Cannot get pdf custom properties values - custom metadata ex

[shikarishambu]

I have a custom content model with a number of attributes. When I create pdf or Word documents that need to be specialized to this content type I create custom properties on the document and let the custom metadata extractor work at extracting the metadata info

This seems to work well with Word document. I have a rule on the space where the documents are loaded to - specialize the content type, extract content. However, in case of pdf it seems to only pull the default pdf extract properties. What is it that I am missing? How do I pull custom properties out of pdf i.e. pdfx:mycustomproperty and not pdf:mycustomproperty.

Here is a snippet of my custom-metadata-extrators-context.xml

  <bean id="extracter.OfficeDocument"
  class="org.alfresco.repo.content.metadata.OfficeMetadataExtracter"
  parent="baseMetadataExtracter">
    <property name="inheritDefaultMapping">
      <value>true</value>
    </property>
    <property name="mappingProperties">
      <props>
        <prop key="namespace.prefix.mymodel">http://www.mymodelsoln.com/model/content/1.0</prop>
        <prop key="namespace.prefix.cm">http://www.alfresco.org/model/content/1.0</prop>
      <prop key="docid">mymodel:docid</prop>
        <prop key="comments">mymodel:comments</prop>
        <prop key="category">mymodel:category</prop>
      <prop key="subcategory">mymodel:subcategory</prop>
        <prop key="policyno">mymodel:policyno</prop>
        <prop key="claimno">mymodel:claimno</prop>
      <prop key="agentno">mymodel:agentno</prop>
      <prop key="providerno">mymodel:providerno</prop>
      <prop key="contractno">mymodel:contractno</prop>
       <prop key="govtid">mymodel:govtid</prop>
      <prop key="effectivedate">mymodel:effectivedate</prop>
      <prop key="product">mymodel:product</prop>
      <prop key="region">mymodel:region</prop>
      <prop key="keywords">mymodel:keywords</prop>
      </props>
    </property>
  </bean>
  <bean id="extracter.PDFBox"
  class="org.alfresco.repo.content.metadata.PdfBoxMetadataExtracter"
  parent="baseMetadataExtracter">
    <property name="inheritDefaultMapping">
      <value>true</value>
    </property>
    <property name="mappingProperties">
      <props>
        <prop key="namespace.prefix.mymodel">http://www.mymodelsoln.com/model/content/1.0</prop>
        <prop key="namespace.prefix.cm">http://www.alfresco.org/model/content/1.0</prop>
      <prop key="docid">mymodel:docid</prop>
        <prop key="comments">mymodel:comments</prop>
        <prop key="category">mymodel:category</prop>
      <prop key="subcategory">mymodel:subcategory</prop>
        <prop key="policyno">mymodel:policyno</prop>
        <prop key="claimno">mymodel:claimno</prop>
      <prop key="agentno">mymodel:agentno</prop>
      <prop key="providerno">mymodel:providerno</prop>
      <prop key="contractno">mymodel:contractno</prop>
       <prop key="govtid">mymodel:govtid</prop>
      <prop key="effectivedate">mymodel:effectivedate</prop>
      <prop key="product">mymodel:product</prop>
      <prop key="region">mymodel:region</prop>
      <prop key="keywords">mymodel:keywords</prop>
      </props>
    </property>
   </bean>‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍

[derek]

Hi,
Look at the javadocs of the metadata extractor implementations.  The PDFBox metadata extractor doesn't extract custom properties.  It will be possible to do something like this:

                // Extract remaining custom properties
                for (String customProp : super.getMapping().keySet())
                {
                    if (rawProperties.keySet().contains(customProp))
                    {
                        // Ignore it
                        continue;
                    }
                    String customValue = docInfo.getCustomMetadataValue(customProp);
                    putRawValue(customProp, customValue, rawProperties);
                }
‍‍‍‍‍‍‍‍‍‍‍‍‍

after the catch code.  I am putting this into the extractor and testing it now.  Implement your own class to get the change now.

Regards

[shikarishambu]

Derek, thanks for the info. Are you committing the change to the codebase? Will it be available out-of-box in a later version of Alfresco? Or, should I continue to rely on my changes?

TIA

[derek]

Hi,
I have committed the change to 3.2.1 (Entperprise customers) and will merge the change in for 3.3 Community as soon as it passes sanity checking.
Regards

[cerberos]

Hi,
got some news about this? I've created a PDF using PDF Box and added a custom property called "prop1". PDFBox library can read it.
I've set a custom mapping for the PDFBox Extractor, using this line:

…
<prop key="prop1">cm:description</prop>
…
‍‍‍‍‍

but when i add a new PDF file nothing appears. I'm using Alfresco Community 3.2-r2, got some news with Alfresco Community 3.3?
How can I develop a custom extracter to work with PDFs, overriding the default extractor???

Thank you.

[cerberos]

Ok, seems to work with Alfresco 3.3g. Now PDFBox Metadata Extractor uses Apache TIKA. Dunno how it works, but seems to be ok!

Cheers