# How to correctly and consistently typeset proofs with implications

I want to typeset proofs that show equivalence of several (numbered) statements. Is there any consistent and nice way to typeset proofs like that, when I want to label the "subproofs" of one statement implying the other with "(x) -> (y)"? I included a minimal working example of what I want it to look like, except for the labels being written on the margin.

\documentclass{scrartcl}
\usepackage{enumitem}
\setlist{
listparindent=\parindent,
parsep=0pt,
}
\usepackage[thmmarks, hyperref, amsthm]{ntheorem}
\usepackage{mathtools, dsfont}
\usepackage{xcolor}

\begin{document}

\noindent\fcolorbox{black}{white}{\parbox{\linewidth -2\fboxsep -2\fboxrule}{%
\begin{proof} This should become the proof of Morera's theorem.
\begin{enumerate}
\item[(i) $\Rightarrow$ (ii):] This implication is trivial.
\item[(ii) $\Rightarrow$ (iii):]
\begin{align*}
F \in \mathcal{O}( D(z,r) ) : F' = f|_{D(z,r)}  &\overset{2.2.4}{\Longrightarrow} f|_{D(z,r)} \in \mathcal{O}(D(z,r)) \\                                                                                    &\Longrightarrow f \in \mathcal{O}(\Omega).
\end{align*}
\item[(iii) $\Rightarrow$ (iv):] Goursat's theorem (Theorem 2.17).
\item[(iv) $\Rightarrow$ (i):] As in theorem 2.19 we see, that on $D(z,r)$
$F : D(z,r) \longrightarrow \mathds{C} \quad , \quad w \mapsto \int_{\gamma_{z \to w}} f(\zeta) \mathop{d\zeta}$
is a holomorphic antiderivative to $f$ on $D(z,r)$.
\end{enumerate}
\end{proof}
}}

\end{document}


• How about just adding leftmargin=63pt, to your \setlist? – Steven B. Segletes May 22 '17 at 12:14
• If there was another way, I guess I would prefer that. 63pt leftmargin for all enumerations in the document would be too much. I ment consistent as in proofs, not necessarily throughout the entire document. But thanks for the effort, really! – Student May 22 '17 at 12:23
• Alternately, ditch the leftindent specifier and add itemindent=36pt, to the \setlist. In both alternatives, the specifier may be added as an optional argument to the enumerate list rather than in the \setlist, if each proof/list must be different. – Steven B. Segletes May 22 '17 at 12:24
• I find that no list is necessary: just start the paragraph with (i)$\implies$(ii) and let it go. – egreg May 22 '17 at 12:28

Considering the comment of the OP, I think the best alternative is to employ the itemindent=... optional argument to each particular enumerate list requiring a tweak. That way, the global indent of lists is not affected (if one used the \setlist, instead), and the indent of subsequent lines in each \item is also not affected (if one used the leftmargin= approach, instead).

Of course, this is not as simple as egreg's suggestion of ditching the \item specifier altogether; however, that approach provides no leftward offset or right-alignment for the implications.

\documentclass{scrartcl}
\usepackage{enumitem}
\setlist{
listparindent=\parindent,
parsep=0pt,
}
\usepackage[thmmarks, hyperref, amsthm]{ntheorem}
\usepackage{mathtools, dsfont}
\usepackage{xcolor}

\begin{document}

\noindent\fcolorbox{black}{white}{\parbox{\linewidth -2\fboxsep -2\fboxrule}{%
\begin{proof} This should become the proof of Morera's theorem.
\begin{enumerate}[itemindent=36pt]
\item[(i) $\Rightarrow$ (ii):] This implication is trivial.
\item[(ii) $\Rightarrow$ (iii):]
\begin{align*}
F \in \mathcal{O}( D(z,r) ) : F' = f|_{D(z,r)}  &\overset{2.2.4}{\Longrightarrow} f|_{D(z,r)} \in \mathcal{O}(D(z,r)) \\                                                                                    &\Longrightarrow f \in \mathcal{O}(\Omega).
\end{align*}
\item[(iii) $\Rightarrow$ (iv):] Goursat's theorem (Theorem 2.17).
\item[(iv) $\Rightarrow$ (i):] As in theorem 2.19 we see, that on $D(z,r)$

$F : D(z,r) \longrightarrow \mathds{C} \quad , \quad w \mapsto \int_{\gamma_{z \to w}} f(\zeta) \mathop{d\zeta}$

is a holomorphic antiderivative to $f$ on $D(z,r)$.
\end{enumerate}
\end{proof}
}}

\end{document}