How to sort glossaries entries by alphabetical order then by order of appereance

Question

I would like my glossaries entries to be sorted by alphabetical order, and, if some entries have identical sort keys, by the order they are called in the text. Any way to do this ?

Here is the MWE, with the two entries div and commut exhibiting identical sort keys.

\documentclass[]{article}

\usepackage[
            nogroupskip,
%           xindy, 
            ]{glossaries}
% I compile with xindy, but I do not think it is relevant here.
% xindy -L english -C utf8 -I xindy -M % -t %.glg -o %.gls %.glo


\newglossaryentry{div}
{ name={\ensuremath{\protect\vec{\nabla}.}},
  description={Divergence operator},
  sort={0}
  }

\newglossaryentry{commut}
{ name={\ensuremath{\protect[~,~]}},
  description={Commutator},
  sort={0}
  }

\newglossaryentry{b0}
{ name={\ensuremath{\protect\vec{B}_0}},
  description={Static magnetic field},
  sort={B0}
  }

\makeglossaries

\begin{document}
Text.\gls{div} \gls{commut} \gls{b0}

\printglossaries
\end{document}

Here, for example, the commutator do not appear in the glossary. I would like to have it appearing after the divergence operator despite their identical sorting keys.

Answer 1

I would suggest using \makenoidxglossaries together with \printnoidxglossary[sort=standard] which sorts the elements first according to the sort-key and then according to usage.

\documentclass{article}
\usepackage[nogroupskip]{glossaries}

\newglossaryentry{div}
{ name={\ensuremath{\protect\vec{\nabla}.}},
  description={Divergence operator},
  sort={0}
  }

\newglossaryentry{commut}
{ name={\ensuremath{\protect[~,~]}},
  description={Commutator},
  sort={0}
  }

\newglossaryentry{b0}
{ name={\ensuremath{\protect\vec{B}_0}},
  description={Static magnetic field},
  sort={B0}
  }

\newglossaryentry{b1}
{ name={\ensuremath{\protect\vec{B}_1}},
  description={Static magnetic field},
  sort={B0}
  }

\makenoidxglossaries

\begin{document}
Text.\gls{b1} \gls{div} \gls{commut} \gls{b0}

\printnoidxglossary[sort=standard]
\end{document}

Then you get

First the 0 entries show up according to usage, then the two B0 entries.

(xindy might change the entries for the sorting, so I'm not sure if it works there...)