gerlowskija commented on PR #2079:
URL: https://github.com/apache/solr/pull/2079#issuecomment-1814728136

   > Our Admin UI handling of /select covers almost ever property so probably 
couldn't use the JS version.
   Yep definitely.  The "approximation" in this PR falls pretty far short of 
what the Admin UI exposes.
   
   If we cared to, we could close a lot of that gap with very little effort.  
Most parameters would be trivial to add.  But there's a few sticking points 
that are going to require a good bit more effort I think: in particular the 
arbitrary query-params that /select supports as placeholders/variables, and the 
JSON DSL.
   
   That's one of the reasons I didn't go beyond the absolute basics in this PR 
- I wanted the approximation to be driven by actual dogfood-able usage, and the 
Admin UI query screen felt out of reach for an initial pass.
   
   > However, for the V2 Query syntax, is that actually simpler from a API 
perspective since it's mostly "generate a lot of JSON and pass it to this end 
point"? Would that be easier to approximate?
   
   By the "V2 Query syntax" I assume you're talking about the JSON DSL?
   
   I guess it depends how we wanted to represent that in the OAS.  For the 
approximation, we could definitely have a POST that takes an arbitrary map in 
the request body (without trying to dictate exactly what properties/fields are 
in that map).  Eventually we're definitely going to want something more 
DSL-syntax aware...but for an approximation maybe treating the request-body 
syntax as a black box is a good-enough start?
   
   > Whether we can model each and every feature in the JSON DSL, such as query 
parsers, attributes, builders etc in a manner that lets us generate all those 
domain objects, which in turn converts the object graph to JSON, for each 
language?
   
   Yeah, it's a good question.  Do we even want our OAS (and consequently, the 
generated clients) to understand the syntax for each and every one of our 
QParsers, nested facets, etc?  Or maybe an "approximate" solution is the best 
one here in terms of usability for clients.  (Even setting usability aside, you 
could imagine an approximate solution might be more flexible for supporting 
user's custom server-side plugins, etc.)


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: issues-unsubscr...@solr.apache.org

For queries about this service, please contact Infrastructure at:
us...@infra.apache.org


---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscr...@solr.apache.org
For additional commands, e-mail: issues-h...@solr.apache.org

Reply via email to