BioCASe ABCD 1.2 metadata synchronisation doesn't work

14083
Reporter: omeyn
Assignee: lfrancke
Type: Bug
Summary: BioCASe ABCD 1.2 metadata synchronisation doesn't work
Priority: Critical
Resolution: Fixed
Status: Closed
Created: 2013-09-27 10:03:29.546
Updated: 2016-02-15 12:16:24.249
Resolved: 2013-09-27 14:36:49.329
        
Description: Many datasets are failing this validation, and while it may be a correct interpretation of the protocol(s), it means datasets that we've happily indexed in the past are no longer working. We should be more lenient on Title at least, possibly others.

Here's one example (this one happens to be abcd 1.2):

INFO  [2013-09-27 09:39:17,394+0200] [pool-4-thread-6] org.gbif.crawler.registry.metasync.MetasyncService: Done syncing. Processing result.
INFO  [2013-09-27 09:39:17,394+0200] [pool-4-thread-6] org.gbif.registry.metasync.resulthandler.DebugHandler: Installation [60416b42-f762-11e1-a439-00145eb45e9a] synced successfully. [0] added, [0] deleted, [4] updated
INFO  [2013-09-27 09:39:17,394+0200] [pool-4-thread-6] org.gbif.registry.metasync.resulthandler.RegistryUpdater: Updating dataset [82bb6da8-f762-11e1-a439-00145eb45e9a]
WARN  [2013-09-27 09:39:17,507+0200] [pool-4-thread-6] org.gbif.common.messaging.MessageConsumer: Error handling message, will be acknowledged anyway and not retried
javax.validation.ValidationException: com.sun.jersey.api.client.UniformInterfaceException: HTTP 422: 
Validation of [title] failed: may not be null
        at org.gbif.ws.client.interceptor.HttpErrorResponseInterceptor.invoke(HttpErrorResponseInterceptor.java:74) ~[crawler-cli.jar:na]
        at org.gbif.registry.ws.client.BaseNetworkEntityClient.update(BaseNetworkEntityClient.java:29) ~[crawler-cli.jar:na]
]]>
    

Author: lfrancke@gbif.org
Created: 2013-09-27 10:20:41.487
Updated: 2013-09-27 10:20:41.487
        
[~jlegind@gbif.org] What do you suggest we do?

I see two options:
* Registry could allow empty titles ([~trobertson@gbif.org])
* Metadata sync could fill in a "mock" title. Any suggestions what that should be?
    

Author: jlegind@gbif.org
Created: 2013-09-27 10:28:12.328
Updated: 2013-09-27 10:28:12.328
        
[~lfrancke@gbif.org] - I suspect that this might be limited to ABCD 1.2 resources because they do not have the title element, but something called "SourceName":

#--DATASET.DATASET.TITLE --
metadatadataResourceName=*/DataSets/DataSet/OriginalSource/SourceName

See:
http://code.google.com/p/gbif-indexingtoolkit/source/browse/trunk/harvest-webapp/src/main/resources/org/gbif/harvest/biocase/biocase_1_3/mapping/indexMapping_abcd_1_2.properties?r=1419

 
    

Author: lfrancke@gbif.org
Created: 2013-09-27 14:35:25.47
Updated: 2013-09-27 14:35:25.47
        
Jan was totally correct. Thanks for pointing out the link to ABCD 1.2

Here is the fix (one character): https://code.google.com/p/gbif-labs/source/detail?r=632
    

Author: lfrancke@gbif.org
Comment: The underlying issue still might occur with datasets that don't have a title though. I'll rename the title of this issue and will close it. I propose creating a new one should we run into genuine datasets without titles.
Created: 2013-09-27 14:36:19.242
Updated: 2013-09-27 14:36:19.242