The committers listed above are authorized under a signed CLA.

• ✅ login: VladimirPodolyan (4b46f6a, a07da53)

Each call to self.context creates extra HTTP request to Appium server, which could significantly slow down automated tests. Unfortunately we cannot merge this PR

This one should work accurately when autoWebview is enabled, without explicit set context, for example.

@mykola-mokhnach Hi, and thank you for answer.
Maybe it's possible to pass that as class argument for both classes?

For WebDriver:

class WebDriver(...):
    def __init__(command_executor, desired_capabilities, ..., selenium_compatibility)
        ...
        self.selenium_compatibility = selenium_compatibility

    ...

    def find_element():
            if self.selenium_compatibility:
                if by == By.ID:
                    ...

    ...

For WebElement (need to define __init__)

class WebElement(AppiumWebElementSearchContext):
    def __init__(self, selenium_compatibility=False):
        self.selenium_compatibility = selenium_compatibility

    ...
    
    def find_element():
            if self.selenium_compatibility:
                if by == By.ID:
                    ...

    ...

Or only for WebDriver, even you don't accept the specifying the __init__ in WebElement

This has nothing to do with selenium compatibility. Selenium itself has this code as a workaround to support legacy non-w3c compatible locators (the standard only defines css and xpath locator types).

To be honest I'm not very happy about keeping these workarounds here as locators id/name locator types are not going to be supported anyway, which encourages customers to update their tests instead of having the legacy stuff forever.

In the worst case it is possible to do the same id/name -> css conversion directly in the client code with the same success.

@mykola-mokhnach Understood. But what is the final decision here? You don't want to have this "Selenium workaround" in your own code, but it's commented right now. Can I just delete this "comment section" in this PR and do a self workaround in my side?

The comment-out (TODO mark) was still we were learning selenium in the past.

The TODO comment can be removed, but I need the comment-out ones to keep the diff about the workaround for future updates so far, so please keep the comment-out itself so far.

The committers listed above are authorized under a signed CLA.

• ✅ login: VladimirPodolyan (4b46f6a, a07da53)