Issue 17333

Wikipedia checklist normalization fails with nub matching unavailable

17333
Reporter: mdoering
Assignee: mdoering
Type: Bug
Summary: Wikipedia checklist normalization fails with nub matching unavailable
Priority: Critical
Resolution: Fixed
Status: Closed
Created: 2015-03-02 11:19:03.878
Updated: 2015-03-02 13:27:12.079
Resolved: 2015-03-02 13:27:12.052
        
Description: But there is nothing in the nub-ws logs and the service seems to be up and running all the time. Logging of the actual error is on debug level which is not very helpful in this case...

java.lang.IllegalStateException: Backbone matching service unavailable for at least 12h. Interrupting normalization!
    at org.gbif.checklistbank.cli.normalizer.UsageMetricsAndNubMatchHandler.matchToNub(UsageMetricsAndNubMatchHandler.java:154) ~[checklistbank-cli.jar:2.12-SNAPSHOT]
]]>
    


Author: mdoering@gbif.org
Created: 2015-03-02 12:18:22.723
Updated: 2015-03-02 12:41:39.818
        
DEBUG [2015-03-02 12:16:31,365+0100] [pool-9-thread-2] org.gbif.checklistbank.cli.normalizer.UsageMetricsAndNubMatchHandler: Nub matching for >Nereis southerni {{nowrap Abdel-Moez & Humphries, 1955< failed. Sleep and then retry

javax.ws.rs.core.UriBuilderException: java.net.URISyntaxException: Illegal character in query at index 153: http://api.gbif-uat.org/v1/species/match?phylum=Annelida&rank=SPECIES&kingdom=Animalia&family=Nereididae&verbose=false&name=Nereis+southerni+%7B%7Bnowrap Abdel-Moez & Humphries, 1955&class=Polychaeta&genus=Alitta&strict=true

    


Author: mdoering@gbif.org
Comment: Added wikipedia builder bug: https://github.com/mdoering/wikipedia-dwca/issues/17
Created: 2015-03-02 12:47:03.361
Updated: 2015-03-02 12:47:03.361


Author: mdoering@gbif.org
Created: 2015-03-02 13:08:07.001
Updated: 2015-03-02 13:08:07.001
        
The jersey client URI encoder treats curly brackets special and thinks of them as template variables.
Escaping does not seem to work, so replacing them with other brackets is the only solution for the name matching I can think of
    


Author: mdoering@gbif.org
Created: 2015-03-02 13:27:12.076
Updated: 2015-03-02 13:27:12.076
        
Escaping { into url encoded values works fine.
Fixed in https://github.com/gbif/checklistbank/commit/ab560bc842c764538cee1290396732131dc6af0b