When the first parameter to inreplace was an array, and the replacement
failed, InreplaceError would end up crashing with an undefined method
exception because the order of operations resulted in super not being
passed the value of the entire inject block.

`any?` is not the opposite of `empty?`. Besides the case that
`[false, nil].any?` will return false, `any?`(O(n)) has much worse
performance than `empty?`(O(1)).

Provides feature parity between the block and non-block forms of
inreplace by creating a four-argument version of the non-block form,
where the fourth argument is an optional Boolean value, defaulting to
true, which specifies whether a failed inreplace should cause an
InreplaceError error to be raised. The fourth argument is passed along
to StringInreplaceExtension#gsub!, which already supports an optional
audit_result argument.

This resolves the Catch-22 that single replacements aren't permissible
in the block form (in that they now cause `brew audit` to complain), but
the audit_result argument is not available in the non-block form.

Closes #552.

Signed-off-by: ilovezfs <ilovezfs@icloud.com>

And remove the documented stuff from the `example-formula.rb`.

Closes Homebrew/homebrew#43241.

Signed-off-by: Mike McQuaid <mike@mikemcquaid.com>

Closes Homebrew/homebrew#42354.

Signed-off-by: Mike McQuaid <mike@mikemcquaid.com>

Closes Homebrew/homebrew#32649.
Closes Homebrew/homebrew#32703.

Closes Homebrew/homebrew#22628.

Signed-off-by: Jack Nagel <jacknagel@gmail.com>

As discussed in https://github.com/mxcl/homebrew/pull/21936



Closes Homebrew/homebrew#21942.

Signed-off-by: Adam Vandenberg <flangy@gmail.com>

Closes Homebrew/homebrew#21163.