Hi,
Is there a dynamic_template setting, or a dynamic/object setting that can
for type "object" be indexed as a searchable string blob?
Our structure looks like this:
{
"doc_type": "foo",
"date_received": "2013-01-01", # mapped as date
"form": {
"username": "dan",
"timestart": "2013-01-01 11:58:00", #mapped as date
"timeend": "2013-01-01 11:59:00", #mapped as date,
"prop1": "lala",
"prop2": "dada",
"prop3": {
"suba": "something",
"subb": "anything",
}
}
}
The issue with our data is that due to the it being user defined document
structures, we don't know if someone will create "prop3" as a string or a
date, or a mapping.
We've set date_detection to False, but the possible nested objects have
prevented us from indexing unknown properties (we have a common, system
defined index for our known properties)
What we're wondering is can the dynamic_template work to:
1: store objects like form.prop3 as a searchable string (knowing full well
that it'd be difficult to search)
2: if we set store=True, can we ensure that form.prop3 will be in its
original json format and not stringified?
My attempt at the dynamic template:
"dynamic_templates": [
{
"object_fields": {
"match": "form.*",
"match_mapping_type": "object",
"mapping": {
"type": "string"
}
}
}
],
doesn't seem to pick up any of the properties despite the mapping being set
to dynamic=true for the form property.
Is what I'm asking for possible?
Thanks,
Dan
--
You received this message because you are subscribed to the Google Groups "elasticsearch" group.
To unsubscribe from this group and stop receiving emails from it, send an email to elasticsearch+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.