add-playwright-ground-truth
@imenelydiaker
and
@dhuynh95
We can merge this pull request. It just adds a new column for playwright_ground_truth. You can preview the data on my hf repo which i will delete once we merge this.
https://huggingface.co./datasets/shubhamofbce/playwright_wave
@shubhamofbce : It looks great!
Just a small point: it looks like there are quotation marks sometimes, could you clean them?
Also, have you tested the elements they select are the same to the ones the Selenium code selects? You could print the HTML associated to the element to check they match.
Thanks for the help
@dhuynh95
I found that in hf dataset viewer they add a quotation mark if they encode the data. So it is not something that i can control. I tried traversing my data for rows starting " and it returned none. so the quotation marks are not going to affect anything.
I also found this in some urls of our actual dataset, see the ss attached but when we pull the data it comes without those marks.
@dhuynh95
I did run a test and here is what i found about it.
35/50 - Full match same to same outer html
7/50 - My outer html is span Hello world /span and ground truth is div span Hello world /span /div
8/50 - (Failed cases) There are some queries related to cookies, like i accept and manage cookies. these elements are not visible at all visible in my browser during testing. So i have just guessed the playwright code for this in the data.
Let me know your thoughts about it.