If the font size is fixed then just compute:

row = (y-verticaloffset)/characterheight

column = (x-horizontaloffset)/characterwidth

whats x and y? if its width and height of view:

?
something is wrong.

to get row and col is easy the problem is to map that to char id.

The [t] char:

has same row and col of:

however, they have differents char ids.

What do you mean by "char id". Do you mean the offset into some memory buffer?

If your memory buffer is the raw text with CR and LF etc. then you can scan and count the LFs to find the row. And then move forward to find the column. Make sure your column counter does not pass a CR and move into the next line.

yep, offset.

I will need to count? there is not a math way to do that?

Well if your text buffer is free form with CR/LF in random places then no simple mathematics can help you.

i was wondering about make each line with 128 characters in max, so i can calc each position easily. However i don't know if its a good limitation and it will take too many memory space. So i think the way is to count. (although this seems slow).

There are a few things you can do to enhance it.

Starting with your basic linear buffer you can create a table of line lengths for fast line indexing. Of course it requires you to update the index also when updating the text.

However, perhaps the basic rule is appropriate here. Get it working, then get it fast. If, once it is working, you find that the speed is a non-issue then forget about making it faster and instead make more code for other things.