CMAP and IL2 encoding

CMAP and IL2 encoding

[Vit Zyka]

[-- Attachment #1: Type: text/plain, Size: 2211 bytes --]

Nice Sunday evening Hans,

I played with ToUnicode resource (for correct copy&paste and search 
opperation in PDF). I noticed a package cmap.sty was done by Mr. 
Volovich for LaTeX a year ago; I did one (cmap.tex) for plain TeX 
yesterday and started to be interested in the situation in ConTeXt. No 
wonder you did it 4 years ago. But I met some minor problems:

1) if you run example file
     \useencoding[pfr]
     \setupbodyfont[csr]

     \starttext
       test \`z \'z \bf test \sl test \bs quite funny \`z \page
       test \`z \'z \bf test \sl test \bs quite funny \`z \page
     \stoptext
   you obtain /ToUnicode resource ONLY for first roman variant. It is due
   to a condition in the file enco-pfr.tex
     \def\doincludepdffontresource
      {\ifcsname\s!ucmap\fontfile\endcsname\else
         \dodoincludepdffontresource
       \fi}
   The condition is correct but the \fontfile is not changed when \bf,
   \it, etc. is used (it is still cscsc10 (?)).

   Might you look at a proper \fontfile definition?

2) Technical note: in ConTeXt the il2 encoding is defined. But the
    definition is not ISO-8859-2 (Latin 2, IL2) encoding sharp. The
    definition is CS font encoding (IL2cs or 8z). It is derived from il2
    but it differ in 49 positions (!). For
    usage in Czech or Slovac languages it causes no problem but what if
    somebody will want to use a proper il2 font?
    If you want I can prepare
    both IL2 and (a bit richer) il2cs (8z?) enco-??.tex files.

3) File pdfr-il2.tex with CMAP definition is proper IL2 mapping. So now
    we combine IL2 CMAP mapping with IL2cs ConTeXt encoding. It is
    nosence. Suggested solution: use csr fonts with new IL2cs encoding
    and create new pdfr-il2cs.tex.

4) Definitions of CMAP encodings are stored in pdfr-??.tex files - it
    means in TeX format. But the common way is to be a stand-alone PS
    file, see attachement. Now I know about 5 files from Mr. Volovich
    (Cork, 3 russian, vietnamese) and two from me (IL2, IL2cs). Such file
    can be easily included by \pdfobj stream file {} mechanism.
    Question: I can prepare pdfr-il2cs.tex but is it not better to
    use more standard stand-alone files in ConTeXt?

Vit Zyka

[-- Attachment #2: il2.cmap --]
[-- Type: application/postscript, Size: 1886 bytes --]

[-- Attachment #3: il2cs.cmap --]
[-- Type: application/postscript, Size: 1846 bytes --]

[-- Attachment #4: Type: text/plain, Size: 139 bytes --]

_______________________________________________
ntg-context mailing list
ntg-context@ntg.nl
http://www.ntg.nl/mailman/listinfo/ntg-context

Re: CMAP and IL2 encoding

[Vit Zyka]

Some auto-answers ;-)

Add 1: I thing the answer is waiting only for Hans.
Add 2: il2 name is all right for encoding of csr fonts in ConTeXt.
        I was confused of what is called il2. il2 in TeX world is
        encoding of CS fonts (csr). Do not confuse with ISO-8859-2 so
        called Latin 2.
        Never mind, in enco-il2.tex are still missing some
        not-so-important chars.
Add 3: Definition of pdfr-il2.tex (CMAP) is for ISO-8859-2 fonts so it
        is not suited for csr font (I believe it was the first [and
        perhaps last also] intention).
Add 4: Decision for ConTeXt developers(?) According to the decision I
        can prepare the CMAP files for il2, xl2, cp1250, and ISO-8859-2
        encodings.

Vit Zyka

> I played with ToUnicode resource (for correct copy&paste and search 
> opperation in PDF). I noticed a package cmap.sty was done by Mr. 
> Volovich for LaTeX a year ago; I did one (cmap.tex) for plain TeX 
> yesterday and started to be interested in the situation in ConTeXt. No 
> wonder you did it 4 years ago. But I met some minor problems:
> 
> 1) if you run example file
>     \useencoding[pfr]
>     \setupbodyfont[csr]
> 
>     \starttext
>       test \`z \'z \bf test \sl test \bs quite funny \`z \page
>       test \`z \'z \bf test \sl test \bs quite funny \`z \page
>     \stoptext
>   you obtain /ToUnicode resource ONLY for first roman variant. It is due
>   to a condition in the file enco-pfr.tex
>     \def\doincludepdffontresource
>      {\ifcsname\s!ucmap\fontfile\endcsname\else
>         \dodoincludepdffontresource
>       \fi}
>   The condition is correct but the \fontfile is not changed when \bf,
>   \it, etc. is used (it is still cscsc10 (?)).
> 
>   Might you look at a proper \fontfile definition?
> 
> 2) Technical note: in ConTeXt the il2 encoding is defined. But the
>    definition is not ISO-8859-2 (Latin 2, IL2) encoding sharp. The
>    definition is CS font encoding (IL2cs or 8z). It is derived from il2
>    but it differ in 49 positions (!). For
>    usage in Czech or Slovac languages it causes no problem but what if
>    somebody will want to use a proper il2 font?
>    If you want I can prepare
>    both IL2 and (a bit richer) il2cs (8z?) enco-??.tex files.
> 
> 3) File pdfr-il2.tex with CMAP definition is proper IL2 mapping. So now
>    we combine IL2 CMAP mapping with IL2cs ConTeXt encoding. It is
>    nosence. Suggested solution: use csr fonts with new IL2cs encoding
>    and create new pdfr-il2cs.tex.
> 
> 4) Definitions of CMAP encodings are stored in pdfr-??.tex files - it
>    means in TeX format. But the common way is to be a stand-alone PS
>    file, see attachement. Now I know about 5 files from Mr. Volovich
>    (Cork, 3 russian, vietnamese) and two from me (IL2, IL2cs). Such file
>    can be easily included by \pdfobj stream file {} mechanism.
>    Question: I can prepare pdfr-il2cs.tex but is it not better to
>    use more standard stand-alone files in ConTeXt?
> 
> Vit Zyka

-- 
Vít Zýka --- http://typokvitek.com
     Automatic document typesetting
     Automaticka sazba dokumentu --
--