Path: utzoo!utgpu!news-server.csri.toronto.edu!rutgers!uwm.edu!rpi!zaphod.mps.ohio-state.edu!usc!cs.utexas.edu!chinacat!woody
From: woody@chinacat.Unicom.COM (Woody Baker @ Eagle Signal)
Newsgroups: comp.lang.postscript
Subject: Re: Character side bearings
Summary: Sidebearings.
Message-ID: <1405@chinacat.Unicom.COM>
Date: 18 Jul 90 02:14:28 GMT
References: <2259@dino.cs.iastate.edu>
Distribution: usa
Organization: a guest of Unicom Systems Development, Austin
Lines: 41

In article <2259@dino.cs.iastate.edu>, hascall@cs.iastate.edu (John Hascall) writes:
>    How does one determine the left side bearing of a character (in a
> 
>    charpath flattenpath pathbbox
> 
> may be used to determine character bounding boxes and side bearings."
> 
>    Well, I can see how it gives you the bounding box, but how do you
> get the (left) side bearing?


Well, it turns out that the lower left x coordinate IS the left sidebearing.
Since bounding box just encloses the character, if the character is offset
from the current position (i.e. the leftmost point of the character) either
positive or negative, it will show up as a positive or negative number, which
is the sidebearing.  After struggling for better than 9 months, and after
not getting replies to several inquires, I am now prepared to state that the
RED book appears to be FLAT WRONG when it says that the metrics entry can be
an array, that includes the left sidebearing and the width.  The width
part works, but regardless of what you put in the other element, it does not
work.  Adobe has a major gap between documentation and implemtation.  IF
some one can prove me wrong on this, I'll gladly retract this statement.
To do so, send me or post a fragment of code that clearly demonstrates teh
altering of the sidebearing.  Print some letters in say Times Roman, 
especially 'j' and 'l' and 'W', then alter the the width, and print them
again, then alter the sidebearing and the width and print them again.

The application is monospacing the proportional fonts.  By measuring the
width of say a "W", and taking that to be the cell width, computing the
bounding box for a character, subtracting it from the cell width, dividing
the result by 2, one can get an offset that you need to shift the left
sidebearing by inorder to make the character sit squarely in the "cell"
width.  In addition, you have to set the character width to the value
computed from "W".  If you just change the character widths to the value
obtained by "W", and then print the text, it becomes very obvious that
while the "escapement" of the characters has been changed, the characters
are NOT centered in the cell.  Computing the diffrence, and altering the
sidebearing should fix the problem.  I have not gotten it to work,
inspite of many hours (so far over 20) of hacking at it, and working on it.
Cheers
Woody