From Newsgroup: comp.lang.postscript

Mac font Calibri has glyphs for old-style numbers. Mac Font Book claims that they are glyphs 954 to 963, it not being obvious what those numbers mean. Best efforts with glyph names resembling /oneoldstyle, rCa, /threeoldstyle, rCa, produce the notdef block. :-(
Please, how can these non-Unicode glyphs be accessed from PostScript?
--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.postscript

In comp.lang.postscript, jdaw1 <jdawiseman@gmail.com> wrote:

Mac font Calibri has glyphs for old-style numbers. Mac Font Book
claims that they are glyphs 954 to 963, it not being obvious what
those numbers mean. Best efforts with glyph names resembling
/oneoldstyle, rCa, /threeoldstyle, rCa, produce the notdef block. :-(

Please, how can these non-Unicode glyphs be accessed from PostScript?

I'm not enough of a Postscript expert to help here, but such glyphs have
a bunch of names:

https://en.wikipedia.org/wiki/Text_figures

Text figures (also known as non-lining, lowercase, old style,[1]
ranging, hanging, medieval, billing,[2] or antique[3] figures or
numerals) are numerals designed with varying heights in a fashion
that resembles a typical line of running text, hence the name. They
are contrasted with lining figures (also called titling or modern
figures), which are the same height as upper-case letters.[4][5]
Georgia is an example of a popular typeface that employs text
figures by default.

I generally think of them as "lower case" numbers, and would hunt using
names like that.

Elijah
------
found that wikipedia page using "lower case numbers" as the search
--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.postscript

Agree with everything you have said. But PLRM3, Appendix E.9 (rCLExpert Character SetrCY), pp788rCo789, lists: centoldstyle, dollaroldstyle, eightoldstyle, fiveoldstyle, fouroldstyle, nineoldstyle, oneoldstyle, sevenoldstyle, sixoldstyle, threeoldstyle, twooldstyle, zerooldstyle. So those ought to be valid glyph names.
Unicode rCo wrongly rCo decided not to have separate characters for old-style numbers. Could that be the problem? Could it be that, somehow, macOS 10.13.6 doesnrCOt allow PostScript to access non-Unicode characters?
--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.postscript

On Monday, October 5, 2020 at 2:33:40 AM UTC-5, jdaw1 wrote:

Agree with everything you have said. But PLRM3, Appendix E.9 (rCLExpert Character SetrCY), pp788rCo789, lists: centoldstyle, dollaroldstyle, eightoldstyle, fiveoldstyle, fouroldstyle, nineoldstyle, oneoldstyle, sevenoldstyle, sixoldstyle, threeoldstyle, twooldstyle, zerooldstyle. So those ought to be valid glyph names.

Unicode rCo wrongly rCo decided not to have separate characters for old-style numbers. Could that be the problem? Could it be that, somehow, macOS 10.13.6 doesnrCOt allow PostScript to access non-Unicode characters?

Can you get the available names from the CharStrings dictionary?
Sthg like
/Calibri findfont /CharStrings get { pop ==only ( )print } forall
If so then you can patch the encoding vector
/Calibri fontfont dup dup /Encoding get dup length array copy
(0) 0 get
[ /zero /one /two ... ] putinterval
put
--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.postscript

On Monday, October 5, 2020 at 10:00:43 AM UTC-5, luser droog wrote:

On Monday, October 5, 2020 at 2:33:40 AM UTC-5, jdaw1 wrote:

Agree with everything you have said. But PLRM3, Appendix E.9 (rCLExpert Character SetrCY), pp788rCo789, lists: centoldstyle, dollaroldstyle, eightoldstyle, fiveoldstyle, fouroldstyle, nineoldstyle, oneoldstyle, sevenoldstyle, sixoldstyle, threeoldstyle, twooldstyle, zerooldstyle. So those ought to be valid glyph names.

Unicode rCo wrongly rCo decided not to have separate characters for old-style numbers. Could that be the problem? Could it be that, somehow, macOS 10.13.6 doesnrCOt allow PostScript to access non-Unicode characters?

