I am using SOLR 7.7.2 and trying to index binary data that is stored in
postgresql's large object feature (OID type / lo module) and not directly
in the database. Is this possible? If so are there any examples of others
configuring SOLR in this way?
Attached are my db-data-config and managed-schema files for reference.
This same file works against an oracle database with the same data.
<dataConfig>
<dataSource
type="JdbcDataSource"
name="ISO17025V3"
driver="org.postgresql.Driver"
url="jdbc:postgresql://IS-Config-DB:5432/ISO17025V3"
batchSize="0"
user="postgres"
password="xxxxx"
/>
<!-- always needed regardless of database -->
<dataSource name="fieldReader"
type="FieldStreamDataSource"
/>
<!-- document and field definitions - DB_DOCUMENT table -->
<uniqueKey>file_name</uniqueKey>
<document>
<entity
name="root"
query="select d.file_name, dbf.file_contents, d.file_label, d.version, d.dir_num
from db_document d
inner join db_files dbf on (d.file_name = dbf.original_file_name and d.version = dbf.document_version and d.revision_no = dbf.revision_no )
where dbf.file_contents is not null and dbf.parent_file_name is null
and d.version = (select max(version) from db_document d2 where d.file_name = d2.file_name)
order by d.file_label"
deltaImportQuery="select d.file_name, dbf.file_contents, d.file_label, d.version, d.dir_num
from db_document d
inner join db_files dbf on (d.file_name = dbf.original_file_name and d.version = dbf.document_version and d.revision_no = dbf.revision_no )
where dbf.file_contents is not null and dbf.parent_file_name is null
and d.version = (select max(version) from db_document d2 where d.file_name = d2.file_name) and d.file_name = '${dih.delta.file_name}'
order by d.file_label"
deltaQuery="select d.file_name as file_name
from db_document d
inner join db_files dbf on (d.file_name = dbf.original_file_name and d.version = dbf.document_version and d.revision_no = dbf.revision_no )
where dbf.file_contents is not null and dbf.parent_file_name is null
and d.version = (select max(version) from db_document d2 where d.file_name = d2.file_name) and dbf.changed_on > to_date('${dataimporter.last_index_time}', 'YYYY-MM-DD HH:MI:SS')
order by d.file_label"
transformer="TemplateTransformer"
onError="skip"
dataSource="ISO17025V3">
<field column="file_name" name="file_name" />
<entity
name="blob2"
dataSource="fieldReader"
processor="TikaEntityProcessor"
dataField="root.file_contents" format="text" onError="skip" extractEmbedded="true">
</entity>
</entity>
</document>
</dataConfig>