I might not fully understand how you would like to combine them. The possible reason is that [subquery] expect regular Solr Response to act on, but [child] might yield something hairish.
On Wed, Oct 9, 2019 at 2:40 PM Bram Biesbrouck < bram.biesbro...@reinvention.be> wrote: > Hi Mikhail, > > You're right, I should file an issue for the doc thing, I'll look into it. > > Thanks for pointing me towards parsing the _nest_path_ field. It's exactly > what ChildDocTransformer does, indeed. > > Would you by any chance know why [child] and [subquery] can't be combined? > They don't look too related to me and I can't seem to find any logical > reason why they couldn't coexist in the same query. > > b. > > > On Wed, Oct 9, 2019 at 1:08 PM Mikhail Khludnev <m...@apache.org> wrote: > > > Hello, Bram. > > > > I guess [child] was recently extended. Docs might be outdated, don't > > hesitate to contribute doc improvement. > > [subquery] is a neat thing, it's just queries without relying on > particular > > use case, if my understanding is right one may request something like > > _path_ field in [subquery], which may let to reconstruct hierarchy. > > > > On Wed, Oct 9, 2019 at 1:36 PM Bram Biesbrouck < > > bram.biesbro...@reinvention.be> wrote: > > > > > Hi all, > > > > > > I'm diving deep into the ChildDocTransformer and its > > > related SubQueryAugmenter. > > > > > > First of all, I think there's a bug in the Solr docs about [child]. It > > > states: > > > "This transformer returns all descendant documents of each parent > > document > > > matching your query in a flat list nested inside the matching parent > > > document." > > > This is not exact: the descendant documents are "wired into" the > parent, > > > creating a hierarchical structure (which is nice). Or am I > > misinterpreting > > > the docs? > > > > > > Secondly, the [subquery] transformer is super powerful and awesome, but > > it > > > doesn't like to be combined with [child]? I'm getting a "[subquery] > name > > > children is duplicated" error. Is there a way to work around this? Or > > maybe > > > better: is there a way to make the [subquery] transformer behave like > (a > > > more flexible version of) [child]? Because now, the path information > (how > > > the children relate to their parent fields) is lost when using > > [subquery]. > > > > > > Hope to hear more! > > > > > > b. > > > > > > > > > -- > > Sincerely yours > > Mikhail Khludnev > > > -- Sincerely yours Mikhail Khludnev
