Hi Ryan,

For commons, make your changes and make a pull request. Someone in MarkLogic
will review your changes!

Thanks,
Nuno


On Fri, Apr 15, 2011 at 5:19 AM, Geert Josten <[email protected]>wrote:

> Hi Ryan,
>
>
>
> I’d say, contact [email protected]. They can file it as a bug for you.
>
>
>
> It should also be available in the marklogic/commons on github I think. You
> could fork, patch and do a pull request. I could do that too for you. For a
> one-line-fix that shouldn’t be much trouble.. ;-)
>
>
>
> Kind regards,
>
> Geert
>
>
>
> *Van:* [email protected] [mailto:general-b
> [email protected]] *Namens *[email protected]
> *Verzonden:* vrijdag 15 april 2011 0:24
> *Aan:* [email protected]
> *Onderwerp:* [MarkLogic Dev General] Bug in in-mem-update
>
>
>
> The following code throws this error: [1.0] XDMP-ATTRSEQ: (err:XQTY0024)
> $node/@* -- Attribute node cannot follow non-attribute node in element
> content:
>
>
>
> xquery version '1.0-ml';
>
> import module namespace mem = "http://xqdev.com/in-mem-update"; at
> "/MarkLogic/appservices/utils/in-mem-update.xqy";
>
> let $test :=
>     <root type="a">
>         <message>Hello</message>
>     </root>
>
> return mem:node-insert-child($test, <name>bob</name>)
>
>
>
> Because the node "root" has an attribute and child elements already, so
> when adding another child node, the attribute node is not added first.
>
> It looks like line 121 in in-mem-update.xqy has a bug where it has:
>
>
> then element { QName(namespace-uri($node), local-name($node)) } { (
>                     $node/node(), $node/@*, $newNode
>                 ) }
>
>
> where I think it should be:
>
> then element { QName(namespace-uri($node), local-name($node)) } { (
>                     $node/@*, $node/node(), $newNode
>                 ) }
>
> Can someone confirm? Works for me if I make the above change. Also, I don't
> know an easy fix for this given that this is code included with MarkLogic. I
> don't really want to have to remember to change this code on all ML
> instances. I think the only other option at this point is to make a
> corrected copy of the code and use that instead.
>
> Suggestions?
>
> Thanks,
> Ryan
>
> _______________________________________________
> General mailing list
> [email protected]
> http://developer.marklogic.com/mailman/listinfo/general
>
>
_______________________________________________
General mailing list
[email protected]
http://developer.marklogic.com/mailman/listinfo/general

Reply via email to