fix: schema-based validation not working

[seriousManual]

We have been trying to use the schema based validation as described in the following sample:
https://github.com/typestack/class-validator/tree/develop/sample/sample5-schemas

We were not able to get the schema based validation to work and even executing the code from the provided sample did not work.

This looks like there has been a regression somewhere in the past?

[pasear]

Same here.
Here's a simple script to reproduce it.

var classValidator = require("class-validator");

var classValidatorSchemas = [
  { name: "uuid", properties: { value: [{ type: "isUuid", constraints: [4] }] } },
  { name: "len", properties: { value: [{ type: "maxLength", constraints: [1] }] } },
  { name: "mail", properties: { value: [{ type: "isEmail" }] } }
];

classValidatorSchemas.forEach((s) => classValidator.registerSchema(s));

console.log("uuid", classValidator.validateSync("uuid", {value: "aaa"}));
console.log("len", classValidator.validateSync("len", {value: "aaa"}));
console.log("mail", classValidator.validateSync("mail", {value: "aaa"}));

[d-otter]

The schema-based validation stopped working with version 0.12.0 - same issue, unfortunately I do not know the internals good enough to spot the root cause

[jbeckton]

Looks like Schema Based validation will be deprecated in 1.0 #178.

Bummer

[braaar]

If you want schema based validation, check out the zod readme's comparison with other validators

[pasear]

I know alternatives, but this feature is in the document and should be a bug. 2022年10月6日(木) 13:52 Brage Sekse Aarset ***@***.***>:

…

IF you need schema based validation, check out the zod readme's comparison with other validators <https://github.com/colinhacks/zod#comparison> — Reply to this email directly, view it on GitHub <#1155 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AB624UGEWEJZLBVTBAHZG2TWBZLKRANCNFSM47SEWUOA> . You are receiving this because you commented.Message ID: ***@***.***>

[braaar]

It is currently a bug, but as this feature is going to be deprecated, it's not wise for this project to prioritise scarce maintainer time and effort towards fixing it. After almost a year of inactivity, we have gotten a little bit of maintainer involvement, and I think we will have to accept that we can't get everything we want, at least for the time being. See #1719

PR is welcome, though. The first step to fixing this would be to write some tests for the schema based validation.

[pasear]

How about removing it from the document? It's still on the homepage (readme) and confusing peoole. It's already broken and no need to keep it in the document. 2022年10月6日(木) 14:10 Brage Sekse Aarset ***@***.***>:

…

It is currently a bug, but as this feature is going to be deprecated, it's not wise for this project to prioritise scarce maintainer time and effort towards fixing it. After almost a year of inactivity, we have gotten a little bit of maintainer involvement, and I think we will have to accept that we can't get everything we want, at least for the time being. See #1719 <#1719> PR is welcome, though. The first step to fixing this would be to write some tests for the schema based validation. — Reply to this email directly, view it on GitHub <#1155 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AB624UD5LGXGBKDUUWTGCZTWBZNNNANCNFSM47SEWUOA> . You are receiving this because you commented.Message ID: ***@***.***>

[braaar]

How about removing it from the document? It's still on the homepage (readme) and confusing peoole. It's already broken and no need to keep it in the document.

That is a good compromise solution. I can make a PR.