Hi, Nested is a big nononno for modelling collections. It's either query time join or field collapsing/grouping or just an aggregation or so.
On Wed, May 21, 2025 at 3:56 PM Yaşar Arabacı <yasar11...@gmail.com> wrote: > Hi, > > I am trying to figure out a schema for indexing documents that I will > download from another document management system. In the original > source, documents are stored inside collections. Collections act like > folders with extra metadata (they cannot be nested though). > > Each document must be part of a collection, so there are no dangling > documents. A document can rarely be in multiple collections, I am > thinking of indexing them for each collection they are in to keep it > simple. > > My two primary concerns are that I should be able to modify collection > metadata and I should be able to add more documents without reindexing > all the documents in the collection. > > I will only be searching for text and metadata (like creation time, > author name etc.) of individual documents but collection metadata > should also be returned with results. > > My second plan is to index documents and collections separately and > join them in query time (documentation seems to indicate query time > joins are possible). > > I am very new to solr, so I don't want to start with a schema that I > will regret later. So any advice is appreciated. > > Best Regards, > Yaşar > -- Sincerely yours Mikhail Khludnev
