Meaning most likely gatherNodes will not be able to satisfy the requirements?
Is there any thing in Solr that allow us to stored output into things like memory variable, and use that variable to pass it as an input for the next part of the query? Regards, Edwin On 5 May 2017 at 23:39, Joel Bernstein <joels...@gmail.com> wrote: > Yeah gatherNodes just supports one key traversals. > > Joel Bernstein > http://joelsolr.blogspot.com/ > > On Fri, May 5, 2017 at 10:54 AM, Zheng Lin Edwin Yeo <edwinye...@gmail.com > > > wrote: > > > Hi Joel, > > > > I may have several fields which requires to pass, so not sure if this > will > > works? > > Eg: I may have field1a, field1b, field1c from collection1 which requires > to > > be pass to field2a, field2b and field2c from collection2 respectively. > > > > And yes, my /export is working now. > > > > Regards, > > Edwin > > > > > > On 5 May 2017 at 20:24, Joel Bernstein <joels...@gmail.com> wrote: > > > > > But /export has to be working to support the gatherNodes() expression. > > > > > > Joel Bernstein > > > http://joelsolr.blogspot.com/ > > > > > > On Fri, May 5, 2017 at 7:40 AM, Joel Bernstein <joels...@gmail.com> > > wrote: > > > > > > > This may work for you: > > > > https://cwiki.apache.org/confluence/display/solr/Graph+Traversal > > > > > > > > Joel Bernstein > > > > http://joelsolr.blogspot.com/ > > > > > > > > On Fri, May 5, 2017 at 5:39 AM, Zheng Lin Edwin Yeo < > > > edwinye...@gmail.com> > > > > wrote: > > > > > > > >> Hi, > > > >> > > > >> I'm using Solr 6.4.2. > > > >> > > > >> Is it possible to do a query based on output from another collection > > in > > > >> Solr Streaming, so that the entire query can be passed at one go? > > > >> > > > >> For example, I have a field called reference_s in collection1. In > > > >> collection2, I will need to pass what is in the reference_s from > > > >> collection1 into the search in collection2. > > > >> > > > >> Currently, I will need to do a normal query to collection1, to > > retrieve > > > >> all > > > >> the required values in reference_s, and put them into this join > query, > > > >> which is a two step process. This is probably not so effective and > > also > > > >> slow down the search. > > > >> > > > >> > > > >> hashJoin(parallel(collection2, > > > >> workers=3, > > > >> sort="id asc", > > > >> innerJoin(search(collection1, q="*:*", > > > >> fl="id,reference_s", > > > >> sort="id asc", qt="/export", > > > >> partitionKeys="id"), > > > >> search(collection2, > > > >> q="reference_s:?", > fl="id, > > > >> year_i,reference_s", sort="id asc", qt="/export", > > > >> partitionKeys="id"), > > > >> on="id")), > > > >> > hashed=search(collection3, > > > >> q="day_i:7", fl="id, day_i", > > > >> sort="id asc", qt="/export"), > > > >> on="id") > > > >> > > > >> > > > >> Regards, > > > >> Edwin > > > >> > > > > > > > > > > > > > >
