Denis Oliver Kropp wrote: > Phil Endecott wrote: >>>> I am slowly working through all the places where small changes will be >>>> needed to work with my text rotation feature. >> >> At present, the logical_rect returned by GetStringExtents can have >> negative width (or height?) for non-left-to-right scripts. Is this a >> useful feature, or is it just how it worked out? >> >> I have to decide whether a rotated (e.g. upside-down) font should >> return a rectangle with negative dimensions in the same way, or not. >> >> The easiest thing for me to do is to always return a "valid" rectangle >> with positive height and width: I simply find the coordinates of the >> four corners of the string and use min and max to get the bounding >> box. But this would change the current behaviour for right-to-left scripts. >> >> Does anyone rely on the current behaviour of GetStringExtents? > > Yes, it's important that you can do this: > > 1) DrawString( 10,10 ) > 2) GetStringExtents( &rect ) > 3) region = translate( rect, 10, 10 ) > 4) Flip( ®ion ) > > In this case ink_rect would be used, but logical_rect should not have a > different orientation or offset.
But at present, with fonts for right-to-left scripts, logical_rect can return a rectangle with negative width. If you used that in your example above then the region passed to Flip would still have a negative width, which I guess it would not like (right? - at least, it should be equally happy with an equivalent rectangle with positive dimensions). So I think that that is an example of something that would benefit from the change that I am proposing. (Or maybe I am missing something.... it's late!) Phil. _______________________________________________ directfb-dev mailing list [email protected] http://mail.directfb.org/cgi-bin/mailman/listinfo/directfb-dev
