Expand \FV@EscChars to handle comma, possibly other characters

[gpoore]

\edef\FV@commachar{\string,}
\newcommand{\FV@EscChars}{%
  \let\#\FV@hashchar
  \let\%\@percentchar
  \let\{\@charlb
  \let\}\@charrb
  \let\$\FV@dollarchar
  \let\&\FV@ampchar
  \let\_\FV@underscorechar
  \let\\\@backslashchar
  \let~\FV@tildechar
  \let\~\FV@tildechar
  \let\[\FV@leftsquarebracket
  \let\]\FV@rightsquarebracket
  \let\,\FV@commachar
}

[muzimuzhi]

Perhaps even a hook so users can add their own?

\appto\FV@EscChars{\FV@EscChars@user}

% definition
\define@key{FV}{escape chars}{\def\FV@EscChars@user{#1}}
\define@key{FV}{escape chars*}{\appto\FV@EscChars@user{#1}}
\fvset{escape chars}

Just for fun, here's an implementation of a more user-friendly option new escape char=<char> | \<char> in latex3.

\documentclass{article}
\usepackage{fvextra}

\begin{document}
\makeatletter
% new escape char = <char> | \<char>
\define@key{FV}{new escape char}{%
  \FV@EscChars@add{#1}%
}
\def\FV@EscChars@user{}

\appto\FV@EscChars{\FV@EscChars@user}


\ExplSyntaxOn
\cs_new_protected:Npn \FV@EscChars@add #1
  {
    \tl_if_single_token:nTF {#1}
      {
        \token_if_cs:NTF #1 
          { \exp_args:Ne \FV@EscChars@@add { \cs_to_str:N #1 } }
          { \exp_args:Ne \FV@EscChars@@@add { \tl_to_str:n {#1} } }
      }
      % #1 contains multiple tokens
      { \msg_error:nnx {fvextra} {invalid-escape-char} { \tl_to_str:n {#1} }}
  }

\cs_new_protected:Npn \FV@EscChars@@add #1
  {
    \exp_args:Ne \tl_if_single_token:nTF { #1 }
      { \FV@EscChars@@@add {#1} }
      % #1 of \FV@EscChars@add is a control sequence but has its csname longer than one
      { \msg_error:nnn {fvextra} {invalid-escape-char} { #1 } }
  }

\cs_new_protected:Npn \FV@EscChars@@@add #1
  {
    % \def \FV@char@<char> { <char> }
    \tl_set:cn { FV@char@#1 } { #1 }
    % append \tl_set_eq:NN \<char> \FV@char@<char> to \FV@EscChars@user
    \exp_args:NNe \tl_put_right:Nn \FV@EscChars@user 
      { \tl_set_eq:NN \exp_not:c {#1} \exp_not:c {FV@char@#1} }
  }

\msg_new:nnn {fvextra} {invalid-escape-char}
  {
    Invalid~ value~ "#1"~ for~ "new~ escape~ char", \\
    should~ be~ either~ a~ single~ char~ or~ length~ 1~ control~ sequence
  }
\ExplSyntaxOff

\fvset{
  new escape char={,}, new escape char=^,
  new escape char=\%, new escape char=\ , new escape char=\#, 
  % should error
  new escape char={:;}, new escape char={\ab}
}

%\ttfamily
%\meaning\FV@EscChars@user
\show\FV@EscChars@user
\makeatother
\end{document}

[muzimuzhi]

Resolving of this issue made it necessary to use escaped comma (\,) in breakbefore and breakafter. It that an expected change?

[gpoore]

@muzimuzhi Using a comma seems to work as expected:

\documentclass{article}

\usepackage{fvextra}

\begin{document}

TEXT

\begin{Verbatim}[breaklines,breakafter={,}]
a,a,a,a,a,a,a,a,a,a,a,a,a,a,a,a,a,a,a,a,a,a,a,a,a,a,a,a,a,a,a,a,a,a,a,a,a,a,a,a,a,a,a,a,a,a,a,a,a,a,a,a,a,a,a,a,a,a,a,a,a,a,a,a,a,a,a,a,a,a,
\end{Verbatim}

\end{document}

Or is this related to minted, before you created gpoore/minted#407?

[muzimuzhi]

Or is this related to minted, before you created gpoore/minted#407?

Ah yes. I made a mistake by testing with breakafter=,, but not breakafter={,}.

Then maybe words "should be backslash-escaped" in the related doc are too strong:

Special characters given to breakafter should be backslash-escaped (usually #,
{, }, %, [, ], and the comma ,; the backslash \ may be obtained via \\ and the
space via \space).