[MarkLogic Dev General] xdmp:unquote issues in 1.0-ml: "invalid root text"

[Eric Palmitesta]

Apologies, just missing a parameter, sorry for the noise.

EP

Eric Palmitesta wrote:
> Sorry Michael, I should have taken a closer look at the release notes.
> 
> xquery version "1.0-ml";
> let $value := <value>one &lt;b&gt;two&lt;/b&gt; three</value>
> return xdmp:unquote($value/node(), "repair-full")
> 
> This produces the same error as without the "repair-full" parameter.  I 
> have a feeling I'm just making a silly mistake here.
> 
> Eric
> 
> Michael Blakeley wrote:
>> Eric,
>>
>> The change is in the release notes 
>> (http://developer.marklogic.com/pubs/4.0/books/relnotes.pdf):
>>
>>> 4.2.10
>>> repair Option of xdmp:document-load, xdmp:unquote, and Others
>>> Now Default to none in 1.0 and 1.0-ml
>>> In the 1.0 and 1.0-ml dialects, the default for XML repairing
>>  > option is now none instead of full.
>>
>> You can still supply the 'repair-full' option in 1.0-ml, if you wish.
>>
>> For your second test, with the extra element wrapper, xdmp:describe() 
>> is your friend:
>>
>> let $value := <value>one &lt;b&gt;two&lt;/b&gt; three</value>
>> return xdmp:describe(
>>   xdmp:unquote(concat('<t>', $value/node(), '</t>')) )
>> =>
>> document{<t>one <b>two</b> three</t>}
>>
>> Per docs 
>> (http://developer.marklogic.com/pubs/4.0/apidocs/Extension.html#xdmp:unquote), 
>> xdmp:unquote() returns document-node()+. This is similar to the 
>> behavior of fn:doc(), fn:collection(), etc.
>>
>> -- Mike
>>
>> On 2009-04-24 12:43, Eric Palmitesta wrote:
>>> Anyone know what's changed in xdmp:unquote when moving from 0.9-ml to
>>> 1.0-ml?
>>>
>>> xquery version "0.9-ml"
>>> let $value :=<value>one&lt;b&gt;two&lt;/b&gt; three</value>
>>> return xdmp:unquote($value/node())
>>> =>
>>> <v:results v:warning="more than one node">
>>>     one
>>>     <b>two</b>
>>>     three
>>> </v:results>
>>>
>>> xquery version "1.0-ml";
>>> (: same code as above :)
>>> let $value :=<value>one&lt;b&gt;two&lt;/b&gt; three</value>
>>> return xdmp:unquote($value/node())
>>> =>
>>> [1.0-ml] XDMP-DOCROOTTEXT: xdmp:unquote("one<b>two</b>  three") --
>>> Invalid root text "one "
>>>
>>> This works:
>>>
>>> xquery version "1.0-ml";
>>> let $value :=<value>one&lt;b&gt;two&lt;/b&gt; three</value>
>>> return xdmp:unquote(concat('<t>', $value/node(),'</t>'))
>>> =>
>>> <t>
>>>     one
>>>     <b>two</b>
>>>     three
>>> </t>
>>>
>>> However, appending /node() doesn't change the output.  I would expect
>>> <t>  and</t>  to vanish:
>>>
>>> xquery version "1.0-ml";
>>> let $value :=<value>one&lt;b&gt;two&lt;/b&gt; three</value>
>>> return xdmp:unquote(concat('<t>', $value/node(),'</t>'))/node()
>>> =>
>>> <t>
>>>     one
>>>     <b>two</b>
>>>     three
>>> </t>
>>>
>>> These examples were run on CQ, using MarkLogic 4.0-3, linux x86_64.
>>>
>>> Eric
>>> _______________________________________________
>>> General mailing list
>>> General at developer.marklogic.com
>>> http://xqzone.com/mailman/listinfo/general
>>
>> _______________________________________________
>> General mailing list
>> General at developer.marklogic.com
>> http://xqzone.com/mailman/listinfo/general
> _______________________________________________
> General mailing list
> General at developer.marklogic.com
> http://xqzone.com/mailman/listinfo/general