WikiDiscuss

On Tue, Mar 16, 2004 at 12:51:25PM -0800, Jorge Llamb?as wrote:
>
> --- Robin Lee Powell <rlpowell@digitalkingdom.org> wrote:
> > On Tue, Mar 16, 2004 at 11:51:05AM -0800, Jorge Llamb?as wrote:
> > > > In a new pass, perform the following absorptions (absorption
> > > > means that the token is removed from the grammar for
> > > > processing in following steps, and optionally reinserted,
> > > > grouped with the absorbing token after parsing is
> > > > completed).
> > >
> > > AS I understand it, the token that is absorbed (and thus removed)
> > > is the BU token, not the one that gets relabeled.
> >
> > That seems unlikely, as the text above refers to the 'absorbing
> > token' as well as the token that is absorbed. In 'mi bu', it seems
> > very strange to refer to 'mi' as the 'absorbing token'.
>
> The absorbing token is the one left behind, labeld as BY.

You must be reading a very different paragraph than the one I'm reading.

"the token is removed ... and optionally re-inserted, *grouped* *with*

• the* *absorbing* *token*"

That implies to me that there are two tokens, the absorbed and the
absorbing, that they are treated seperately, and that neither of them is
the new virtual token, because you wouldn't group that with anything if
you are re-inserting things for human readability.

> It doesn't matter whether you think of it as {mi} or as {bu}, in fact
> it really is {mibu}. In any case, the absorbing token left behind is
> available to interact with the next bu. The paragraph may not be very
> clear, but the meaning you propose does not make sense, what would be
> the point of removing the newly created BY token?

I have no idea, but I'm trying to reconcile that parapgraph with both
the informal grammar production given for BU and the actual behaviour of
the official parser.

> > > The reinsertion occurs after all parsing is done, for the benefit
> > > of the human.
> >
> > In which case, there definately isn't going to be anything there for
> > a second 'bu' to attach to.
>
> The "reinsertion" is the reconversion of the virtual BY into {mi bu}.
>
> The preparser changes {mi bu bu} into BY.

Apparently not.

> Then, after all parsing is done, the first reinsertion changes BY into
> BY bu, and a second reinsertion changes BY bu into mi bu bu. The
> reinsertions are said to be optional because they don't affect the
> parsing.

That makes perfect sense, but it isn't true in practice and I'm not sure
it's what the grammar calls for.

-Robin