fixed memory leak related to video files

Twitter/downloader.py

@@ -68,7 +68,7 @@ async def download_loop(botData: RuntimeBotData):
         await discordHelper.post_result(results, guild, botData.new_accounts)
     except Exception as ex:
         print(ex)
-        await discordHelper.send_error(traceback.format_exc()[0:256], botData)
+        await discordHelper.send_error(str(ex) + " " + traceback.format_exc(), botData)
  
 async def download_post(artist: x_accounts, tweet: Tweet, botData: RuntimeBotData):
     x_post = x_posts(id = tweet.id, account_id = tweet.author.id, date = tweet.date, text = tweet.text)
@@ -79,8 +79,8 @@ async def download_post(artist: x_accounts, tweet: Tweet, botData: RuntimeBotDat
         return
     
     print("New media post:", str(tweet.url))
-    media = await tweetHelper.GetTweetMediaUrls(tweet)
-    image_containers = [x_posts_images(tweet.id, idx, file = url) for idx, url in enumerate(media)]
+    media = await tweetHelper.GetTweetMedia(tweet)
+    image_containers = [x_posts_images(tweet.id, idx, file = med.url) for idx, med in enumerate(media)]
 
     try:
         async with aiohttp.ClientSession() as session:
@@ -104,7 +104,7 @@ async def download_post(artist: x_accounts, tweet: Tweet, botData: RuntimeBotDat
     for idx, attachment in enumerate(downloaded_media):
         container = image_containers[idx]
         container.saved_file = attachment.file_name
-        container.vox_label, container.rating, container.tags, filtered_tags, container.phash, container.dhash, container.error_id = await classify_all(attachment.file_bytes, botData.classifier, botData.vox)
+        container.vox_label, container.rating, container.tags, filtered_tags, container.phash, container.dhash, container.error_id = await classify_all(attachment.file_bytes if not attachment.is_video else None, botData.classifier, botData.vox)
         
         if container.vox_label not in vox_labels:
             vox_labels.append(container.vox_label)

Twitter/tweetHelper.py

@@ -15,18 +15,22 @@ if TYPE_CHECKING:
 class TweetMedia:
     url : str
     file_name : str
+    is_video: bool
 
-    def __init__(self, url, file_name):
+    def __init__(self, url, file_name, is_video: bool):
         self.url = url
         self.file_name = file_name
+        self.is_video = is_video
 
 class DownloadedMedia:
-    file_bytes : str
+    file_bytes : bytes
     file_name : str
+    is_video: bool
 
-    def __init__(self, bytes, file_name):
+    def __init__(self, bytes, file_name, is_video: bool):
         self.file_bytes = bytes
         self.file_name = file_name
+        self.is_video = is_video
 
 async def GetTweetMedia(tweet : Tweet) -> list[TweetMedia]:
     mediaList : list[TweetMedia] = []
@@ -34,12 +38,12 @@ async def GetTweetMedia(tweet : Tweet) -> list[TweetMedia]:
         if media.file_format == 'mp4':
             best_stream = await media.best_stream()
             fileName    = f"{tweet.author.screen_name}_{tweet.id}_{idx}.{media.file_format}"
-            mediaList.append(TweetMedia(best_stream.direct_url, fileName))
+            mediaList.append(TweetMedia(best_stream.direct_url, fileName, True))
         else:
             best_stream = await media.best_stream()
             extension   = best_stream.file_format
             fileName    = f"{tweet.author.screen_name}_{tweet.id}_{idx}.{extension}"
-            mediaList.append(TweetMedia(best_stream.direct_url, fileName))
+            mediaList.append(TweetMedia(best_stream.direct_url, fileName, False))
 
     return mediaList
 
@@ -47,18 +51,18 @@ async def GetTweetMediaUrls(tweet : Tweet):
     mediaList = await GetTweetMedia(tweet)
     return [media.url for media in mediaList]
 
-async def DownloadMedia(post_id, account_id, account_name, url_list : list, session) -> list[DownloadedMedia]:
+async def DownloadMedia(post_id, account_id, account_name, media_list : list[TweetMedia], session) -> list[DownloadedMedia]:
     result : list[DownloadedMedia] = []
     path = f"{Global_Config["x_download_path"]}{account_id}"
     os.makedirs(path, exist_ok=True)
 
-    for idx, file_url in enumerate(url_list):
-        file_name = get_file_name(account_name, post_id, idx, file_url)
+    for idx, media in enumerate(media_list):
+        file_name = get_file_name(account_name, post_id, idx, media.url)
         full_path = f"{path}/{file_name}"
 
-        photo_bytes = await downloadHelper.save_to_file(file_url, full_path, session)
+        photo_bytes = await downloadHelper.save_to_file(media.url, full_path, session)
 
-        result.append(DownloadedMedia(photo_bytes, file_name))
+        result.append(DownloadedMedia(photo_bytes, file_name, media.is_video))
 
     return result