On 2004-04-03 12:20:50 +0200, Hartmut Henkel wrote:
Until now, the base-14 fonts are _never_ embedded, unless one tweaks the BaseName entry in the 2nd mapline field. These BaseName entries are compared against a list of the base-14 fontnames, and if they match one of these, the fontfile is ignored.
I have changed this check in a small experiment so that the following happens:
If a mapline contains a BaseName from the base-14 set, e. g. Helvetica, the further treatment depends on whether a fontfile name is given or not: If the fontfile name (e. g.
If there is a mapline with a BaseName not from the base-14 set, and no fontfile is given, the usual warning is issued.
So one can decide by telling or omitting a fontfile name, whether a font from the base-14 set is embedded or not (which seems to me a rather natural way). No new switch or primitive would be needed. And the usual behaviour could be reached by editing the file pdftex.map, or by making a small modification to the updmap script.
I have separated the flag is_basefont() and is_included() in file mapfile.c, so that one now can distinguish between all these cases.
Attached is a tgz-file with the patch to mapfile.c, and a some test files, which also exercise a few cases of embedding things. Just rough stuff, lots of more tests would be needed to fully understand all embedding cases. What do you think?
Sounds very interesting and like a very natural user interface. But i think it's much too late for 1.20a. Best regards Martin -- Martin Schröder, ms@artcom-gmbh.de ArtCom GmbH, Lise-Meitner-Str 5, 28359 Bremen, Germany Voice +49 421 20419-44 / Fax +49 421 20419-10 http://www.artcom-gmbh.de
On Mon, 5 Apr 2004, Martin Schröder wrote:
On 2004-04-03 12:20:50 +0200, Hartmut Henkel wrote:
Attached is a tgz-file with the patch to mapfile.c, and a some test files, which also exercise a few cases of embedding things...
Sounds very interesting and like a very natural user interface. But i think it's much too late for 1.20a.
Sure nothing for 1.20. And I had overseen that also writefont.c would need some tweaking; currently the patch does more or less nothing :-) Regards, Hartmut
participants (2)
-
Hartmut Henkel
-
Martin Schröder