[metapost] mptopdf loses characters

[Hartmut Henkel]

On Sat, 24 Feb 2007, Stephan Hennig wrote:

> the following file prints two letters 'e' at the same position.
> Looking into MetaPost's eps output both characters seem to be
> available at that stage.
>
> However, after converting the file to pdf by mptopdf the information
> that there were two characters seems to be lost.  Neither can I copy
> both characters from AR7 nor is pdftotext able to extract two
> characters, but only one.  I would regard this a bug.

both characters are still in the PDF file, on top of each other, as you
can verify by pdftk uncompress. So mptopdf does the right thing. What
then happens, is viewer dependent.

> On the other hand, epstopdf fails converting the file without
> information loss, too.  Is this behaviour in compliance with ps/pdf
> specs?

no idea.

> prologues := 1;
> beginfig(1);
>   label("e", origin);
>   label("e", origin);
> endfig;
> end

Regards, Hartmut