[XeTeX] Another tipa/xunicode puzzle: interaction with array.sty?

Ross Moore ross.moore at mq.edu.au
Tue Aug 31 00:26:31 CEST 2010

Hi Alan,

On 31/08/2010, at 5:36 AM, Alan Munn wrote:

> On Aug 30, 2010, at 3:33 PM, Ross Moore wrote:

> Why does the same code work with the IPA environment using tipa.sty and regular latex?

There you are using a font that has the TIPA characters in regular
ASCII positions. The characters are not made active there, 
whereas the \setTIPAcatcodes makes them effectively macros,
expanding to the IPA characters.

It could be done differently, by writing a parser that
looks at each letter in sequence, and choosing whether
to map it to something else or leave it as it is.
But that strategy would only be reliable if you could 
guarantee that only IPA characters were appearing within 
your environment. 

It would be really tricky to get this working if you allowed 
other macros to be in there;
e.g. \textcolor{...}{.. IPA characters ... }. 
You would have to program the parser to be able to recognise
every single construction that you might want to use, since
it needs to know when to turn itself off, and back on again.

The active character mechanism is both easier and more
efficient, so far as the TeX processing is concerned.
If it means editing your legacy source to put in \relax
in a lot of places, then that's the price you pay for getting
proper Unicode output, rather than staying with the old, 
hacky  tipa.sty  methods.

Using a CMap with the tipa fonts is another possible approach.

> Alan
> -- 
> Alan Munn
> amunn at gmx.com

Hope this helps,


Ross Moore                                       ross.moore at mq.edu.au 
Mathematics Department                           office: E7A-419      
Macquarie University                             tel: +61 (0)2 9850 8955
Sydney, Australia  2109                          fax: +61 (0)2 9850 8114

More information about the XeTeX mailing list