I just gave it a try to screen scrape the table inside the PDF file provided, and i see what you mean.
First of all, its not clear if the table are gonna switch in content?
By that i mean, are you gonna screen scrape multiple PDF files with different tables?
(If you are, you need to modify my solution to be more dynamic, this is only to show possibilities in your specific case).
If so i would do something like this. (could be a little tricky at first try)
Open PDF file in Adobe Reader
Use UiPath Explorer to target the table elements one by one
// This will target Adobe Reader, with a title unknown because we set “*” star in the title attribute. <wnd app='acrord32.exe' cls='AcrobatSDIWindow' title='* - Adobe Reader' />
// Targeting the row we want to get data from (in this case row 1) <ctrl idx='1' role='row' />
// Then the column header 1 <ctrl role='column header' idx='1' />
All of these informations are coming from UiPath explorer and can be reused for every “row - column” you want to target and “convert” the result to CSV formatted string.
Use OCR activity “Get OCR Text” and copy the selected item inside the selector property of “Get OCR Text” activity.
Now for every “row - column” you want to read, change the:
// Targeting the row we want to get data from (in this case row 1) <ctrl idx='1' role='row' />
// Then the column header 2 <ctrl role='column header' idx='2' />
And when its not the header you are reading anymore, change the role to cell instead of column header: <ctrl role='column header' idx='2' />
to
<ctrl role='cell' idx='5' />
I know that there is a bit of work in this solution, but you will have full control of the result/output you get from the table.
And i think it gives alot of new knowledge if you just started UiPath journey.
I am able to retrieve single values with the selector . I want to extract entire column data as a string… With this approach i am only able to extract specified cell value .