Parsing a Flat SGF String (with the Package listofitems)

Question

I'm trying to parse an SGF string with the package listofitems. SGFs are trees in text form, but right now, I'm working on only single-branched SGFs, for simplicity.

Here's an example of what I'm trying to do:

\documentclass{article}
\usepackage{listofitems}
\usepackage{tikz}
\newcommand{\parseSgf}[1]{
  % From this answer by @StevenB.Segletes.
  \setsepchar{;}
  \defpair{[]}
\readlist\Z{#1}
\begin{itemize}
    \foreach \i in {\Z}{
      % TODO: if key is either B or W:
      \item Color: {\i}[0] and Coords: {\i}[1]
    }
  \end{itemize}
}
\begin{document}
  \def\sgfA{;B[ab];W[cd]}
  \def\sgfB{(;GM[1]FF[4]CA[UTF-8]AP[Sabaki:0.52.2]KM[6.5]SZ[19]DT[2024-02-05];B[as];W[bs];B[cs])}
\parseSgf{\sgfA}
  \parseSgf{\sgfB}
\end{document}

sgfA isn't proper SGF, just a simple example. I would like it to print an unordered list like this:

• Color: B and Coords: ab
• Color: W and Coords: cd

But I haven't been able to figure out how to use \setsepchar and \defpair to separate things properly, or how to use them in a \foreach.

sgfB is a proper SGF string. And there's the additional problem of ignoring keys that are not B or W. If anyone knows how to do it with listofitems, that would be a plus.

Answer 1

I undestand that you are primarily interested in the parts of the string which have the syntax B[xy] or W[xy] (where x and y are some single letters). The below code therefore ignores everything else.

This should probably not be considered the final version, but maybe something like this can do. I used expl sequences in this case as most of the things are related to splitting strings (note that a "missing item" error may pop up, if the string does not contain any suitable syntax):

\documentclass{article}
\ExplSyntaxOn
\cs_generate_variant:Nn \tl_set:Nn { Ne } % You might need to add this if the version of your TeX installation is not the most recent one 
\cs_generate_variant:Nn \seq_set_split:Nnn { Nne }
\NewDocumentCommand{\parseSgf}{ m }{
  \tl_set:Ne \l_tmpa_tl { #1 }
  % remove ( and ) from string
  \tl_remove_all:Nn \l_tmpa_tl { ( }
  \tl_remove_all:Nn \l_tmpa_tl { ) }
  % store the contents of #1 in a sequence with ; as delimiter
  \seq_set_split:Nne \l_tmpa_seq { ; } { \l_tmpa_tl }
  \begin{itemize}
    % loop over all items in sequence
    \seq_map_inline:Nn \l_tmpa_seq {
      % split string at [ and store result in another sequence
      \seq_set_split:Nne \l_tmpb_seq { [ } { ##1 }
      \bool_if:nT { 
        % test if first part of sequence is B or W
        \str_if_eq_p:ee { \seq_item:Nn \l_tmpb_seq { 1 } } { B } ||
        \str_if_eq_p:ee { \seq_item:Nn \l_tmpb_seq { 1 } } { W } 
      } {
        \tl_set:Ne \l_tmpa_tl { \seq_item:Nn \l_tmpb_seq { 1 } }
        \tl_set:Ne \l_tmpb_tl { \seq_item:Nn \l_tmpb_seq { 2 } }
        % remove ] from string
        \tl_remove_all:Nn \l_tmpb_tl { ] }
        \item Color: ~ \l_tmpa_tl {} ~ and ~ Coords: ~ \l_tmpb_tl
      } 
   }
  \end{itemize}
}
\ExplSyntaxOff
\begin{document}
\def\sgfA{;B[ab];W[cd]}
  \def\sgfB{(;GM[1]FF[4]CA[UTF-8]AP[Sabaki:0.52.2]KM[6.5]SZ[19]DT[2024-02-05];B[as];W[bs];B[cs])}
\parseSgf{\sgfA}
  \parseSgf{\sgfB}
\end{document}

Answer 2

Here is a listofitems approach. However, it will not test for improper syntax in the SGF string, such as ;B[ab[;W]cd], which will respond in the same way as the test case \sgfA.

\documentclass{article}
\usepackage{listofitems,tikz}
\long\def\Firstof#1#2\endFirstof{#1}
\ignoreemptyitems
\newcommand{\parseSgf}[1]{%
  \setsepchar{;/[||]}%
  \readlist*\Z{#1}%
  \begin{itemize}
    \foreachitem \i \in \Z[]{%
      \itemtomacro\Z[\icnt,1]\tmp
      % TODO: if key is either `B` or `W`:
      \expandafter\if\expandafter\Firstof\tmp\endFirstof B
        \item Color: \tmp and Coords: \Z[\icnt,2]
      \else\expandafter\if\expandafter\Firstof\tmp\endFirstof W
        \item Color: \tmp and Coords: \Z[\icnt,2]
      \fi\fi
    }%
  \end{itemize}
}
\begin{document}
  \def\sgfA{;B[ab];W[cd]}
  \def\sgfB{(;GM[1]FF[4]CA[UTF-8]AP[Sabaki:0.52.2]KM[6.5]SZ[19]DT[2024-02-05];B[as];W[bs];B[cs])}
  \parseSgf{\sgfA}\bigskip
  \parseSgf{\sgfB}
\end{document}