tubearchivist/tubearchivist-frontend
1
0
Fork 0
mirror of https://github.com/tubearchivist/tubearchivist-frontend.git synced 2024-11-22 20:00:15 +00:00
Code Issues Projects Releases Wiki Activity

standardize country specific subtitle language codes

Browse Source
This commit is contained in:
simon 2022-02-05 16:07:11 +07:00
parent 3ef35a9d53
commit ac531affb5
No known key found for this signature in database
GPG Key ID: 2C15AA5E89985DD4
1 changed files with 26 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

28
tubearchivist/home/src/index/video.py
View File

@ -68,11 +68,21 @@ class YoutubeSubtitle:
return relevant_subtitles return relevant_subtitles
def _normalize_lang(self):
"""normalize country specific language keys"""
all_subtitles = self.youtube_meta.get("subtitles")
all_keys = list(all_subtitles.keys())
for key in all_keys:
lang = key.split("-")[0]
old = all_subtitles.pop(key)
all_subtitles[lang] = old
return all_subtitles
def get_user_subtitles(self): def get_user_subtitles(self):
"""get subtitles uploaded from channel owner""" """get subtitles uploaded from channel owner"""
print(f"{self.youtube_id}: get user uploaded subtitles") print(f"{self.youtube_id}: get user uploaded subtitles")
all_subtitles = self.youtube_meta.get("subtitles") all_subtitles = self._normalize_lang()
if not all_subtitles: if not all_subtitles:
return False return False
@ -90,6 +100,19 @@ class YoutubeSubtitle:
return relevant_subtitles return relevant_subtitles
def download_subtitles(self, relevant_subtitles):
"""download subtitle files to archive"""
for subtitle in relevant_subtitles:
dest_path = os.path.join(
self.config["application"]["videos"], subtitle["media_url"]
)
response = requests.get(subtitle["url"])
if response.ok:
with open(dest_path, "w", encoding="utf-8") as subfile:
subfile.write(response.text)
else:
print(f"{self.youtube_id}: failed to download subtitle")
class YoutubeVideo(YouTubeItem, YoutubeSubtitle): class YoutubeVideo(YouTubeItem, YoutubeSubtitle):
"""represents a single youtube video""" """represents a single youtube video"""
@ -256,6 +279,7 @@ class YoutubeVideo(YouTubeItem, YoutubeSubtitle):
subtitles = handler.get_subtitles() subtitles = handler.get_subtitles()
if subtitles: if subtitles:
self.json_data["subtitles"] = subtitles self.json_data["subtitles"] = subtitles
handler.download_subtitles(relevant_subtitles=subtitles)
def index_new_video(youtube_id): def index_new_video(youtube_id):