On Saturday 25 June 2016 15:11:09 Akira Kakuto wrote:

...

I found another problem in glyphtounicode.tex file. There are duplicate codes for ligatures, see: ...

I cannot find such duplicate lines in glyphtounicode.tex in TeX Live: % lcdf-typetools glyphtounicode.tex, Version 2.95

Ah right, again I check my TeXLive (in 2012) and svn code in /trunk/... Sorry I forgot to look into /branches/stable/ which is *real* pdftex trunk. Anyway, question about not using unicode characters for ligatures remains. -- Pali Rohár pali.rohar@gmail.com

Hi Karl, Pali and others On Jun 26, 2016, at 8:34 AM, Karl Berry mailto:karl@freefriends.org> wrote: 4) For the ligatures, I can't remember for sure, but I suspect that the entries predate the existence of the ligatures in Unicode. Yes. I think this is most likely true. It is not clear to me that the change is desirable, because the "decomposition" into two f characters is what is actually desired for searches, etc. If you can convince Eddie, fine; it's up to him. A practical usage case is probably needed for that. The current entries have been around a long, long, time and should not be changed merely to "follow" Unicode. Agreed. I was looking at this kind of thing more than 10 years ago, and very few fonts supported the ligatures. So copy/paste would result in missing characters. 5) Regarding your previous message about Delta(greek). glyphtounicode.tex generally follows Adobe's original glyphlist.txt http://partners.adobe.com/public/developer/en/opentype/glyphlist.txt which has the same Delta + Deltagreek definitions. Which are more useful in practice. I can't imagine there will or should be any changes there. This is indeed more useful with mathematical content. However, if the document is about Greek language, or written substantially in Greek, then the proper language code is what would be required. A customised CMAP file is then a good way to go — see my previous response. 5b) I note that glyphlist.txt does have the FB0* for the f-ligatures. I don't know if we intentionally diverged from that for TeX. 6) If you want to make changes for your own purposes, you can just make your own version of the file. Or create a package that loads a customized CMAP with your particular fonts. You do *not* want to use \pdfglyphtounicode with glyph names that clash with those that are otherwise loaded automatically, else you could be killing the correct mapping for other fonts. Hope this helps, Karl Cheers Ross Dr Ross Moore Mathematics Dept | Level 2, S2.638 AHH Macquarie University, NSW 2109, Australia T: +61 2 9850 8955 | F: +61 2 9850 8114tel:%2B61%202%209850%209695 M:+61 407 288 255tel:%2B61%20409%20125%20670 | E: ross.moore@mq.edu.aumailto:rick.minter@mq.edu.au http://www.maths.mq.edu.auhttp://mq.edu.au/ [cid:image001.png@01D030BE.D37A46F0]http://mq.edu.au/ CRICOS Provider Number 00002J. Think before you print. Please consider the environment before printing this email.http://mq.edu.au/ This message is intended for the addressee named and may contain confidential information. If you are not the intended recipient, please delete it and notify the sender. Views expressed in this message are those of the individual sender, and are not necessarily the views of Macquarie University.http://mq.edu.au/

On Sunday 26 June 2016 00:34:51 Karl Berry wrote:

3) glyphtounicode.tex is no longer maintained in pdftex (or in TeX Live).

Then would be better to remove it from pdftex tree, so other people do not look at incorrect place (as me).

4) For the ligatures, I can't remember for sure, but I suspect that the entries predate the existence of the ligatures in Unicode. It is not clear to me that the change is desirable, because the "decomposition" into two f characters is what is actually desired for searches, etc. If you can convince Eddie, fine; it's up to him. A practical usage case is probably needed for that. The current entries have been around a long, long, time and should not be changed merely to "follow" Unicode.

Ok, so problem is that pdf viewers do not support searching for unicode ligatures and decomposition help users. Understand.

5) Regarding your previous message about Delta(greek). glyphtounicode.tex generally follows Adobe's original glyphlist.txt

http://partners.adobe.com/public/developer/en/opentype/glyphlist.txt which has the same Delta + Deltagreek definitions. Which are more useful in practice. I can't imagine there will or should be any changes there.

More mathematical fonts have under glyph Delta greek delta character. And in mathematical text is greek Delta used as Delta, not increase character. This reason for my proposal of update. -- Pali Rohár pali.rohar@gmail.com