[MLton] Long identifiers and def-use data

Stephen Weeks MLton@mlton.org
Tue, 18 Apr 2006 11:12:35 -0700

One advantage to reporting more precise def-use information as Vesa
suggests is that it disambiguates the case in which the same
identifier appears multiple times in a long identifier (e.g. S.S.x).
With the more precise information, the emacs mode can have a flag that
controls whether the entire long identifier is highlighted or just the
component is.  With the current information, no such flag is possible,
because the elisp code can't always determine which component of the
long identifier is being used.