Put each "Case ..." on its own line in the proof-by-cases environment.

diff --git a/mjo-proof_by_cases.tex b/mjo-proof_by_cases.tex
index b85daefa54a80b9e7365feb42cbcc82a30c0b460..89a813b6d72c7d875a3f180a31ede6d419f6df78 100644 (file)
--- a/mjo-proof_by_cases.tex
+++ b/mjo-proof_by_cases.tex
@@ -13,7 +13,11 @@
 \usepackage{enumitem}
 
 % Using the enumitem package, we define a new type of list, called
-% ``pcases'' (proof by cases).
+% ``pcases'' (proof by cases). Each case has a label with an arabic
+% numeral (the case number), but also a \thiscase identifier. The
+% macro \thiscase is defined below by the \case command, and gives the
+% name or conditions or whatever that distinguish one case from
+% another.
 \newlist{pcases}{enumerate}{1}
 \setlist[pcases]{
   label=\textbf{Case~\arabic*}~(\protect\thiscase)\textbf{:},
@@ -21,7 +25,12 @@
   align=left,
   leftmargin=\parindent}
 
+% The optional argument here gets stuffed into the \thiscase macro, to
+% be called by pcases when it creates this list item. The \hfill is a
+% hack intended to force the proof to start on a new line, rather than
+% right after the colon. A \newline where the \hfill is does not work,
+% so we consume the rest of the line instead.
 \newcommand{\case}[1][]{
   \def\thiscase{#1}%
-  \item
+  \item \hfill
 }