test

If you use aquery_differ to diff 2 aquery proto files and you pass both --attrs=inputs and --attrs=cmdline, if there is a difference in the Environment keys of actions it will not appear in the final diff

Currently, when entering epic mode the README is frozen in the last level of the tower. When you're trying to fine-tune the score for a level other than the last one, it would be helpful if we had the README for that level available. The proposal is that when entering epic mode, the README is updated with all levels, one following the other.

Example:

# Starbolt - beginner

uiautomator2的版本号：2.11.1

使用device.image.match时匹配错误

模版图像是：

截图：

匹配结果：{'similarity': 0.9961444735527039, 'point': [319, 673]}

直接使用findit模块测试也是这么高匹配度，估计是findit模块不给力，是

I'd like an option for shouldEqualJson to ignore keys in actual that are unspecified in expected. It should probably be included for shouldNotEqualJson as well.

I think this could be achieved either by:

• Expanding CompareMode with new modes: CompareMode.StrictIgnoringUnknownKeys and CompareMode.LenientIgnoringUnknownKeys
• Adding a new flag to matcher signatures: `ignoreUnknown

A "normal" event on an instance ignores null subscriptions. If I do so on a substituted interface, a null subscription leads to an exception. I have a not so nice legacy code with different listeners to an event an one class that does null-attach internally under some conditions, would be nice if this behaviour would be able to be tested as well. Example code for the problem:

        publ