Skip over doc_count_error_upper_bound when deserializing aggregations

[tjrobinson]

Deserializing a search response with bucket aggregations in it fails because the doc_count_error_upper_bound is unexpected. Take this response snippet for example:

... "aggregations":{"myBucket":{"doc_count_error_upper_bound":105322,"buckets":[{"key":"00000000-1111-1111-0000-070000000000","doc_count":202391...

This commit skips over this value and lets the method continue as intended. This has been tested on elasticsearch 1.4.0.beta1.

Without this change I'm unable to use bucket aggregations.

[gmarz]

Hey @tjrobinson,

Thank you very much for this PR and pointing this out!

There are actually quite a few changes in 1.4 Beta1 that are not compatible with NEST at the moment, this of course being one of them. We are in the process of going through these changes and making sure 1.4 will be fully compatible with NEST (without breaking compatibility with earlier versions).

That said, I don't think we can simply skip over this property as it's a valid part of the terms aggregation response and should be included. Working on a solution now for this as well as fixing the other aggregations that no longer work in 1.4. Stay tuned.

[gmarz]

@tjrobinson Just pushed a3dcb39 which fixes the issue while supporting the new response. Thanks again for bringing this to our attention!

[tjrobinson]

@gmarz Great, thanks.