Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > comp.lang.postscript > #2149 > unrolled thread

kshow for real?

Back to article view | Back to comp.lang.postscript

Contents

  kshow for real? luser- -droog <mijoryx@yahoo.com> - 2015-01-24 00:30 -0800
    Re: kshow for real? Ross Presser <rpresser@gmail.com> - 2015-01-24 00:44 -0800
      Re: kshow for real? luser- -droog <mijoryx@yahoo.com> - 2015-01-24 15:01 -0800
        Re: kshow for real? Dan Coby <adcoby@earthlink.net> - 2015-01-24 15:36 -0800
          Re: kshow for real? luser- -droog <mijoryx@yahoo.com> - 2015-01-24 22:08 -0800
    Re: kshow for real? Phyl <psexton19@gmail.com> - 2015-01-24 16:56 -0800
      Re: kshow for real? luser- -droog <mijoryx@yahoo.com> - 2015-01-25 20:57 -0800
        Re: kshow for real? Phyl <psexton19@gmail.com> - 2015-01-26 06:13 -0800
          Re: kshow for real? luser- -droog <mijoryx@yahoo.com> - 2015-01-26 12:55 -0800

#2149 — kshow for real?

I've only ever used kshow for stupid tricks like color-changing
letters and stair-step effects. I know I'm not using it 
properly because my proc always begins `pop pop`.

So what does "properly" look like?

My only thought (so far) is to encode the two chars so the
table can use a single key for pairs. That might look sthg like

/pair {256 mul add} def
/pairs <<
    (l) 0 get (l) 0 get pair some-number
    (l) 0 get (i) 0 get pair some-other-number-or-the-same-maybe
>> def

50 100 moveto
(lullali-lililullulalala) {
    pair pairs exch 2 copy known {
        get 0 rmoveto
    }{
        pop pop
    } ifelse
} kshow

It's got to be scaled with the fontsize, right?
And probably tuned to a specific face, right?

[toc] | [next] | [standalone]

#2150

On Saturday, January 24, 2015 at 3:30:55 AM UTC-5, luser- -droog wrote:
> I've only ever used kshow for stupid tricks like color-changing
> letters and stair-step effects. I know I'm not using it 
> properly because my proc always begins `pop pop`.
> 
> So what does "properly" look like?
[snip]
> It's got to be scaled with the fontsize, right?
> And probably tuned to a specific face, right?

Since it's intended for *kerning* (that's what the k is for),
definitely it has to be specific to the face and the fontsize.
This page has some samples.
http://www.hegny.de/stefan/pskern-en.html

[toc] | [prev] | [next] | [standalone]

#2151

On Saturday, January 24, 2015 at 2:44:43 AM UTC-6, Ross Presser wrote:
> On Saturday, January 24, 2015 at 3:30:55 AM UTC-5, luser- -droog wrote:
> > I've only ever used kshow for stupid tricks like color-changing
> > letters and stair-step effects. I know I'm not using it 
> > properly because my proc always begins `pop pop`.
> > 
> > So what does "properly" look like?
> [snip]
> > It's got to be scaled with the fontsize, right?
> > And probably tuned to a specific face, right?
> 
> Since it's intended for *kerning* (that's what the k is for),
> definitely it has to be specific to the face and the fontsize.
> This page has some samples.
> http://www.hegny.de/stefan/pskern-en.html

Awesome. So it's not all that complicated. Essentially, the
proc's signature determines the data structure.

    int int  -proc-  -

So we need a 2-key or 2-index map to {some-change-to-state}.

In the link above, he uses an array of {name name offset} tuples.

[
[ /A /C -36 ]
[ /A /Ccedilla -37 ]
[ /A /G -38 ]
[ /A /O -38 ]
[ /A /Odieresis -38 ]
[ /A /Q -38 ]
[ /A /T -50 ]
[ /A /U -29 ]
...
]

That's nice because it's independent of the font encoding,
unlike my chickenscratch upthread (which also had the arguments
in the wrong order).

But the linear data-structure seems slow. I haven't tested,
but I feel in my bones that it'd be better in dictionaries,
maybe even two-deep.

    <<
        /A <<
            << /C -36 >>
            << /Ccedilla -37 >>
            << /G -38 >>
            << /O -38 >>
            << /Odieresis -38 >>
            << /Q -38 >>
           >>
       ...
    >>

[toc] | [prev] | [next] | [standalone]

#2152

On 1/24/2015 3:01 PM, luser- -droog wrote:
> But the linear data-structure seems slow. I haven't tested,
> but I feel in my bones that it'd be better in dictionaries,
> maybe even two-deep.
>
>      <<
>          /A <<
>              << /C -36 >>
>              << /Ccedilla -37 >>
>              << /G -38 >>
>              << /O -38 >>
>              << /Odieresis -38 >>
>              << /Q -38 >>
>             >>
>         ...
>      >>
>

My guess is that you meant:
	<<
	    /A <<
		   /C		-36
		   /Ccedilla	-37
		   /G		-38
		   /O		-38
		   /Odieresis	-38
		   /Q		-38
	       >>
	>>


