Re: SDATA entity mapping

Subject: Re: SDATA entity mapping
From: James Clark <jjc@xxxxxxxxxx>
Date: Wed, 21 May 1997 16:38:10 +0700
At 10:26 21/05/97 +0200, Dylan van Rijsbergen wrote:

>>2. Is it possible to force Jade to use an 8 bit character set and map these
>>>entities to letters in this character set? 
>>I don't know what you mean.  Are you talking about input or output?
>I am talking about output: I want to map my SDATA entities to the correct

It doesn't make much sense to talk about forcing Jade to use an 8-bit
character set in the context of output.

The appropriate strategy depends on what type of font you've got. To find
out, get and then do

ttfdump -nx -tcmap yourfont.ttf >yourfont.txt

Then look at yourfont.txt.  If it's a Windows TrueType font, the cmap will
include a subtable with platform id 3.  The specific id will either be 0 or 1.

- 0 means it's a symbol font and the cmap will have characters in the range
F020 - F0FF.  If your font does this, it's basically broken (it should be
using a a specific id of 1) but not hopelessly so. The best thing to do at
the moment is to explicitly specify the font and use characters of the form
"\U-F0NN" (ie F000 more than what you probably have been thinking the
character code is).  If you don't add in the F000, it will work OK at the
moment, but will probably stop working when the RTF backend implements more
sophisticated font mapping machinery.

- 1 means it's a text font with a Unicode encoding.  If your font includes a
glyph for the Unicode character 1F10 then the cmap should have an entry for
1F10; if it does, all is well and you should be able to use the font without
problems with Jade.  However it may not (especially if it is an amateur
font) and may just have entries for 0000 to 00FF.  If so, you should
complain to your font vendor: you will always have to resort to hacks to use
such a font with Jade.


 DSSSList info and archive:

Current Thread