refactor(cubesql): Improve EGraph debugger, part 3: parsing and data preparation in UI side #8934
Conversation
Codecov ReportAttention: Patch coverage is
Additional details and impacted files@@ Coverage Diff @@
## master #8934 +/- ##
==========================================
+ Coverage 82.66% 82.72% +0.06%
==========================================
Files 221 221
Lines 78303 78244 -59
==========================================
+ Hits 64726 64727 +1
+ Misses 13577 13517 -60
Flags with carried forward coverage won't be shown. Click here to find out more. ☔ View full report in Codecov by Sentry. |
| @@ -18,6 +18,7 @@ import type { | |||
| NodeProps, | |||
| } from 'reactflow'; | |||
| import 'reactflow/dist/style.css'; | |||
| import { z } from 'zod'; | |||
There was a problem hiding this comment.
Probably, not a big deal but still:
We already have a joi scheme validator in our codebase, so maybe we can reuse it here too? I know, node_modules is a black hole and could hold any number of packages) but... If we can reduce and reuse...Why not?
There was a problem hiding this comment.
I'd rather switch to zod in those places 😅
Joi does not introduce static types in TypeScript along with building parser, so one need to manually support and ensure compatibility for both. There are solutions like joi-to-typescript, but, given the usual process of working with codegen in JS, it's not very comfortable.
Also egraph debugger is a completely separate package from monorepo, so it wouldn't pollute anything with unnecessary dependencies in "usual" case, and it would install them second time even if dependencies were same.
mcheshkov
force-pushed
the
refactor-egraph-debugger-3
branch
from
676b716
to
5212ee8
Compare
|
The latest updates on your projects. Learn more about Vercel for Git ↗︎ 8 Skipped Deployments
|
Used just for Id for now
mcheshkov
force-pushed
the
refactor-egraph-debugger-3
branch
from
5212ee8
to
05e20cb
Compare
Check List
Description of Changes Made (if issue reference is not provided)
Add proper parsing on UI side, to catch data format changes.
Move most of data preparation related to UI (like a concepts of edge or combo) to UI completely, simplify data collection on rewriter side.