Environment definitions with arguments in \enddef

Question

I was hoping to create a custom proof environment where along with the qed symbol it also mentions what was the claim that was proved. I thought the following would work:

\newenvironment{myproof}[1]%
{\vspace{1ex}\noindent{\em Pf.}\hspace{0.5em}}%
{\hfill{\tiny \qed (#1)}\vspace{1ex}}

I thought \begin{myproof}{Claim 42} ..\end{myproof} would work but it doesn't. After some googling I figured that it was because the argument is not passed on to \endmyproof.

Is there any fix to this? Or is this just bad TeX?

PS: FWIW, I was writing down a proof and had a small claim inside. The QED on the claim was a bit misleading that the proof of the Theorem was done. So this seemed like a natural(?) thing to do.

Answer 1

The \newenvironment{myproof} command behind the scenes actually defines two macros: \myproof and \endmyproof. These will be called by the \begin and \end commands to do the job. By the way of how \begin and \end work, only \begin can pass any arguments to the underlying macro, and thus \newenvironment only defines arguments for \myproof.

You can work around this by defining a temporary macro in \myproof and using its value in \endmyproof:

\newenvironment{myproof}[1]%
{\vspace{1ex}\noindent{\em Pf.}\hspace{0.5em}\def\myproofname{#1}}%
{\hfill{\tiny \qed (\myproofname)}\vspace{1ex}}

(If you put this environment definition in a package file, you could name this temporary macro \myproof@name, to avoid usage in the document and clashes.)