#PFM Search
13 messages in this thread
I haven't got such .PFMs, Gary, but I _am_ looking for a guide to the .PFM
file format. Any ideas where I might find such a thing?
Mike (Whapping from the UK)
Mike:
The only place I can think of looking is Adobe (though you would think
Microsoft would have some docs in their forum… but NO!!!…. no wonder
IBM wants to leap out of bed with those guys). If I can find the time (and
I'd better for my next project!), I'll call Adobe and see if they have some
docs on PFM's. Could also bug SoftLogik. They have a converter or some such
thing. There's also a couple of IBM AFM->PFM converters here on CIS. I
could try to bug the authors for the docs, but you know how programmers
are…:)
Gary
– Mirror Image Prod. (416) 495-7469, via WHAP!
Thanks, Gary … I've hunted high and low for the .PFM specs, but can't
find them anywhere. I might have a look at the AFM->PFM converters in the
IBM areas … I could always try and reverse-engineer them (or use them to
do the conversions on the PCs I now have in my office).
Mike (Whapping from the UK)
Gary, don't forget that there is that one Amiga PFM to AFM converter on the
SL BBS with C source code. It doesn't work well, but will give you a head
start.
Mike @ SL
I didn't know about that one … unfortunately, the SL BBS is just too
expensive to access from the UK 🙁 … I had hoped that this was the type
of file that SL might upload to the AmigaVendor area.
Mike (Whapping from the UK)
In the future I'll be uploading files on a more regular basis. But I do CIS
support from Canada, and I have to long distance dload it from the SL BBS,
then upload it to CIS over long-distance again because I don't have a local
node. I'll be moving in the late spring and things will change then.
Mike
Mike,
Look for a file called FMDTP2.LZH. Its in AmigaArts, I believe. I think
it does what you want and was actually written by another "Red Coat," ;^)
B. Aylward author of Post fame.
BTW, you need at least the post.library portion of the Post PostScript
interpreter program which is also here on CIS as POST16.LZH, also in
AMigaArts, I believe. Those files are roughly 89K and 200K so they take a
while to DL.
You might also need conv.arc which converts .PFB files to something that
can be downloaded to a PostScritp printer. As I recall, the output of the
unpacked .PFB is used by the routines in FMDTP2 to create .metric files for
ProPage and amiga bitmap files.
wmc
Mike, Your looking for a way to generate .PFM files? If you can't find a
way to do it on the Ami or IBM and have AMAXII, look at a program called
Metamorphosis Professional. It converts MAC printer fonts into IBM versions
and includes a .PFM and .AFM file in the conversion. I've been deleting the
.PFMs 'cause I saw no use for them on the Amiga. BTW, what is your need for
them?
Chris
Actually, it's the other way around … I'd like to create complete .AFMs
from
PFMs. Generating pure character bounding info for the .AFM is easy, it's
getting the other info (kerning etc) from the .PFMs which I want to do.
Sounds like some of the other suggestions may have the solution for me.
Mike (Whapping from the UK)
Mike,
What were the other suggestions that may have the solution for you
in regards to generating .AFMs from .PFMs?
Chris
Mike:
From what I've discovered, you're going to have to include the Type 1 in
your AFM from PFM utility if you expect it to work flawlessly. The PFM's do
not contain character name information, for one thing; the whole format is
based on the IBM character layout (so why are IBM fonts are becoming the
Amiga standard?!?), so you'd be screwed if you tried converting a symbol
font or one of the Adobe Expert series, which use non-standard character
names. Where would they go in the PFM? As for the bounding boxes: that's
the only thing about the AFM's I can't figure out! I don't think what you
think they are is what they are, but I could be wrong about that… in any
case, if you know how to derive them, let me know. The kerning information
is a breeze; the PFMs store it just like the Mac, as a pair of single bytes
(the first byte is the first char in the pair, the second byte is the
second char), followed by a 2-byte word describing the actual PostScript
value of the kern (negative is backwards kern, pos is forwards). In any
case, the whole idea for converting PFM's to AFM's seems kind of silly to
me, at this point. If SoftLogik has decided to support IBM Type 1 fonts
only, you would think that they would accomodate the PFMs, since that's all
that people are using anymore on IBMs. And what the heck do they need from
the AFM, anyway, that they can't get from the actual Type 1 (except
kerning, which is painfully simple to extract from the PFM, so we're back
to square one again…) Anyway, maybe if you want to share resources, we
can write a PFM-AFM converter for the Amiga and get our names in the
papers! (BTW, the method for storing kerning information in PFMs is
compact, but rather bulky and slow to access, unless a program using it
translates it to an internal table. I like the way PPage's metric file's do
it, because you can find a kerning pair almost instantly, except that it
takes up alot (relatively speaking) of disk space. I developed a way for
storing kerning information that is as compact as the Mac/PFM method, but
requires
[ MORE ]
[ continuation ]
only 2 jumps in the file to find out about any kerning pair value. But it
uses string arrays, and I guess that sort of thing is frowned upon in the
Land of Programmers…:)
Gary
– Mirror Image Prod. (416) 495-7469, via WHAP!