I have this code:

I always use \MyBrand , it's the best!

Which results in the following pdf:

I always use MyBr👌nd, it’s the best!

I want to be able to copy the complete text from the pdf to a text editor (Notpad, Word, …), including the text in the image.

When I mark the sentence in a pdf-Viewer (Adobe Acrobat) and copy it, I get "I always use , it’s the best!". But I want "I always use MyBr👌nd, it’s the best!".

So added an invisible tikz-layer fitted to the image width, in which I can type my text (because I used the resizebox, the text can even be very long):

        \tikz\node[inner sep=0pt, opacity=0]{myBrand};%

With the added tikz node I get "I always use myBrand, it’s the best!". That is a start, but I want "I always use MyBr👌nd, it’s the best!".

If I change the node to \tikz\node[inner sep=0pt, opacity=0]{myBr👌nd};, I get garbage like "I always use myBr￿nd, it’s the best!" when compiling with XeLaTeX and nothing with pdfLaTeX (inputenc Error).

How can I print ANY utf-8 character (even if the font doesn't have a symbol for it) for the purpose of copying it from the pdf into a text editor (which hopefully has an appropriate font)? Is there a solution for both Xe- and pdfLaTeX?

Again: I don't care if the symbol doesn't show up or shows as a placeholder… as long as it can be copied by marking the text in a pdf-Viewer and pasted to a text editor (where I can chose a font supporting all utf-8 characters).


You can use the accsupp package. But it won't work in every pdf-viewer, e.g. sumatra doesn't copy it.



I always use \MyBrand, it's the best! 



as discovered in the comments it doesn't work if the picture is a bitmap (e.g. jpg or png). In this case one work-around is to a glyph, e.g.

  • Thanks for your answer. I tried it with Acrobat Reader XI and it doesn't work (neither compiling with XeLaTeX, nor with pdfLaTeX). It just copy-pasted "I always use , it’s the best!" as if there were only the image. With what pdf-viewer did it work for you? – genericFJS Oct 23 '17 at 9:41
  • Adobe reader dc. The example is for pdflatex, with lualatex you need also luatex85, i didn't try with xelatex. Do you have a current tex system? – Ulrike Fischer Oct 23 '17 at 9:44
  • Yes, I have a full texlive 2017 installation (updated just now). I'm compiling with pdflatex -synctex=1 -interaction=nonstopmode -shell-escape %.tex. I tried opening it with Google Chrome and Microsoft Edge too, but it didn't copy anything from the "ActualText". – genericFJS Oct 23 '17 at 9:53
  • I just installed Adobe Reader DC and I couldn't copy anything "from" the image as well. Just "I always use , it's the best!", as before :-( – genericFJS Oct 23 '17 at 10:13
  • You really should use nonstopmode -- or if you do you should check the log ;-). My code works with a development version of hyperref. But I don't have the time now to check which commit enable the handling of this unicode range. – Ulrike Fischer Oct 23 '17 at 10:32

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.