module Mongoid::Matcher::EqImpl

def matches?(exists, value, condition, original_operator) 

def matches?(exists, value, condition, original_operator)

invokes $eq, the exception message needs to handle
tors.
::InvalidQuery, "Range is not supported as an argument to '#{original_operator}'"
a?(Array)
 { |elt| condition.include?(elt) }
include?(value)
 a comparison with Time objects, compare using millisecond precision
ompare_time_by_ms
ind_of?(Time) && condition.kind_of?(Time)
(value, condition)
e.is_a?(Array) && condition.kind_of?(Time)
p do |v|
ind_of?(Time)
_rounded_to_millis(v)
ude?(time_rounded_to_millis(condition))
 condition ||
_a?(Array) && value.include?(condition)
ondition ||
?(Array) && value.include?(condition)

def time_eq?(time_a, time_b)

As such, perform a similar operation to what the bson-ruby gem does

unexpected results.
calculations using integer math, as inexactness of floating point calculations may produce
> Because of this flooring, applications are strongly recommended to perform all time

are serialized to BSON or Extended JSON, the times are floored to the nearest millisecond.
> Times in BSON (and MongoDB) can only have millisecond precision. When Ruby Time instances
Per https://www.mongodb.com/docs/ruby-driver/current/tutorials/bson-v4/#time-instances, 
def time_eq?(time_a, time_b)
o_millis(time_a) == time_rounded_to_millis(time_b)

def time_rounded_to_millis(time) 

def time_rounded_to_millis(time)
son_to_i * 1000 + time.usec.divmod(1000).first