Fork me on GitHub

Congrats on the CT grant :)


Will malli have something like spec/fdef + instrument + unstrument? And will the sequential destructuring be part of the first production release?


@borkdude the first release is planned to be just the current and (hopefully) immutable core on which everything builds on, Here’s the roadmap for the together funding:

1) get a stable release out! lot's of small and some bigger design decisions, tracked via metosin/malli#116

2) help early adopters (users and libraries like reitit, regal, aave and gungnir) to upgrade to use the initial version

After the release, would work on the following:

3) finalize sequence schemas, 

4) enhance developer tooling: 
  - function schemas with clj-kondo integration 
  - pull out and reuse the reitit development time error pretty printer as a separate library ()

5) implement pluggable schema inference

6) parsers


destucturing might be easy to implement using the current explain api, need to visit that for the first release to see will it require changes to the m/explain format.


Vincent has done a great initial work already with the sequence schemas, will dig into that soon.

bubblebobble 3
Vincent Cantin19:09:05

Thx ^_^ I recommend to take the impl from Minimallist as a reference, as it is the most up-to-date and (most importantly) well tested. The work remaining to be done is integration and optimization.


was just about to ask from which codebase should I look this for. Thanks!

Vincent Cantin10:09:12

the latest commit of the main branch all-work-and-no-play


so, no, but hopefully soon has both.


merged Adds :double,  `:boolean`,  `:keyword`,  `:symbol`,  `:qualified-keyword`, `:qualified-symbol` and `:uuid`. There is now m/-simple-schema to easily build custom schemas using properties in validation. Could rewrite m/-predicate-schema to use that too.