Re: [docbook-apps] Alternatives to MS Arial Unicode for PDF=?UTF-8?Q?output=3F?=

[maxwell <maxwell@umiacs.umd.edu>]

On Fri, 8 Jul 2011 12:05:56 EDT, DeanNelson@aol.com wrote:
> We use DejaVu fonts that are freely available from 
> _http://dejavu-fonts.org/_ (http://dejavu-fonts.org/) 
>  
> They have several unicode fonts that incorporate almost all of the
unicode
> sets. 
> ...
>> From: Ron Catterall  <ron@catterall.net>
>> ...
>> "the Arial Unicode MS font can be downloaded for  free."
> ...
> This font was useful for fop, to  produce PDFs of documents containing,
> for example, Chinese or Japanese  characters.

AFAIK, the DejaVu fonts don't include any CJK characters.  In fact, there
are few if any fonts that cover the entire Unicode range, or even most of
it.  (The only one I know of simply puts up a box containing the code point
of the character--not any glyphs for the character.)  And once you get
beyond Latin or Cyrillic characters, many of the character sets place
extraordinary demands on the rendering system.  Arabic scripts are
connected, and the Nasta'liq versions of Arabic scripts even more so;
glyphs flip over preceding glyphs in many Indic languages, or show up on
both the left and the right of a preceding glyph; and so on.  It's hard to
find a good font for any one such script, much less a font that covers all
or most of them.

That said, you can look here:
   http://en.wikipedia.org/wiki/Unicode_typefaces

I think the right solution is to use multiple fonts for a document that
contains multiple scripts.  We have that problem with multilingual
documents, and it's reasonably (not completely) straightforward to tag
sequences of characters in this or that Unicode block for the font that
they should use.  (The tags will be dependent on your typesetting system,
of course.)  

   Mike Maxwell