[comp.lang.postscript] Nested Composite Fonts ?

nad@tegra.COM (Nancy Durgin) (10/24/90)

I have a question about Nested Composite Fonts.  This is based on reading
Adobe's document on "Composite Font Extensions" that I got from the
Adobe Server (I have the April 5, 1989 version).

Anyway, on page 8, about nested composite fonts it says:

"1.  If the descendant font that is selected by an 8/8, 1/7, 9/7, or
     SubsVector mapping is itself a composite font, the second
     part of the value extracted from the show string is reused as
     the first part during the descendent's mapping algorithm."

Now, I can understand how this would work for the Byte-based mappings
such as 8/8 and SubsVector, but what happens for the 1/7 and 9/7 mappings?

Suppose I have a 1/7 mapped composite font, each of whose descendents
are also 1/7 mapped?  Do I start dealing with shifting around and
using partial bits of bytes, or what?

For example, if I'm using 1/7 mapping, would I take the first bit to
indicate the font, and then re-use the 7-bits of character plus the
first bit of the next byte, to construct the 1 font bit and 7 char bits
of the actual character?  Or do I mask off the upper bit and reuse the
whole first byte for the second layer of mapping?

Or is it just that nobody in their right mind would use a 1/7 font as
the descendant of a 1/7 font, so it doesn't work at all?

Does anybody out there have any enlightenment on this?

		Nancy

-- 
==============================================================================
Nancy Durgin          | (Usual disclaimers | Tegra-Varityper, Inc. 
tegra!nad@uunet.com   | apply...)          | Billerica, Massachusetts
==============================================================================