I am a little puzzled by the issue I came across but before I get to it, what is the best way to search for a plain text on a webpage? I have some success using "if text on screen (OCR)" function and had very inconsistent result with "Move mouse to text on screen (OCR)" function.
I have following flow: after opening an Excel file, and getting necessary data from it, I split the text with a comma character, and I end up creating a variable that contains a unique list of values (containing alpha-numerical characters) that I am going to search for on screen (internal company's website) one-by-one (the website screen also has a list of strictly unique values). If one is found, it will click on a checkmark next to the text.
The issue: while testing, some text consistently is being skipped as if OCR function did not locate it, when I clearly see it on the screen (and can confirm it is present using Ctrl+F function). For example, the first two values are found, but the third one is skipped, the last two are also found. There's no error message, so nothing to troubleshoot.
0 ABC123
1 XYZ456
2 QWE789
3 TYU987
4 YUI654
Where can I go from here troubleshooting wise?
- All of the text is visible during the flow
- I tried to move the 'offender' value to a different row just to see if that would make any difference. Of course it didn't.
- I tried creating "subflow" for testing purposes, and declared a variable manually to test the same steps. The results were inconsistent (I found the value a couple of times, but majority was a miss)
As a side note: the OCR functionality sometime appear to work on a subflow but not the main flow. Very odd.
I haven't been using Power Automate Desktop for long but everything seems logical until issues like that....
