On 26.05.14 09:51, Philippe Marschall wrote:
Hi
I have been investigating why Dictionary look up performance with String
keys is not as good as I would expected. Something I noted is that
String >> #= is implemented in terms of #compare:with:collated:. There
is no short circuit if Strings are not the same size. In my case some
Strings have the same prefix but a different length eg 'Content-Type'
and 'Content-Length'. In that case a #compare:with:collated: is
performed even though we know in advance the answer will be false
because they have different sizes.
Some follow up on this:
In Seaside we have a custom dictionary GRSmallDictionary which is more
or less an OrderedCollection of Associations (not actually but
conceptually). I did a microbenchmark with 11 keys [1] and out of the
box Dictionary is about twice as fast as GRSmallDictionary. However when
I change GRSmallDictionary to first send and compare #size before
sending #= then GRSmallDictionary is slightly faster than Dictionary.
[1] #('Content-Type' 'Content-Language' 'Content-Length' 'Date'
'Last-Modified' 'Location' 'Set-Cookie' 'Set-Cookie2' 'Servlet-Engine'
'Status' 'WWW-Authenticate')
Cheers
Philippe
