Sorry, please disregard. Most recent documentation clarifies that we were doing "anonymous" child indexing which is no longer recommended.
On Thu, Jul 29, 2021 at 2:16 PM Tim Allison <[email protected]> wrote: > > Hi All, > > I'm working with Nicholas DiPiazza on the Solr emitter for Tika 2.x. > We're trying to model all embedded documents for a given input > document as children of that input document. > > We're using Solrj [1] and addChildDocument(). It feels from the > documentation [2] that we should get the _nest_parent_ and _nest_path_ > populated automatically, but we're not seeing that even though we're > defining those fields. > > Can someone take a look and let us know what we're doing wrong at > indexing time or in our expectations for search time? The details are > here: https://issues.apache.org/jira/browse/TIKA-3495 > Thank you! > > Best, > > Tim > > [1] > https://github.com/apache/tika/blob/main/tika-pipes/tika-emitters/tika-emitter-solr/src/main/java/org/apache/tika/pipes/emitter/solr/SolrEmitter.java#L108 > > [2] https://solr.apache.org/guide/8_9/indexing-nested-documents.html
