4 Replies Latest reply on Feb 5, 2013 6:30 AM by aroxenhag

Initial content xml text properties are split into multi valued properties

aroxenhag Feb 4, 2013 2:48 PM

I have some initial content that gets imported. It appears though, that for some reason text properties are split on commas and set on my node as multivalued properties. But I'd like the strings stay as is and get imported as one value. This is a part of the xml:

<?xml version="1.0" encoding="UTF-8"?>

<jcr:root xmlns:jcr="http://www.jcp.org/jcr/1.0">

<story jcr:primaryType="wmod:story"

title="Winter vomiting bug threatens to ruin Christmas for thousands"

lead="Families face ruined Christmas celebrations as the worst-ever outbreak of the winter vomiting bug escalates sharply.">

<body>This is the body text. Please beware of commas until as they will turn this single valued property into a multi valued one..</body>

</story>

</section>

...

I suspect there is some sequencing going on, replacing my text with a split version of it. I've seen there is config for text sequencers to do that anyway.

How do I tell the import to not split the strings?

My repository config is quite simple, without any sequencer config.

1. Re: Initial content xml text properties are split into multi valued properties

hchiorean Feb 5, 2013 3:50 AM (in response to aroxenhag)

Hi Anders,

If there aren't any sequencers configured, then there isn't any sequencing going on. This could be a potential bug with the parsing of the initial content files.
Is there any chance you could attach your xml file (with commas in it) ?

Thanks
Actions
2. Re: Initial content xml text properties are split into multi valued properties

aroxenhag Feb 5, 2013 5:30 AM (in response to hchiorean)

This is a full example where lead and body text are split by the comma in the text. I tested providing the strings as attribute and sub element but there is no difference in the handling.

<?xml version="1.0" encoding="UTF-8"?>
<jcr:root xmlns:jcr="http://www.jcp.org/jcr/1.0">
<sites jcr:primaryType="wmod:section" title="Sites">
    <mst jcr:primaryType="wmod:section" title="Times">
      <section jcr:primaryType="wmod:section" title="News">
        <section jcr:primaryType="wmod:section" title="Politics">
          <story jcr:primaryType="wmod:story"
            title="Story with commas in the text lead and body text"
            lead="Lead text in attribute, split with a comma.">
            <body>Body text in sub element, also split by the comma.</body>
          </story>
        </section>
      </section>
    </mst>
</sites>
</jcr:root>

I am using 3.1.1.Final. Not sure what other info is useful. As I said, I have a very simple repository config, not specifying any sequencers.

The story type is defined like this:

[wmod:content] > nt:unstructured, mix:referenceable, mix:created, mix:lastModified
[wmod:story] > wmod:content
Actions
3. Re: Initial content xml text properties are split into multi valued properties

hchiorean Feb 5, 2013 6:14 AM (in response to aroxenhag)

Had a look at the code and it seems we always use the "," character as an indicator of multi-valued properties and split accordingly.

Can you please log an enhancement request and we'll try to see if there's another way of handling/specifying them in the XML format.

Thanks
Actions
4. Re: Initial content xml text properties are split into multi valued properties

aroxenhag Feb 5, 2013 6:30 AM (in response to hchiorean)

Created https://issues.jboss.org/browse/MODE-1788.

Thanks
Actions

Go to original post