another automatic beaming question

[Russ Ross]

As I indicated in a previous message, I'm using a script to give me
shortcuts for \melisma and \melismaEnd that also turn on automatic
beaming during the melisma.  This uncovered a bug where \autoBeamOff
doesn't take effect until after the current group of beamed notes
finishes, so something like:

\autoBeamOn a8 b c \autoBeamOff d r2

has all four notes under a beam, including the d.  To work around it,
I had my script insert a \noBeam  after the d.  Unfortunately, this
cancels the beam for the entire beam group, all the way back to the a.

Erik filed a bug report for me and indicated that autobeaming is
slated for an overhaul, so for now I'm mostly looking for a workaround
that I can put in my script.  Is there another way to force an end to
the current beam group?  I tried putting a zero-length space:

\autoBeamOn a8 b c \autoBeamOff s0 d r2

but LilyPond didn't like it.  I get the desired behavior with this:

\autoBeamOn a8 b c \autoBeamOff d[] r2

namely, the first three notes are beamed and the last one isn't,
although there are a bunch of warnings to go with it so I'm a bit
reluctant to use this solution.  Is there a bit of scheme code I can
insert or something like that to achieve the same effect without all
the warnings?  Ideally it would be something that is ignored when it
isn't neccessary, e.g., when the d is a quarter note or they are all
quarter notes so they wouldn't be beamed anyway.  Then my script can
just apply it every time it inserts \autoBeamOff.

Thanks for any suggestions,

Russ