On 4 Feb 2011, at 11:11 , Florent Georges wrote: > G. Ken Holman wrote: > > Hi, > >> This output surprises me, though perhaps you are running >> "1.0-ml" and not "1.0" ... you don't say which. > >> You are not specifying the collation URI and so in XQuery 1.0 >> the collation order should be Unicode code-point order, which >> has upper-case first. > > My understanding of the spec is that the default collation in > the static context is implementation-defined. So MarkLogic can > define its own default collation, as a specific collation URI, or > based on a config switch. If it does not do so, then yes, the > default default collation is the Unicode code point order. > > So the questions are: which is the default in ML regarding case > ordering, and has the OP any collation configured in the admin > console?
Yes, that's correct as I understand it. xquery version "1.0-ml"; default-collation() => http://marklogic.com/collation/ The guide at http://developer.marklogic.com/learn/4.2/search-dev-guide has a section on "Encodings and Collations" which documents this. The default is the UCA collation (http://www.unicode.org/reports/tr10/). According to the documentation it "adds more useful case and diacritic sensitivity to the Unicode codepoint order". You can also declare a different default: xquery version "1.0-ml"; declare default collation "http://marklogic.com/collation/codepoint";; ... Or you can set the default collation for each application server. -- Mike _______________________________________________ General mailing list [email protected] http://developer.marklogic.com/mailman/listinfo/general
