feat: named constraints + improvements #919
Conversation
|
@do4gr Could you check if I missed something here? |
| @@ -179,7 +227,8 @@ | |||
| ] | |||
| }, | |||
| { | |||
| "label": "@default()", | |||
| "label": "@default", | |||
There was a problem hiding this comment.
on SQL Server defaults are also named. So there they will have a map argument. I'm not sure we currently have the capability to give connector aware autocomplete?
There was a problem hiding this comment.
If we have that, we also should adjust the map argument on primary keys, since on mysql and sqlite these are not named. Same goes for foreign keys on sqlite, we use no names there.
There was a problem hiding this comment.
Indeed, we could build that I think using some regex then.
| { line: 34, character: 13 }, | ||
| { | ||
| isIncomplete: false, | ||
| items: [fieldsProperty, nameProperty, mapProperty], |
There was a problem hiding this comment.
Do I understand it correctly that the items are the generated suggestions?
| { line: 40, character: 33 }, | ||
| { | ||
| isIncomplete: false, | ||
| items: [fieldsProperty, nameProperty, mapProperty], |
There was a problem hiding this comment.
Why are then fields still suggested if already present?
There was a problem hiding this comment.
In the case of @@unique() it's expected to have all suggestions
In the case of @@unique([email, something], ) it is indeed not expected to have all suggestions, fields should be removed though this looks like something we should do in another PR.
There was a problem hiding this comment.
Note that @@unique(fields: [email, something], ) would then work as expected (not suggest fields)
Closes #825
The goal of this PR is to add auto-completion for named index / constraints added in 2.29.
I also used the available feature
insertTextwhich makes it possible to display something but insert a different version which we insert in snippet mode and$0means that the cursor will move to that position for better auto-completion experience.