Confusion about dictionaries - keys use value or identity?

Steve Holden sholden at holdenweb.com
Mon Jul 9 00:38:36 CEST 2001


"Tim Peters" <tim.one at home.com> wrote in message
news:mailman.994626808.10039.python-list at python.org...
> [Steve Holden]
> > ...
> > Key comparison of strings will test the lengths fisrt, and if the
> > lengths are the same will stop once two unequal characters are found...
>
> That's backwards.  If it tested the lengths first, then e.g. "z" < "aaa"
> would be true.
>
> "perl"<"python"-is-true-either-way-though-ly y'rs  - tim
>
We are talking about has_key here, right?

I was simply trying (perhaps rather too hurriedly, as it turned out) the
fact that strings if unequal lengths could not be equal, and only equality
is required when indexing a dictionary. Without looking at the code I'm
presuming that actually some fairly crafty hashing algorithm is used to test
for the presence of a particular key. But when more that one entry hashes to
the same bucket, I hope the lengths are checked before
character-by-character comparison goes ahead ... you know the details, I
don't, so I'm not going to press this issue any further. And no, I haven't
forgotten that strings are only a subset of the possible keys.

Which is why I said "key comparison of strings". I know you know that
locating a dicitonary entry isn't the same thing as comparing strings
lexically.

and-we-all-assume-python>perl-anyway-ly y'rs  - steve
--
http://www.holdenweb.com/








More information about the Python-list mailing list