[midPoint] Binary attribute, ScriptedSQL, Groovy and Oracle Blob - the ultimate guide

[Wojciech Staszewski]

For now, the code below is working as expected: 

1) MidPoint
"stamp" (image) attribute in schema extension with "xsd:base64Binary"
type. 

2) SchemaScript.groovy:


account.addAttributeInfo(AttributeInfoBuilder.build("stamp",
byte[].class)); 

3) SearchScript.groovy (read the stamp from database):


def stampquery = "select stamp from users" + where;
stamp_blob =
sql.firstRow(stampquery).stamp;
if ( stamp_blob ) { stamp_in =
stamp_blob.getBinaryStream().getBytes(); } else { stamp_in = null }


and set the attribute: stamp:stamp_in, 

4) UpdateScript.groovy (write
the stamp to datablase) 

case "stamp":
 // oracle "blob" is an
interface, write to the blob can be done using BinaryOutputStream.
 //
first, we need to lock the blob in the database by running "select for
update" query:
 blob_out = (oracle.sql.BLOB) sql.firstRow("select stamp
from users where id = ? for update",[uid]).stamp;
 // now we have to
define binary output stream to write the blob:
 outstream =
((BLOB)blob_out).getBinaryOutputStream();
 // and write the attribute
received as a byte array to this stream:
 stamp_out =
attributes?.get("stamp")?.get(0);
 outstream.write(stamp_out);
 // at
the end we close the stream.
 outstream.close();
 break;


---------
Have fun.
Regards!
WS 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.evolveum.com/pipermail/midpoint/attachments/20171122/84e7421f/attachment.htm>