If you are using GLASS (
http://seaside.gemstone.com), you add your data to
an UnorderedCollection (typically an IdentitySet) and then ask it to keep
things indexed for you.
myCollectionOfData
createEqualityIndexOn: title withLastElementClass:
String;
createEqualityIndexOn: creationDate withLastElementClass:
Date ;
createEqualityIndexOn: modifiedDate withLastElementClass:
Date ;
createEqualityIndexOn: type withLastElementClass:
Symbol;
yourself.
>From then on the indexes are automatically updated based on adds and
removes. Queries use a special "select block" syntax:
recentDate := Date today subtractDays: 10.
recentlyModified := myCollectionOfData select: {:each |
each.modifiedDate >= recentDate}.
James
On Nov 25, 2007 10:22 PM, <
[hidden email]> wrote:
> Hi,
>
> my data class has following attributes;
>
> title
> content
> creation date
> modified date
> type
>
> and except content, the collection of
> elements of this data shlould be indexed; that is,
> title, creation/modified date and type be key.
>
> If single key is required, then dictionary will be my choice(or
> should I use tree?). What is the best data structure for multiple
> keyed data? This question has been one of my hurddle to the
> OODB; if like now, sql db will be used, I know how to handle this.
>
> Thanks in advance.
_______________________________________________
seaside mailing list
[hidden email]
http://lists.squeakfoundation.org/cgi-bin/mailman/listinfo/seaside
Locked
