Abbreviate resolved abbreviations in authors' first names in biblatex

Question

I like to resolve abbreviated first names in my .bib file. For example, if the author of a book is given as Paul G. J. Smith, I will resolve the abbreviations as Paul G[eorge] J[ohn] Smith in my .bib file.

Some journals, however, specifically want the names of the authors to be printed as they appear in the cited work. The entry for Smith, therefore, should appear as Smith, Paul G. J. in the printed bibliography.

What do I need to do in biblatex in order to tell it to replace a sequence beginning with [ and ending with ] with a dot .? Please suggest a command I can put in the preamble of my .tex file. I have a long list of biblatex tweaks in my preambles, and I would prefer to keep them all there, if possible.

A MWE:

\documentclass{article}
\usepackage{csquotes}
\usepackage[
        bibstyle = authoryear,
        citestyle = authoryear-comp,
        sorting = nyt,
        language = american,
        abbreviate = false,
        backend = biber]{biblatex}
\usepackage{filecontents}
\begin{filecontents}{\jobname.bib}
@BOOK{smith2006,
    AUTHOR = "Paul G[eorge] J[ohn] Smith",
    TITLE = "My life with the Beatles",
    YEAR = "2006"}
\end{filecontents}
\addbibresource{\jobname.bib}
\begin{document}
\noindent
Text \parencite{smith2006}.
\printbibliography
\end{document}

Answer 1

The built-in way to do such things is with the biblatex \DeclareSourcemap macro. To use this, you have to be using biber. This method is preferable as it has semantic access to biblatex native objects. Try adding this to the preamble of your example above:

\DeclareSourcemap{
  \maps[datatype=bibtex]{
    \map{
      \step[fieldsource=author, 
            match=\regexp{\[[^]]+\]},
            replace=.]
    }
  }
}

There are many other things you can do with this, see section 4.5.2 of the current biblatex manual. This is completely style independent, it alters the data stream from the .bib before biblatex even sees it.

Answer 2

I'm not sure if this is the right way, but it seems to work (part of the code is from this answer by Marco Daniel), particularly in the usage of \DeclareNameFormat and \DeclareNameAlias. Maybe experts in biblatex can improve it.

The following method uses l3regex to get rid of the bracketed parts:

\documentclass{article}
\usepackage{csquotes}
\usepackage[
        bibstyle = authoryear,
        citestyle = authoryear-comp,
        sorting = nyt,
        language = american,
        abbreviate = false,
        backend = biber]{biblatex}

\DeclareNameAlias{sortname}{last-first} % for styles where names are sorted last-first
\DeclareNameFormat{last-first}{%
  \usebibmacro{name:last-first}{#1}{\removebrackets{#3}}{#5}{#7}%
  \usebibmacro{name:andothers}}

\DeclareNameAlias{sortname}{first-last} % for styles where names are sorted first-last
\DeclareNameFormat{first-last}{%
  \usebibmacro{name:first-last}{#1}{\removebrackets{#3}}{#5}{#7}%
  \usebibmacro{name:andothers}}

\usepackage{xparse,l3regex}
\ExplSyntaxOn
\NewDocumentCommand{\removebrackets}{m}
 {
  \sverre_remove_brackets:n { #1 }
 }
\cs_new_protected:Npn \sverre_remove_brackets:n #1
 {
  \tl_set:Nn \l__sverre_input_tl { #1 }
  \regex_replace_all:nnN { \[.*?\] } { \. } \l__sverre_input_tl
  \tl_use:N \l__sverre_input_tl
 }
\tl_new:N \l__sverre_input_tl
\ExplSyntaxOff


\usepackage{filecontents}
\begin{filecontents}{\jobname.bib}
@BOOK{smith2006,
    AUTHOR = "Paul G[eorge] J[ohn] Smith",
    TITLE = "My life with the Beatles",
    YEAR = "2006"}
@BOOK{smith2006x,
    AUTHOR = "Paul G[eorge] J[ohn] Smith and Knuth, Donald E[rving]",
    TITLE = "My life with the Beatles",
    YEAR = "2006"}
@BOOK{smith2006y,
    AUTHOR = "Paul G[eorge] J[ohn] Smith and X, Y and A, B and C, D and E, F and G, H",
    TITLE = "My life with the Beatles",
    YEAR = "2006"}
\end{filecontents}
\addbibresource{\jobname.bib}
\begin{document}
\noindent
Text \parencite{smith2006}.

\parencite{smith2006x}
\parencite{smith2006y}
\printbibliography
\end{document}