Rob Browning <[EMAIL PROTECTED]> writes:

> I've finally gotten around to looking at this, and are you sure it's
> correct?  It seems to assume that hashes will always be founded on
> vectors of empty lists, and I don't think that's guaranteed.  In fact
> I've always started my hashes with no values.  For example, I believe
> this is a functionally acceptable guile hash:

OK, I've checked with the guile people, and while (make-vector 17)
might *work*.  You're really supposed to use (the completely AFAICT
undocumented) make-hash-table.  I don't know whether or not a prime
number is preferred for the size, but I wouldn't be surprised.  Oh,
and these hashes are not resizing.

In any case, if we use make-hash-table, as you're supposed to, then
your original code should work just fine.

Also, for anyone who's interested (and if we need it), Dirk mentioned
that Jay Glascoe had implemented a self-resizing hash-table that's
much more flexible than the default guile ones.  If someone feels
motiviated, the source should be at
http://www.giss.nasa.gov/~jglascoe/GUILE/

Thanks

-- 
Rob Browning <[EMAIL PROTECTED]> PGP=E80E0D04F521A094 532B97F5D64E3930

--
Gnucash Developer's List 
To unsubscribe send empty email to: [EMAIL PROTECTED]

Reply via email to