[NTG-pdftex] problems still, within tounicode.c

Frank Mittelbach frank.mittelbach at latex-project.org
Tue Jun 13 10:13:06 CEST 2017

Am 13.06.17 um 09:46 schrieb Ross Moore:
>> What is the common expectation when one does copy/paste this text from 
>> a pdf viewer?
>> I personally expect to get the same text, not the unicode points for 
>> the *.sc glyphs from the actual font.
> Perhaps.
> But then the author is stating that it is a heading, but not using 
> proper markup that
> might be used to include tagging that could convey this.

or the author has written \textsc{ctan} or \textsc{ibm} in which case 
one would perhaps expect CTAN and IBM not ctan and ibm in the 
cut-and-paste. So I think it is more or less impossible to give a 
general rule that works universally.

However, for cut and paste I guess that you usually wouldn't want to 
receive anything other than the base text not the *.sc glyph code points 
in most of the cases similar to wanting plain text when selecting, say, 
parts of a bold heading.

So in my opinion the default should be plain base text

As to deducing a heading structure if it is not properly marked up, that 
seems to me a totally different kind of usage and yes, there knowing 
that something is in a special font (and on a line by itself etc) might 
be a useful indicator, but it screws up the other use, so something 
would need to give


More information about the ntg-pdftex mailing list