Five glyphs: a minor PostScript challenge

[G. Branden Robinson]

Hi folks,

I was wondering if anyone could help with completing, for practical
purposes, the glyph coverage of the special characters listed in
groff_char(7) for the PostScript output device.

We're missing only six glyphs.

One, the Bell Systems logo, I propose to skip over, and let warnings
continue to be issued for any document that attempts to use it,
including our own [1].

The rest _seem_, to my eye wholly untrained in font drawing, like they
might be relatively easy.  As I understand it, the PostScript language
has operators or built-in functions for performing basic transformations
like reflection and skew.

In what I _think_ is increasing order of difficulty, here they are.

A. \[-+]: The minus-plus.  We should be able to dynamically generate
   this from \[+-] by reflecting the latter about a horizontal axis.  If
   the glyph is flipped within its bounding box, I guess the result
   would need to be vertically shifted to align it on the baseline.

B. \[coproduct]: Similar story; we should be able to produce it from the
   product sign, \[product].

C. \[+e]: Variant epsilon.  This is supposed to be a lunate epsilon.  We
   should be able to produce it by right-skewing \[mo], "member of,
   element of a set".

D. \[vA]: Bidirectional vertical double arrow.  Unfortunately,
   overstriking \[dA] and \[uA] doesn't quite work; the arrow stems are
   a bit too long.  Does PostScript have an erasure operator?  What if
   we overstruck the top half of \[uA] with the bottom half of \[dA]?

E. \[.j]: Dotless j.  Assuming the existence of an erasure operator, of
   which I am not certain, the problem here lies in knowing where the
   tittle is to erase it.  It should lie above the x-height of the font.
   If that information is available to the PostScript interpreter, then
   maybe synthesizing this glyph is feasible.

Does anyone have enough PostScript wizardry to achieve any of this?

Regards,
Branden

[1] Not only do I suspect it of being harder to draw than the others,
    but the logo changed over time[2] and all forms may be subject to
    trademark restrictions that would make them a headache to deal with.
    Further, only legacy documents from AT&T would require it, and it's
    just as well if it's missing from any re-renderings a stock groff
    performs of such old documents, groff not being a "pure"/"official"
    troff implementation.

[2] https://www.bell-labs.com/usr/dmr/www/

signature.asc
Description: PGP signature