[pdftex] AR 5 not working with pdftex?

Tom Kacvinsky tjk at ams.org
Mon Jun 4 11:54:45 CEST 2001

I am sort of happy to say that, yes indeed, the gaps in the Subrs array is
causing printing problems.  Happy, because I am now convinced that this is
indeed the problem, but not happy, because Acrobat used to *not* do this, and
there was no ChangeLog type entry for this modification.

Incidentally, the problem is this:

If the Type 1 font inthe PDF file has gaps in the Subrs array, the font is not
even embedded in the PS stream sent to the printer.  It is *not* the case that
the font is sent unaltered and the file does not print because of bugs in the
printer's interpreter.

Finally, I have discovered if PDF file has a font which has no Encoding object
associated with it, Acrobat builds the encoding vector in the PS stream from the
Type 1 font's encoding vector, BUT IT STRIPS ENTRIES 0 -- 31.  Unfortunately,
changing the Flags entry in the FontDescriptor object to 12 (for CMR10, treat it
as seriffed and as symbolic) did not help.  I thought that changing the flags so
that Acrobat thought the font was symbolic would keep it from stripping the
duplicate entries.  No go...

This is bad, because the page objects actually use codes 0 -- 31.  Oops!

However, if the font has an Encoding object assocaited with it, it uses that to
build the correct encoding vector for the PS file (that is, all of the char
codes used on the page are encoded).

So there are "bugs" with both pdftex *and* Acrobat.


On Mon, 4 Jun 2001, Tom Kacvinsky wrote:

> This is because slots 0 -- 31 are not encoded in the PS file generated by
> Acrobat Reader 5.0.  At least this is the case in the PS file you sent me.
> As for the subsetted fonts and hinted glyphs -- that *may* have something to do
> with gaps in the Subrs array.  The one test file you sent me used char code 13,
> which is one of the char codes that gets "deleted" in the PS file, so it would
> not print regardless of hint replacement (or no hint replacement).  I would need
> to see a file that uses a glyph which uses hint replacement and has char code
> above 31.
> Finally, the Subrs problem may very well have something to do with your printer.
> I have never had a problem with printing a PS file which uses a Type 1 font in
> which the Subrs array had gaps, but I use HP 5Sis or Xerox N2125s.
> On Mon, 4 Jun 2001 rhowlett at mail.usyd.edu.au wrote:
> > I think now the printing problem when one embeds the font file cmr10.pfb
> > entirely is somewhat unrelated to the problem with the hinted characters,
> > which may indeed be something to do with gaps in the Subrs array
> > when the font is subsetted. It seems that when I embed the font entirely
> > and then print to a ps printer from AR 5.0 (under Windows98), it's the
> > characters 0 to 31 that are missing. The others are OK, hinted and
> > unhinted alike. This is not what I thought I witnessed before -- I guess
> > I was mistaken (although that was on a different machine ...)
> >
> _______________________________________________
> pdftex mailing list
> pdftex at tug.org
> http://tug.org/mailman/listinfo/pdftex

More information about the pdftex mailing list