Fixes ui

[zorox112-practicum]

replace strong to red *
add possible reasons
add ClipboardButton response body

[zorox112-practicum]

allOf use for find ref

[moki]

i would prefer styles being factored out into css file and span having class name applied here

[zorox112-practicum]

fixed:
e9461d1

[moki]

this whole function might as well be:

if (value.enum) {
  return 'string';
}

return value.type

or

return value.enum ?? value.type;

[zorox112-practicum]

enum has the type "JSONSchema6Type[] | undefined" and it has possible values for enum, not their type. The purpose of the function is to set the string value "string" if the type was not specified in the specification

 export type JSONSchema6Type =
    | string 
    | number
    | boolean
    | JSONSchema6Object
    | JSONSchema6Array
    | null;

[zorox112-practicum]

This function should output the value of the "type" field if it is set explicitly, substitute "type: 'string'" if it is an enum and return the original type if it is not an enum, I can write in the last return like this:
return undefined

It seems that in the current version it is clearer that the function will return the original type than hardcoding undefined

[moki]

This function should output the value of the "type" field if it is set explicitly, substitute "type: 'string'" if it is an enum and return the original type if it is not an enum, I can write in the last return like this: return undefined

It seems that in the current version it is clearer that the function will return the original type than hardcoding undefined

is there the case of:
type being set explicitly(to value other than 'string') and enum being provided for one entity?

[zorox112-practicum]

type being set explicitly(to value other than 'string')

yes

enum being provided for one entity

I don't understand this part of the question

[moki]

type being set explicitly(to value other than 'string')

yes

enum being provided for one entity

I don't understand this part of the question

do we have a case when type is being set to value other that string and enum is also provided at the same time

[zorox112-practicum]

do we have a case when type is being set to value other that string and enum is also provided at the same time

yes.
example#1

        name:
          type: integer
          enum:
            - 1
            - 2

Generated type:

export const PetNameEnum = {
    NUMBER_1: 1,
    NUMBER_2: 2
} as const;

example#2:

        name:
          type: string
          enum:
            - 1
            - 2

Generated type:

export const PetNameEnum = {
    _1: '1',
    _2: '2'
} as const;

In the OPEN API specification, enum is always an array.
https://github.com/OAI/OpenAPI-Specification/blob/main/schemas/v3.0/schema.json#L404

I fixed the type and made the logic, if type was not passed, then type is taken from enum[0]

[moki]

@zorox112-practicum i can approve ui changes as they are rn, and we will move on with this PR.

if you want to finish implementation of json6schema normalization, cherry pick commits that are about it onto another separate branch -> make PR then we will continue discussion there.

don't forget to cleanup commit history tho.

[zorox112-practicum]

divided into a pool of requests:
#301
#302

here only UI

[zorox112-practicum]

@moki ^

[zorox112-practicum]

@moki
this is the final version, sorry ^

[moki]

@zorox112-practicum

cleanup commit history:

rebase squash:

f0ec15e into dc051c2
c78258e into 88e2794

give meaningful name to each commit to have clear changelogs

describe functionality that you fix/add concisely in header and optionally go in depth inside the commit body.

commit messages:

f7bde27 i see that you set data attribute for the link element why tho i have no idea, describe what problem you are fixing with it

dc051c2 same goes for the replacing strong with red

6adb8e0 better commit message would be display request failure debug information

[moki]

i've missed the point we started declaring constants in a camelCase

you wouldn't have to do as TextEnum if usual convention of having constants in uppercase SNAKE_CASE were followed.

about naming FETCH_FAILURES would be clear enough to describe constant values and more concise.