JQOM - 'Duplicate' rows returned

[Former Community Member]

[img]structure.png[/img]

Hi guys, I am encountering some trouble with JCR-JQOM. The image above shows the structure I am working on.
Briefly:

• ​I have a set of pages in /content/mySite/myFolder
• Each one of these pages has a couple of sub-pages
• In the jcr:content of these sub-pages there is a property, let's say it's called myProperty, which is a String

Let's assume that the value of myProperty is "someValue" for the sub-pages of /content/mySite/myFolder/firstPage, while the sub-pages of /content/mySite/myFolder/secondPage have some different values for this property.

If you want to find all the pages in /content/mySite/myFolder that have a sub-page with the property "myProperty" set to "someValue", you can do something like

SELECT * FROM [cq:Page] AS page INNER JOIN [cq:Page] AS subPage ON ISCHILDNODE(subPage, page) INNER JOIN [nt:unstructured] AS subPageJcrContent ON ISCHILDNODE(subPageJcrContent, subPage) WHERE ISCHILDNODE(page, [/content/mySite/myFolder]) AND CONTAINS(subPageJcrContent.myProperty, 'someValue')

Running this query via CRX DE, you get the one row you are expecting to be returned, which is the node /content/mySite/myFolder/firstPage.

Everything sounds just fine till now, but...
The equivalent JQOM for this should be the following:

QueryObjectModelFactory qf = session.getWorkspace().getQueryManager().getQOMFactory(); Selector pageSelector = qf.selector("cq:Page", "page"); Selector subPageSelector = qf.selector("cq:Page", "subPage"); Selector subPageJcrContentSelector = qf.selector("nt:unstructured", "subPageJcrContent"); Join join = qf.join(pageSelector, subPageSelector, QueryObjectModelFactory.JCR_JOIN_TYPE_INNER, qf.childNodeJoinCondition(subPageSelector.getSelectorName(), pageSelector.getSelectorName())); join = qf.join(join, subPageJcrContentSelector, QueryObjectModelFactory.JCR_JOIN_TYPE_INNER, qf.childNodeJoinCondition(subPageJcrContentSelector.getSelectorName(), subPageSelector.getSelectorName())); Constraint constraint = qf.childNode(pageSelector.getSelectorName(), "/content/istituzionale20/myFolder"); constraint = qf.and(constraint, qf.fullTextSearch(subPageJcrContentSelector.getSelectorName(), "ingredients", qf.literal(vf.createValue("lol")))); QueryObjectModel query = qf.createQuery(join, constraint, null, null);

The query runs fine but the problem is: if I print the rows (this sample code is running within a JSP page) with the following code:

<% QueryResult queryResult = query.execute(); RowIterator rowIterator = queryResult.getRows();%> <% while(rowIterator.hasNext()) { Node currentNode1 = rowIterator.nextRow().getNode(ricettaPageSelector.getSelectorName()); %> This is the path: <%=currentNode1.getPath()%><br> <% } %>

the output is

This is the path: /content/mySite/myFolder/firstPage This is the path: /content/mySite/myFolder/firstPage

That is, the same node is returned twice. If the page /content/mySite/myFolder/firstPage would have had 3 pages with that property set at the desired value, the output would have been

This is the path: /content/mySite/myFolder/firstPage This is the path: /content/mySite/myFolder/firstPage This is the path: /content/mySite/myFolder/firstPage

I don't see why this is happening, and I am also curious about this difference between the result I get in CRX DE and the output of my JSP.

Did any of you know how to avoid this behaviour?

Thanks!

[Sam205505050]

This thread might help you

[Former Community Member]

Hi Sam, thanks for your reply.

Unfortunately, it seems to be a different issue. In his post, Dmitri says he got 1 node when querying via the javax.jcr.query.QueryManager, and 3 when using QueryObjectModelFactory, while I get "duplicated" rows in both cases.

[smacdonald2008 · Accepted solution]

Any reason why you using this API as opposed to the AEM QueryBuilder API - i would recommend the AEM QueryBuilder API? 

[Former Community Member]

I just found some tutorial using the QueryObjectModelFactory and I stuck to it since it was suggested as the "right way" to query the CRX (rather than just passing a query as a String). I know that QueryBuilder works in a similar way, but I read that it translates your parameters into an XPath query; isn't XPath deprecated?