Rule /require-data-selectors should not warn on cy.get(selector)

[ana-moraes]

Hello,

This issue was discussed before on #44, where judemorrissey reported the warnings being applied on aliases. This part seems to be working fine, but the situation described by SeanPercy and morficus is still ongoing.

For example:

// This works
cy.get('[data-test=a]').as('a'); 
cy.get('@a'); 

// This gives a warning
const ASSESSMENT_SUBMIT = '[data-cy=assessment-submit]';
cy.get(ASSESSMENT_SUBMIT).click();

This also gives a waning when using separate files for selectors

// selectors.js

const addPrefix = testId => `[data-cy=${testId}]`;
export const ASSESSMENT_SUBMIT = addPrefix('assessment-submit');

// test.js
cy.get(ASSESSMENT_SUBMIT).click();

Is it possible to change the rule to accommodate this type of situation?

Thank you.

[MikeMcC399]

@ana-moraes

• Thanks for highlighting that the additional comment in require-data-selectors rule should not apply to cy.get('@alias') #44 (comment) was not addressed by PR fix: Allow aliases in require-data-selectors rule #59 in eslint-plugin-cypress when it was released in May 2020.

This is about the rule cypress/require-data-selectors.

• The construct:

const ASSESSMENT_SUBMIT = '[data-cy=assessment-submit]'
cy.get(ASSESSMENT_SUBMIT)

generates the linting error:

error  use data-* attribute selectors instead of classes or tag names  cypress/require-data-selectors

It is not shown in the documentation docs/rules/require-data-selectors.md which lists only the following correct code examples:

cy.get('[data-cy=submit]').click()
cy.get('[data-QA=submit]')

I've marked this as an enhancement request. Most of the enhancement implementations in the last couple of years have come from the community. Perhaps there will be a community member who can contribute to implementing this?

[MikeMcC399]

@ana-moraes

• @duranbe has submitted PR fix: dont throw warning on variable data selector #272, so you might like to review this

[MikeMcC399]

• PR fix: dont throw warning on variable data selector #272 has been merged and a new version v5.1.1 has been released
• Although @duranbe noted that not all aspects of this request could be implemented, I am going to close it, as it significantly implements the request as far as possible