There are three different ways to get at this information.
First, the data on the screen is derived from an HMTL document. This is common with all programs receiving data from a web-server. In this case, ScreenDataPro accesses the HMTL source in the memory. According to your specifications, we check for markers to find the information that has to be used and then fulfil the required actions.
Second, the access is reading controls in the memory of your computer. It might help to understand this when you imagine that ScreenDataPro is running on the same processor as the program displaying the data. Hence, it has access to the memory, the other program is using as well. Controls are specifications within the screen. Programmatically speaking, the contained information is about the size of a box, the color of the background, the font used and so forth. And they contain the data that is displayed. This is what we need.
Finally, whereas the two previous methods retrieve the information 100-percent error free, we have a third way that is more complex. If we have no HMTL and the programming language used for the program displaying your date does not work with controls, we need to get the information out of the screen memory. With the use of OCR (Optical Character Recognition) we would scan the text and use OCR to recognize the characters and reconstruct the text. ScreenDataPro gets the pixels of the characters directly from the screen and can trace the font used. Therefore it can build up an index to quickly find out to what character each pixel combination refers to. Reading digits is 100% correct. A problem with this method is that some characters cannot be distinguished in some fonts. As an example a capital I (as in India) can be displayed the same way like a lower l (as in Lima). Of course there are workarounds but they cannot reach 100% security. We do not recommend this method when “rendering” is used to display data.