I need to do the following: having a normal webpage, being able to use a mouse, stylus or finger (in case of a tablet or smartphone) to “draw shapes” over the text, and then using the information of the shapes, determine certain “actions”.
Let me give an example. Imagine this question on stackoverflow, and I “draw” on an ipad a circle around this paragraph. The circle (or the attempt of a circle, manual drawing is always very imperfect), “capture” the paragraph, so I know that I will make an action “over” the paragraph, like, for example, select all the inside text.
My question is: Is this feasible? there is any library that does this? If not, which technologies could be the ones to investigate?
I have been making my own research, but nothing has given me all that I need. For example, canvas draws exactly as I need, but a canvas cannot be used over an html page. It is another element on the page, not one who “wraps” the other.