Determine hex unicode character code from raw character?

Question

I found coloremoji which allows inserting apple's color emoji's into an uptex document. This library first extracts all emoji's as PDFs named according to their hex unicode character codes (e.g. 1F36E.pdf). Then there is a stylesheet (coloremoji.sty) which seemingly looks up an incoming emoji's hex code and includes the corresponding PDF.

However, this seems to require some upTeX specific features. I'd like to get this working with pdflatex using \usepackage{stringenc},\usepackage{utf8},\usepackage{fontenc}, etc.

Is this possible to look up a raw unicode character's hex code using pdflatex (TeXLive on Mac OS X)?

Here's the imagined interface (as a screenshot to preserve emoji)

The coloremoji command should expand to an includegraphics so that the resulting document is a cheeseburger.

Answer 1

If you have already the converted PDF files, here's a version that works with all engines:

\documentclass{article}
\usepackage{graphicx,stringenc}
\usepackage{xparse,l3regex}

\ExplSyntaxOn
\tl_const:Nn \c_coloremoji_dir_tl { ./emoji_images/hires }
\tl_new:N \l_coloremoji_input_tl

\cs_new_protected:Npn \coloremoji_unicode:nn #1 #2
 {
  \includegraphics[#1]{\c_coloremoji_dir_tl / \int_to_Hex:n{`#2}.pdf}
 }

\cs_new_protected:Npn \coloremoji_eightbit:nn #1 #2
 {
  \StringEncodingConvert \l_coloremoji_input_tl
    { \tl_to_str:n { #2 } } { utf8 } { utf32 }
  \tl_set:Nx \l_coloremoji_input_tl { \pdfescapehex\expandafter{\l_coloremoji_input_tl} }
  \regex_replace_once:nnN { \A 0* } { } \l_coloremoji_input_tl
  \includegraphics[#1]{\c_coloremoji_dir_tl / \l_coloremoji_input_tl.pdf}
 }

\NewDocumentCommand{\coloremoji}{O{}m}
 {
  \bool_if:nTF { \xetex_if_engine_p: || \luatex_if_engine_p: }
   {
    \coloremoji_unicode:nn { #1 } { #2 }
   }
  {
   \coloremoji_eightbit:nn { #1 } { #2 }
  }
 }

\ExplSyntaxOff

\begin{document}

\coloremoji{}

\end{document}

Compile with pdflatex, XeLaTeX or LuaLaTeX. Change the meaning of \c_coloremoji_dir_tl to point to the directory where the PDF files are.

---

This is a pdftex only solution:

\documentclass{article}
\usepackage{stringenc,graphicx}

\makeatletter
\newcommand{\gethex}[1]{%
  \StringEncodingConvert{\gh@tmp}{\detokenize{#1}}{utf8}{utf32}%
  \edef\gh@tmp{\pdfescapehex\expandafter{\gh@tmp}}%
  \expandafter\get@hex\gh@tmp\get@hex
}
\def\get@hex#1{%
  \if0#1%
    \expandafter\@firstoftwo
  \else
    \expandafter\@secondoftwo
  \fi
  {\get@hex}{\get@hex@aux#1}%
}
\def\get@hex@aux#1\get@hex{%
   \def\gh@string{#1}%
}
\newcommand{\coloremoji}[2][]{%
  \gethex{#2}%
  \includegraphics[#1]{./emoji_images/hires/\gh@string.pdf}%
}
\makeatother

\begin{document}

\coloremoji{}

\end{document}

---

A version without stringenc that works with all engines. Note that the argument to \coloremoji must be a single Unicode character. The \newunicodechar part is just to show compatibility. The optional argument to \coloremoji is passed to \includegraphics.

\documentclass{article}

%\usepackage[utf8]{inputenc} % not necessary, but recommended if pdflatex is used
\usepackage{graphicx}
\usepackage{xparse}
\usepackage{newunicodechar}

\newunicodechar{}{\coloremoji{}}

\ExplSyntaxOn

\NewDocumentCommand{\coloremoji}{O{}m}
 {
  \coloremoji_unicode:nn { #1 } { #2 }
 }

%%% the location where the images are found
\tl_const:Nn \c_coloremoji_dir_tl { ./emoji_images/hires }

\cs_new_protected:Npn \coloremoji_unicode:nn #1 #2
 {
  \includegraphics[#1]{\c_coloremoji_dir_tl / \manual_get_hex:n { #2 }.pdf}
 }

\cs_new:Npn \manual_get_hex_unicode:n #1
 {
  \int_to_Hex:n { `#1 }
 }

\cs_new:Npn \manual_utfviii_to_unicode:n #1
 {
  \tl_if_blank:oTF { \tl_to_str:n { #1 } }
   { 20 }
   { \__manual_utfviii_to_unicode_aux:o { \tl_to_str:n { #1 } } }
 }

\cs_new:Npn \__manual_utfviii_to_unicode_aux:n #1
 {
  \int_case:nn { \tl_count:n { #1 } }
   {
    {1}{ \int_to_Hex:n { `#1 } }
    {2}{ \__manual_utfviii_twobyte:f { #1 } }
    {3}{ \__manual_utfviii_threebyte:f { #1 } }
    {4}{ \__manual_utfviii_fourbyte:f { #1 } }
   }
 }
\cs_generate_variant:Nn \__manual_utfviii_to_unicode_aux:n { o }

\cs_new:Npn \__manual_utfviii_twobyte:n #1
 {
  \__manual_utfviii_twobyte_aux:nn #1
 }
\cs_generate_variant:Nn \__manual_utfviii_twobyte:n { f }

\cs_new:Npn \__manual_utfviii_twobyte_aux:nn #1 #2
 {
  \int_to_Hex:n { (`#1-192)*64 + (`#2-128) }
 }

\cs_new:Npn \__manual_utfviii_threebyte:n #1
 {
  \__manual_utfviii_threebyte_aux:nnn #1
 }
\cs_generate_variant:Nn \__manual_utfviii_threebyte:n { f }

\cs_new:Npn \__manual_utfviii_threebyte_aux:nnn #1 #2 #3
 {
  \int_to_Hex:n { (`#1-224)*4096 + (`#2-128)*64 + (`#3-128) }
 }

\cs_new:Npn \__manual_utfviii_fourbyte:n #1
 {
  \__manual_utfviii_fourbyte_aux:nnnn #1
 }
\cs_generate_variant:Nn \__manual_utfviii_fourbyte:n { f }

\cs_new:Npn \__manual_utfviii_fourbyte_aux:nnnn #1 #2 #3 #4
 {
  \int_to_Hex:n { (`#1-240)*262144 + (`#2-128)*4096 + (`#3-128)*64 + (`#4-128) }
 }

\bool_if:nTF { \xetex_if_engine_p: || \luatex_if_engine_p: }
 {
  \cs_set_eq:NN \manual_get_hex:n \manual_get_hex_unicode:n
 }
 {
  \cs_set_eq:NN \manual_get_hex:n \manual_utfviii_to_unicode:n
 }

\ExplSyntaxOff

\begin{document}
\coloremoji{}

\coloremoji[width=1cm]{}


\end{document}