.error(e)
self.show_toast(_("This video is not available"), self.main_overlay)
self.show_toast(_("Error attaching video, please try again"), self.main_overlay)
elif url_regex.match(text):
dialog_widget.simple(
_('Attach Website? (Experimental)'),

5
src/window.ui
View File

@@ -884,14 +884,13 @@
<child>
<object class="GtkBox">
<child>
<object class="GtkTextView" id="file_preview_text_view">
<object class="GtkLabel" id="file_preview_text_label">
<property name="margin-top">12</property>
<property name="margin-bottom">12</property>
<property name="margin-start">12</property>
<property name="margin-end">12</property>
<property name="hexpand">true</property>
<property name="vexpand">true</property>
<property name="editable">false</property>
<property name="selectable">true</property>
</object>
</child>
<child>