It's normal to process documents (e.g. in your load function if any, or
using a CPF pipeline) to add an attribute that contains a normalized
dateTime value that you can index.
Wayne.
On 07/06/2010 10:25 AM, Bob Runstein wrote:
Our application must accept date information in multiple formats. We
may receive xs:dateTime, xs:date, xs:gYearMonth and xs:gYear and even
values such as "Unknown".
We want to be able to create range indexes on the date fields.
Naturally a dateTime index throws a cast exception when there is no
time component and we do not want to artificially modify the incoming
data. A string index would handle the formats correctly. Is there a
performance penalty for using a string index to do range searches on
dates formatted as above? What is the order of magnitude of the
penalty if there is one. Thanks.
Bob
--
Wayne Feick
Lead Engineer
MarkLogic Corporation
Phone +1 650 655 2378
Cell +1 408 981 4576
www.marklogic.com
_______________________________________________
General mailing list
[email protected]
http://developer.marklogic.com/mailman/listinfo/general