Can you get the available names from the CharStrings dictionary?
Sthg like

/Calibri findfont /CharStrings get { pop ==only ( )print } forall

If so then you can patch the encoding vector

/Calibri fontfont dup dup /Encoding get dup length array copy
(0) 0 get
[ /zero /one /two ... ] putinterval

/Encoding exch

put

--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.postscript



On 10/04/2020 01:14 PM, jdaw1 wrote:

Mac font Calibri has glyphs for old-style numbers. Mac Font Book claims that they are glyphs 954 to 963, it not being obvious what those numbers mean. Best efforts with glyph names resembling /oneoldstyle, rCa, /threeoldstyle, rCa, produce the notdef block. :-(

Please, how can these non-Unicode glyphs be accessed from PostScript?

What software are you using?

Calibri is a Microsoft proprietary font from Microsoft Office and
apparently a lot of people replace it with Carlito from Google which is
freely licensed.

PostScript uses a font encoding vector to map glyphs to a byte value
with a range of 0 to 255 decimal (or 0 to 377 octal) so I don't
understand what 954 to 963 means.

Jeff Coffield
www.digitalsynergyinc.com
--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.postscript

On Monday, October 5, 2020 at 10:05:15 AM UTC-5, Jeffrey H. Coffield wrote:

On 10/04/2020 01:14 PM, jdaw1 wrote:

Mac font Calibri has glyphs for old-style numbers. Mac Font Book claims that they are glyphs 954 to 963, it not being obvious what those numbers mean. Best efforts with glyph names resembling /oneoldstyle, rCa, /threeoldstyle, rCa, produce the notdef block. :-(

Please, how can these non-Unicode glyphs be accessed from PostScript?

What software are you using?

Calibri is a Microsoft proprietary font from Microsoft Office and
apparently a lot of people replace it with Carlito from Google which is freely licensed.

PostScript uses a font encoding vector to map glyphs to a byte value
with a range of 0 to 255 decimal (or 0 to 377 octal) so I don't
understand what 954 to 963 means.

Jeff Coffield
www.digitalsynergyinc.com

Right. A Type 1 font is a dictionary with the /Encoding member holding
a 256 element array of glyph names. These names should be keys in the /CharStrings member which is a dictionary mapping glyph names to encoded/encrypted drawing commands.
If it's a True Type font, the ghostscript does some magic to wrap it
in a Type 42 font and I don't know the details of how or to what
extent it populates the /CharStrings dictionary.
--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.postscript

Code simplified to
/Calibri findfont /CharStrings get {pop ==} forall
Which, embarrassingly, has no threeoldstyle. But! But PLRM3, Appendix E.9! Sigh.
Nor does Carlito (good suggestion anyway rCo thank you).
Re encodings, if one is willing to access through glyphshow, generally there's no need to mess with encodings. But it does matter if one wants a charpath effect for a glyph, as I asked in May 2006:
http://groups.google.com/g/comp.lang.postscript/c/rvAGfbsbLVw
--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.postscript

With further rummaging, I still cannot find a glyph name. Sigh.
--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.postscript

In article <980c54c3-8027-4050-af0c-c358e9661db0n@googlegroups.com>, jdawiseman@gmail.com says...

With further rummaging, I still cannot find a glyph name. Sigh.

PostScript doesn't support TrueType fonts (directly), you can use Type42 fonts, which are TrueType outlines wrapped up with additional
information.

You could start with this Stack Overflow question:

https://stackoverflow.com/questions/63743952/showing-glyphs-with- unicodes-higher-than-decimal-256


Ken

--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.postscript

On Wednesday, October 7, 2020 at 4:51:44 PM UTC-4, jdaw1 wrote:

With further rummaging, I still cannot find a glyph name. Sigh.

I downloaded Carlito-Regular.ttf from here https://fontlibrary.org/en/font/carlito

and opened it in Fontforge and found:

65645 0x1006d U+???? "three.oldstyle"

and

65635 0x10063 U+???? "three.pnum.oldstyle"

But I guess these are not Postscript glyphnames but somehow stored in
the TTF, because this

(Carlito-Regular.ttf) findfont /CharStrings get { pop ==only ( )print } forall

did not emit those names.

There's apparently an enormous amount I don't know about fonts.

--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.postscript

And MyriadPro-Regular works for me, so my problem is solved, even though the general problem remains difficult.
--- Synchronet 3.21d-Linux NewsLink 1.2