Fixes building the database from scratch in twitter - #691
Conversation
The DOM for tweets was not loading completely in the 500ms, causing the selector to return 0 items. Increasing it to 2000ms solves that.
Some object paths used to parse tweets and userCore seem to have been removed from the data. Removing the particular paths in this commit from the check does not affect anything since that path was not getting used.
|
@SaptakS do you think this also fixes cyd not archiving bookmarks properly anymore when using this option? It logs in, open bookmarks then instantly says it's finished without fetching anything new. This stopped working about 2 weeks ago. 
|
|
@jonatino I tested bookmarks also, and have adjusted the sleep time for that as well. Locally testing, this PR seems to fix the bookmarks also. |
|
Based on #693 I think there is another reason/step why building database from scratch might fail. I will check if I can add that fix/fallback also to this PR. |
|
Trying this out, and for some reason with my test account I'm getting into an infinite loop when trying to save tweets: 
|
|
Ah maybe I'm also hitting #693 |
|
@redshiftzero Yeah. I think I need to look into #693. With the chat when they updated the route, they also updated the MITM paths that we track. So hopefully its not the same. I will try to take a look this week into it. |
3 participants
Fixes #669
It actually almost fixes the issue. The two main reasons seemed to be that the DOM wasn't fully loading, and the response data didn't have some objects.
However, the DM/conversations is completely broken. I did update the paths that they use now, but that itself doesn't fix the issue. I am still looking into it and will create a separate draft PR, with more information related to that.