Doesn't work with the "NewTwitter"

[TolstoyDotCom]

Twitter recently obfuscated their CSS and removed some of the meta info the app was using. The below only applies when using a test user and logging in. So far the non-logged-in version has the same info. However, since logging in is how to determine which tweets are censored, and since Twitter probably won't ditch "New Twitter", I'll have to change the app.

The biggest issue is that the "HighQuality", "LowQuality", and "AbusiveQuality" values that were in the HTML for each tweet might have been removed or could be obfuscated.

I tried to go back to the old Twittter but it didn't work (so far at least). The noscript tag at the top has a form letting you go back but I haven't tried submitting it yet.

It looks like it's possible to find tweets since they're HTML articles and it looks like I can get the tweet text, user, and date. Other info is probably missing or represented by CSS classes (look for CSS code that doesn't change the appearance or duplicates other CSS to find such cases).

There's also a new chunk of JS at the top that might have some info.

In any case, I can change the code to assign "quality" rankings to tweets based on the section they're in.

Another issue is that "New Twitter" doesn't work with Firefox 44. As could be expected from such a company, it just shows "something went wrong" or similar. So, I'll have to upgrade the Webdriver code to use a supported browser (maybe an ESR in the 50s, to avoid having to constantly upgrade the Webdriver library).

[TolstoyDotCom]

First attempt at parsing tweets from "NewTwitter".

[TolstoyDotCom]

This has long since worked (on and off) with NT, but there are issues due to constant Twitter changes.