[XeTeX] accented character ṛ within \section{ṛ}

[Philipp Stephani]

Am 26.04.2010 um 21:23 schrieb Herbert Schulz:

> 
> On Apr 26, 2010, at 11:44 AM, Khaled Hosny wrote:
> 
>> On Mon, Apr 26, 2010 at 12:11:52PM -0400, Peter Baker wrote:
>>> This is because Junicode does not have precisely the same character
>>> set in all faces.
>> 
>> That is one reason I personally think combining marks + ccmp are better
>> than pre-composed glyphs, the less glyph work the easier to sync fonts.
>> Sure it means a bit extra OpenType work, but this can easily be shared
>> between faces.
>> 
>> Regards,
>> Khaled
>> 
> 
> Howdy,
> 
> Just curious... what happens when you try to do search within or a copy from a pdf which has such combined characters?

Precomposed characters are defined to be canonically equivalent to the corresponding composite character sequences. Although it is not a requirement, software programs should treat them identically, especially when they are built to work with text. Furthermore there should be an option to ignore accents when searching as they cannot be entered easily by all users. I think at least Adobe Reader implements both.