Fix fulltext extraction for blobs fetched from a document dynamic facet

Steps to reproduce:

1. define a "custom" schema with a blob property

   <xs:element name="pdfprop" type="nxs:content" />
   

2. define a "customFacet" facet

   <extension target="org.nuxeo.ecm.core.schema.TypeService" point="doctype">
   <facet name="customFacet">
    <schema name="custom" />
    </facet>
   </extension>
   

3. dynamically add this facet to a document and set the blob property with a Blob value
4. make a search against a work from the Blob value
5. observe that the newly created document is not returned among the results

What happened is that BlobsExtractor#getBlobPaths seems to read only the static document schemas and therefore the blob property from the attached facet is missed. And therefore the FulltextExtractorWork does not extract the text from this blob and the blob is not fulltext indexed.

Expected behavior: blobs from attached facet are recognized as blob paths and are fulltext extracted/indexed