add snowId

coderbruis · coderbruis · commit 2e219748affb · 2021-08-05T01:25:20.000+08:00
diff --git a/JdkLearn/src/main/java/com/learnjava/concurent/SnowIdDemo.java b/JdkLearn/src/main/java/com/learnjava/concurent/SnowIdDemo.java
@@ -0,0 +1,12 @@
+package com.learnjava.concurent;
+
+/**
+ * @author lhy
+ * @date 2021/7/27
+ */
+public class SnowIdDemo {
+    public static void main(String[] args) {
+        // 通过雪花秀算法获取分布式id
+        System.out.println(SnowIdUtils.uniqueLongHex());
+    }
+}
diff --git a/JdkLearn/src/main/java/com/learnjava/concurent/SnowIdUtils.java b/JdkLearn/src/main/java/com/learnjava/concurent/SnowIdUtils.java
@@ -0,0 +1,114 @@
+package com.learnjava.concurent;
+
+/**
+ * @author lhy
+ * @date 2021/7/27
+ */
+public class SnowIdUtils {
+    /**
+     * 私有的 静态内部类
+     */
+    private static class SnowFlake {
+
+        /**
+         * 内部类对象（单例模式）
+         */
+        private static final SnowIdUtils.SnowFlake SNOW_FLAKE = new SnowIdUtils.SnowFlake();
+        /**
+         * 起始的时间戳
+         */
+        private final long START_TIMESTAMP = 1609464335121L;
+        /**
+         * 序列号占用位数
+         */
+        private final long SEQUENCE_BIT = 12;
+        /**
+         * 机器标识占用位数
+         */
+        private final long MACHINE_BIT = 10;
+        /**
+         * 时间戳位移位数
+         */
+        private final long TIMESTAMP_LEFT = SEQUENCE_BIT + MACHINE_BIT;
+        /**
+         * 最大序列号  （4095）
+         */
+        private final long MAX_SEQUENCE = ~(-1L << SEQUENCE_BIT);
+        /**
+         * 最大机器编号 （1023）
+         */
+        private final long MAX_MACHINE_ID = ~(-1L << MACHINE_BIT);
+        /**
+         * 生成id机器标识部分
+         */
+        private long machineIdPart;
+        /**
+         * 序列号
+         */
+        private long sequence = 0L;
+        /**
+         * 上一次时间戳
+         */
+        private long lastStamp = -1L;
+
+        /**
+         * 构造函数初始化机器编码
+         */
+        private SnowFlake() {
+//            String ip = instance.getDockerIp().replace(".", "");
+            // 模拟获取机器节点ip
+            String ip = "127.0.0.1";
+            long localIp = Long.valueOf(ip.replace(".", ""));
+            machineIdPart = (localIp & MAX_MACHINE_ID) << SEQUENCE_BIT;
+        }
+        /**
+         * 获取雪花ID
+         */
+        public synchronized long nextId() {
+            long currentStamp = timeGen();
+            while (currentStamp < lastStamp) {
+                throw new RuntimeException(String.format("时钟已经回拨.  Refusing to generate id for %d milliseconds", lastStamp - currentStamp));
+            }
+            if (currentStamp == lastStamp) {
+                sequence = (sequence + 1) & MAX_SEQUENCE;
+                if (sequence == 0) {
+                    currentStamp = getNextMill();
+                }
+            } else {
+                sequence = 0L;
+            }
+            lastStamp = currentStamp;
+            return (currentStamp - START_TIMESTAMP) << TIMESTAMP_LEFT | machineIdPart | sequence;
+        }
+        /**
+         * 阻塞到下一个毫秒，直到获得新的时间戳
+         */
+        private long getNextMill() {
+            long mill = timeGen();
+            //
+            while (mill <= lastStamp) {
+                mill = timeGen();
+            }
+            return mill;
+        }
+        /**
+         * 返回以毫秒为单位的当前时间
+         */
+        protected long timeGen() {
+            return System.currentTimeMillis();
+        }
+    }
+
+    /**
+     * 获取long类型雪花ID
+     */
+    public static long uniqueLong() {
+        return SnowIdUtils.SnowFlake.SNOW_FLAKE.nextId();
+    }
+    /**
+     * 获取String类型雪花ID
+     */
+    public static String uniqueLongHex() {
+        return String.format("%016X", uniqueLong());
+    }
+}
diff --git a/JdkLearn/src/main/java/com/learnjava/optimization/OptimizeDemo.java b/JdkLearn/src/main/java/com/learnjava/optimization/OptimizeDemo.java
@@ -2,6 +2,7 @@
 
 import java.util.HashMap;
 import java.util.Map;
+import java.util.concurrent.ConcurrentHashMap;
 
 /**
  *
@@ -14,6 +15,9 @@ public class OptimizeDemo {
     public static void main(String[] args) {
         Map<String, Integer> map = new HashMap<>();
         mergeData(map);
+
+        Map<String, Integer> concurrentMap = new ConcurrentHashMap<>();
+        concurrentMergeData(concurrentMap);
     }
 
     /**
@@ -34,7 +38,33 @@ public static void mergeData(Map<String, Integer> map) {
         Integer mapValue = map.get(key);
         if (null != mapValue) {
             mapValue += value;
+        } else {
+            mapValue = value;
         }
         map.put(key, mapValue);
     }
+
+    /**
+     * 针对mergeData里map的put操作，在并发情况下会存在put的时候，以及有其他线程已经put成功了，导致线程不安全，
+     * 所以需要使用并发集合列的putIfAbsent方法
+     * @param map
+     */
+    public static void concurrentMergeData(Map<String, Integer> map) {
+        String key = "mapKey";
+        int value = 1;
+        Integer mapValue = map.get(key);
+        if (null != mapValue) {
+            mapValue += value;
+        } else {
+            mapValue = value;
+        }
+        map.putIfAbsent(key, mapValue);
+
+        // computeIfAbsent方法对map中的key只进行重新计算，如果不存在这个key，则添加到map中
+        map.computeIfAbsent(key, (k) -> {
+            // 其他计算
+            int a = 1, b = 2;
+            return a + b;
+        });
+    }
 }
