request: ^ and constant searching

[Russ Cox]

> I don't really see how this would help searching for sequences
> differing only in the initial term(s), unless you know exactly which
> variation you are looking for.  To see if another sequence is similar,
> grab several terms starting at the fifth or so, and search for that.  

This doesn't help Eric's initial problem, namely finding whether
a sequence beginning 2,3,4,5,6,7,8,9 is in the database.
I am leaning toward ranking matches starting at the beginning
of a sequence higher than matches in the middle, so that
the early matches sift to the top.

> The second suggestion seems more useful to me.  However, care should
> be taken; decimal strings can occur in text as well, and we don't want
> to lose the ability to search for them.  I would suggest using a unique
> prefix (perhaps cons:) if this feature is supported.

The decimal search is now implemented.  Decimal strings can occur
in text too, but they do not occur very often, and you don't really have
the ability to search for them anyway, since most punctuation (including dots)
turns into spaces in generic searches.  (Search for "russ.cox".)

Russ