AttributeMap (was Re: Announcement: SAX 1998-01-12 Draft)

[Matthew Gertner]

>[Matthew Gertner]
>>Okay, I thought you were talking about the instantiation of the map
>>implementation itself. The fact of the matter is that, although there is
>>some overhead to instantiating an iterator object, looping over a list of
>>attributes and doing n string compares is not all that efficient either.
>
>Given that attribute ordering is never significant could the attributes
>be provided sorted by name so that by-name look up can be achieved with
>a binary chop?


This kind of implementation detail should be hidden from the consumer
application. If Jame's AttributeList interface were used, the String
get(String) method could be made more efficient through alphabetical
sorting, as you suggest. This would, however, require that the attributes be
sorted in the first place, so with any significant number of attributes it
might be more efficient to instantiate an iterator on a map than to sort the
attributes. Of course, the map insert probably has log complexity instead of
constant for the list...

When you get down to it, we need an iterator-type interface and a map-type
interface. The area of controversy seems to be whether these are both
provided on top of one implementation, or whether a new implementation is
instantiated for the iteration interface. I personally prefer the second
variant but I guess it doesn't matter much.

Matthew



xml-dev: A list for W3C XML Developers. To post, mailto:xml-dev at ic.ac.uk
Archived as: http://www.lists.ic.ac.uk/hypermail/xml-dev/
To (un)subscribe, mailto:majordomo at ic.ac.uk the following message;
(un)subscribe xml-dev
To subscribe to the digests, mailto:majordomo at ic.ac.uk the following message;
subscribe xml-dev-digest
List coordinator, Henry Rzepa (mailto:rzepa at ic.ac.uk)