I can run a search like the following,
curl -XGET 'http://localhost:9200/cms-2016-03-30/job/_search?pretty=true&size=1000' -d '{ "_source":"DESIRED_CMSDataset"}'
which gives results like
{
"_index" : "cms-2016-03-30",
"_type" : "job",
"_id" : "crab3-7@vocms0114.cern.ch#6472621.0#1459313328",
"_score" : 1.0,
"_source" : {
"DESIRED_CMSDataset" : "/BTagCSV/Run2015D-16Dec2015-v1/MINIAOD"
}
}
I would like to process this result to only get the MINIAOD part of the DESIRED_CMSDataset string. Various tokenizers exist, but I can't figure out the syntax for any of them.
If you can alternatively tell me how to do this at index time (precisely), that is an acceptable answer. I mostly need the syntax. I don't understand the ElasticSearch syntax