#362 - Extend prefer-web-first-assertion

[MSroczynski3]

In this PR, I extended the src/rules/prefer-web-first-assertions.ts rule to detect another two flaky methods:

-   allInnerTexts()
-   allTextContents().

Additionally, I added handling for options argument that is useful, for example, when replacing innerText() with toHaveText() with useInnerText option. Tests for all changes were added.

This is a follow-up to #362 .

[mskelton]

This isn't the right approach IMO. This isn't a one-to-one replacement, since the behavior of allTextContents is different from toHaveText.

I think we should add a lint error, but not autofix.

[MSroczynski3]

@mskelton I see your point here. My thought process was connected with the fact that toHaveText() is designed to handle multiple elements and multiple text values well. That being said this rule would get lost with this problem:

        const myText = page.locator('foo li').allTextContents();
        expect(myText).toEqual(['foo', 'bar', 'baz'])`

I tried a simplified approach. Could you please let me know your thoughts?

[mskelton]

We can apply an autofix if the autofix is a 1x1 match of behavior, so the example you gave would work.

[MSroczynski3]

I can give it a try, but taking into account how much logic there is in the rule already, I am not sure if it won't complicate things too much. Would this merit a separate rule? Probably unlikely...

[MSroczynski3]

@mskelton I spent some time thinking about this. I think that this rule is in a good place. Adjusting it further will bloat the function for relatively little added value. Do you think we can proceed and merge this PR?