[metapost] Strange infont results

Daniel H. Luecking luecking at uark.edu
Mon Jul 11 18:05:26 CEST 2005

On Mon, 11 Jul 2005, Stephan Hennig wrote:

> Daniel H. Luecking schrieb:
> > Just to point out: pcrr8r (note the last "r") is not Adobe Standard
> > Encoding. It is "TeXBase1Encoding" and (for some characters) a
> > re-encoding is required. Metapost doesn't do re-encoding,
> Admittedly, I'm a bit lost within all the different stages of font
> handling and procesing. So I've taken this as a warning using pcrr8r in
> MetaPost is not save in general. But would pcrr mean pcrr8a or pcrr8r or
> something else?

For a font xyzw, MetaPost requires
 1. A file xyzw.tfm file
 2. If prologues > 0, a line in psfonts.map for xyzw.

When the encoding part of a TeX fontname is missing, it means whatever
the original encoding of the font. For Courier, that should be Adobe
Standard. So pcrr.tfm should be the same as pcrr8a.tfm if they both
exist. I just discovered that my system (TeXLive2004) has only pcrr.tfm.

Most recent psfonts.map files don't have a line for either pcrr or

> > though dvips
> > and pdftex will if the MP output is included in a .tex document.
> Would it be safe to process a file by mptopdf, too?

Yes. What that does is run pdftex with a special format and macros.

> > Courier in the Adobe standard encoding is pcrr8a, but using that will
> > not bring up Courier if prologues is > 0 because there is no name
> > mapping in psfonts.map for pcrr8a. It would also likely cause dvips and
> > pdftex problems as well.
> Are there good reasons not to include those map file entries by default?

I can't think of a reason, except that psfonts.map was prepared for
dvips, which is always used with TeX, and so the nonTeX-related
encodings would be deprecated.

In the circumstances of the OP I would always use btex...etex and
include a LaTeX preamble for my chosen font. That way one can use any
available font package. Also ligatures and kerning pairs will be
available (for the guillemots he was apparently trying to obtain) and
spaces would remain spaces (and not some odd symbol occupying position
32 of the font).


Dan Luecking
Dept. of Mathematical Sciences
University of Arkansas
Fayetteville, AR 72101

More information about the metapost mailing list