Get Text Activity keeps reading a page that isn't even open

Help Studio
, ,
1

I am trying to do the Calculate Client Security Hash training project from the Advanced Developer Course and I have run into an issue that is, as far as I can tell, completely impossible. The process is supposed to open work items on a web page and retrieve a few key details (ID, Name, and Country). When I run the code for that in isolation, it works perfectly every time. no matter which work item I have open, the code scrapes the correct text without issue. When I run the complete process, however, it only reads those items from the first work item for every item, despite the page with that ID, Name, and Country not even being open anymore. there literally does not exist anywhere for the activity to be reading those details from, but it gets the same ones instead of the ones that ARE on the page, and only when I am running the complete process. I am totally at a loss for whee it is reading the wrong data from and how it only sees that when running E2E and not when running test cases for the specific workflow.

2

Even when doing a debug run of the complete process, with a breakpoint set up in the workflow where I am reading the text, when I click Preview Extraction on the Get Text activity it shows the correct text being extracted. When I click Step Over and look at what it read however, it is not the correct text. It is text that doesn’t exist anywhere for the bot to possibly be reading by mistake. The bot is getting incorrect data LITERALLY FROM NOWHERE! EX NIHILO! THAT IS IMPOSSIBLE!

3

It sounds like what’s happening is you have some defaults set somewhere, or even have the variables defined more than once. It’s not reading the data wrong, it’s that the variables still have the same value each time, from the first iteration.

4

Incorrect, During Debug Runs I watched the variables change from null to the wrong value. the default back to null every time the workflow starts. during said debug runs, hitting a breakpoint right before the Get Text, I would click the Preview Extraction option and it would show the correct text in the preview and then when I clicke Step Over it would read the wrong values anyway. I have found a workaround, however. By closing and reopening the browser in between work items, I am able to get the correct data read with the same activities, selectors, settings, etc. every time. This proves that the website UiPath set up for this training assignment is deliberately feeding the wrong data to my bot, since I am able to short that logic and prevent it from doing so.