Tuesday, January 1, 2019

359. Logger Rate Limiter

359Logger Rate Limiter
Design a logger system that receive stream of messages along with its timestamps, each message should be printed if and only if it is not printed in the last 10 seconds.
Given a message and a timestamp (in seconds granularity), return true if the message should be printed in the given timestamp, otherwise returns false.
It is possible that several messages arrive roughly at the same time.
Example:
Logger logger = new Logger();

// logging string "foo" at timestamp 1
logger.shouldPrintMessage(1, "foo"); returns true; 

// logging string "bar" at timestamp 2
logger.shouldPrintMessage(2,"bar"); returns true;

// logging string "foo" at timestamp 3
logger.shouldPrintMessage(3,"foo"); returns false;

// logging string "bar" at timestamp 8
logger.shouldPrintMessage(8,"bar"); returns false;

// logging string "foo" at timestamp 10
logger.shouldPrintMessage(10,"foo"); returns false;

// logging string "foo" at timestamp 11
logger.shouldPrintMessage(11,"foo"); returns true;
-----------------
有好多种方法,可以跟面试官讨论。
用priorityQueue也可以
对比的复杂度O(log (k * n)) vs O(k * n), k为总时间(ex. 10s), n为每秒的消息数量
class Logger {
    
    private Map<String, Integer> map;
    /** Initialize your data structure here. */
    public Logger() {
        map = new HashMap<>();
    }
    
    /** Returns true if the message should be printed in the given timestamp, otherwise returns false.
        If this method returns false, the message will not be printed.
        The timestamp is in seconds granularity. */
    public boolean shouldPrintMessage(int timestamp, String message) {
        prune(timestamp);
        if (!map.containsKey(message)) {
            map.put(message, timestamp);
            return true;
        }
        
        return false;
    }
    
    private void prune(int timestap) {
        Iterator<String> itr = map.keySet().iterator();
        while (itr.hasNext()) {
            String k = itr.next();
            if (map.get(k) + 10 <= timestap) {
                itr.remove();
            }
        }
    }
}

/**
 * Your Logger object will be instantiated and called as such:
 * Logger obj = new Logger();
 * boolean param_1 = obj.shouldPrintMessage(timestamp,message);
 */

另一种方法是按时间存,timestamp as a bucket。 时间超了整个bucket清空掉。bucket里面可以用Set<>做到O(1)查找

No comments:

Post a Comment