Will error message be overwritten by concurrent validations?

[ivawzh]

Hi, I was reading the README, the following line interests me:

Please note: every time validation function or ajv.validate are called errors property is overwritten. You need to copy errors array reference to another variable if you want to use it later (e.g., in the callback). See Validation errors

For example,

const validate = ajv.compile(customerSchema)

function checkCustomer(id) {
  return promiseFetch(id).then(customer => {
    if (!validate(customer)) {
      console.log(`Customer ${id} has errors: ${validate.errors}`)
    }
  })
}

[id1, id2, id3, ...].forEach(id => {
  checkCustomer(id)
})

Am I right that this code has a risk of having a wrong id-error mapping in Customer ${id} has errors: ${validate.errors} output?

Do I have to create and compile a new ajv insance inside each promise to ensure that my error messages won't be overwritten?

[epoberezkin]

JavaScript is single threaded and the validation itself in your case is synchronous for each ID.
So there is no such risk.
If, for example, you wan't to collect all errors from all customers you can:

const validate = ajv.compile(customerSchema)

function checkCustomer(id) {
  return promiseFetch(id).then(customer => {
    validate(customer);
    return { id: id, errors: validate.errors };
  })
}

Promise.all([id1, id2, id3, ...].map(id => checkCustomer(id) ))
.then(results => console.log(results));

It will be mapped correctly.

If the schema itself is asynchronous then errors are not returned as the property of validation function but as the property of error object that the promise rejects with.

[epoberezkin]

In relation to the comment in readme, it means that if you tried to use the same errors in a different execution block they could be errors from the different validation.

For example this code has this problem:

promiseFetch(id)
.then(customer => ({ id: customer.id, valid: validate(customer) }))
.then(result => {
    if (!result.valid) console.log(`Customer ${id} has errors: ${validate.errors}`);
});

But the code below will always work as expected, because errors are used in the same execution block before they can be overwritten:

promiseFetch(id)
.then(customer => {
  var valid = validate(customer);
  return { id: customer.id, valid: valid, errors: validate.errors };
})
.then(result => {
    if (!result.valid) console.log(`Customer ${id} has errors: ${result.errors}`);
});

[ivawzh]

Very well explained. Thank you!