Path: utzoo!news-server.csri.toronto.edu!cs.utexas.edu!usc!snorkelwacker.mit.edu!bloom-beacon!dont-send-mail-to-path-lines
From: mouse@lightning.mcrcim.mcgill.EDU (der Mouse)
Newsgroups: comp.windows.x
Subject: Re:  naive query
Message-ID: <9103040614.AA08270@lightning.McRCIM.McGill.EDU>
Date: 4 Mar 91 06:14:56 GMT
Sender: daemon@athena.mit.edu (Mr Background)
Organization: The Internet
Lines: 42

> (Since I'm much too sleepy to find and wade through the source...)

> Does anyone know exactly the algorithm which the XDrawString
> functions use to draw one character after another?  Specifically, how
> the horiz. position of each subsequent character is determined.

Do you want to know how they *actually* do it or how they're *supposed*
to do it? :-)

In the Xlib document, in the discussion of the font structures, such as
XCharStruct, I find

	For a character origin at [x,y], the bounding box of a
	character (that is, the smallest rectangle that encloses the
	character's shape) described in terms of XCharStruct components
	is [...].

	The origin for the next character is defined to be:
	        [x + width, y]

In other words, you simply step along the X axis by adding the width
field of each character's XCharStruct after drawing that character's
glyph.

As for how they *actually* do it....  XTextWidth and XTextExtents
really do do things this way, at least in the MIT source.  A quick scan
of the mfb code gives me the impression that it does as well, though to
be sure would require a much more careful scan.  (And note that not all
servers use mfb, of course.)

> I've written a small chunk of code which draws text in either a
> line-by-line or character-by-character fashion...the problem is that
> character-by-character just doesn't get the kerning right, and the
> text drawn this way is always slightly off from when the text would
> be drawn in one shot as a line.

If the above doesn't help, I'd have to see the code.

					der Mouse

			old: mcgill-vision!mouse
			new: mouse@larry.mcrcim.mcgill.edu