Dan

[toc] | [prev] | [next] | [standalone]

#2154

On Saturday, January 24, 2015 at 5:37:33 PM UTC-6, Dan Coby wrote:
> On 1/24/2015 3:01 PM, luser- -droog wrote:
> > But the linear data-structure seems slow. I haven't tested,
> > but I feel in my bones that it'd be better in dictionaries,
> > maybe even two-deep.
> >
> >      <<
> >          /A <<
> >              << /C -36 >>
> >              << /Ccedilla -37 >>
> >              << /G -38 >>
> >              << /O -38 >>
> >              << /Odieresis -38 >>
> >              << /Q -38 >>
> >             >>
> >         ...
> >      >>
> >
> 
> My guess is that you meant:
> 	<<
> 	    /A <<
> 		   /C		-36
> 		   /Ccedilla	-37
> 		   /G		-38
> 		   /O		-38
> 		   /Odieresis	-38
> 		   /Q		-38
> 	       >>
> 	>>
> 
> 
> Dan

Yes. Thanks. I plead lyxdexia.

There's more detail in the AFM spec, section 9.2.
http://partners.adobe.com/public/developer/en/font/5004.AFM_Spec.pdf

And .afm files seem to be easy to find
   find / -name \*.afm

And gs conveniently reports the files when it loads a 
font. So, all the pieces seem to be there to hack 
something together.

[toc] | [prev] | [next] | [standalone]

#2153

On Saturday, January 24, 2015 at 2:30:55 AM UTC-6, luser- -droog wrote:
> I've only ever used kshow for stupid tricks like color-changing
> letters and stair-step effects. I know I'm not using it 
> properly because my proc always begins `pop pop`.
> 
> So what does "properly" look like?
> 
> My only thought (so far) is to encode the two chars so the
> table can use a single key for pairs. That might look sthg like
> 
> /pair {256 mul add} def
> /pairs <<
>     (l) 0 get (l) 0 get pair some-number
>     (l) 0 get (i) 0 get pair some-other-number-or-the-same-maybe
> >> def
> 
> 50 100 moveto
> (lullali-lililullulalala) {
>     pair pairs exch 2 copy known {
>         get 0 rmoveto
>     }{
>         pop pop
>     } ifelse
> } kshow
> 
> It's got to be scaled with the fontsize, right?
> And probably tuned to a specific face, right?

Since moving to Windows7 my posscipt is filled with kshow commands I did not change my postscript driver. It looks like it is trying to to variable spacing in my text which makes my user complain. Currently I change all the kshows for my (test) to shows and delete the array.  How do I go back to getting the plain show?

[toc] | [prev] | [next] | [standalone]

#2155

On Saturday, January 24, 2015 at 6:56:23 PM UTC-6, Phyl wrote:
> 
> Since moving to Windows7 my posscipt is filled with kshow commands I did not change my postscript driver. It looks like it is trying to to variable spacing in my text which makes my user complain. Currently I change all the kshows for my (test) to shows and delete the array.  How do I go back to getting the plain show?

What's the source application? Did its version 
change with the upgrade? 

[toc] | [prev] | [next] | [standalone]

#2156

On Sunday, January 25, 2015 at 10:57:05 PM UTC-6, luser- -droog wrote:
> On Saturday, January 24, 2015 at 6:56:23 PM UTC-6, Phyl wrote:
> > 
> > Since moving to Windows7 my posscipt is filled with kshow commands I did not change my postscript driver. It looks like it is trying to to variable spacing in my text which makes my user complain. Currently I change all the kshows for my (test) to shows and delete the array.  How do I go back to getting the plain show?
> 
> What's the source application? Did its version 
> change with the upgrade?

The source application is a UK product Teleform which has been upgraded recently. I have also tried it in Word, which is a new version for me also.  Is it based on my choice of fonts?

[toc] | [prev] | [next] | [standalone]

#2157

On Monday, January 26, 2015 at 8:13:54 AM UTC-6, Phyl wrote:
> On Sunday, January 25, 2015 at 10:57:05 PM UTC-6, luser- -droog wrote:
> > On Saturday, January 24, 2015 at 6:56:23 PM UTC-6, Phyl wrote:
> > > 
> > > Since moving to Windows7 my posscipt is filled with kshow commands I did not change my postscript driver. It looks like it is trying to to variable spacing in my text which makes my user complain. Currently I change all the kshows for my (test) to shows and delete the array.  How do I go back to getting the plain show?
> > 
> > What's the source application? Did its version 
> > change with the upgrade?
> 
> The source application is a UK product Teleform which has been upgraded recently. I have also tried it in Word, which is a new version for me also.  Is it based on my choice of fonts?

That could well be it. Kerning data is optional in font files.
So, perhaps the application has always had this ability and
suddenly found the kerning data it needed.

[toc] | [prev] | [standalone]

Back to top | Article view | comp.lang.postscript

csiph-web