diff --git a/docs/java/collection/concurrent-hash-map-source-code.md b/docs/java/collection/concurrent-hash-map-source-code.md index 74e5faac..015077ca 100644 --- a/docs/java/collection/concurrent-hash-map-source-code.md +++ b/docs/java/collection/concurrent-hash-map-source-code.md @@ -16,7 +16,7 @@ tag: > 下图存在两个笔误 : Segmeng -> Segment ; HashEntity -> HashEntry -![Java 7 ConcurrentHashMap 存储结构](./images/image-20200405151029416.png) +![Java 7 ConcurrentHashMap 存储结构](./images/java7_concurrenthashmap.png) Java 7 中 `ConcurrentHashMap` 的存储结构如上图，`ConcurrnetHashMap` 由很多个 `Segment` 组合，而每一个 `Segment` 是一个类似于 HashMap 的结构，所以每一个 `HashMap` 的内部可以进行扩容。但是 `Segment` 的个数一旦**初始化就不能改变**，默认 `Segment` 的个数是 16 个，你也可以认为 `ConcurrentHashMap` 默认支持最多 16 个线程并发。 diff --git a/docs/java/collection/images/ad28e3ba-e419-4724-869c-73879e604da1.png b/docs/java/collection/images/ad28e3ba-e419-4724-869c-73879e604da1.png deleted file mode 100755 index 8d802d64..00000000 Binary files a/docs/java/collection/images/ad28e3ba-e419-4724-869c-73879e604da1.png and /dev/null differ diff --git a/docs/java/collection/images/java-collection-hierarchy.png b/docs/java/collection/images/java-collection-hierarchy.png index 6cc15baa..346b0203 100755 Binary files a/docs/java/collection/images/java-collection-hierarchy.png and b/docs/java/collection/images/java-collection-hierarchy.png differ diff --git a/docs/java/collection/images/image-20200405151029416.png b/docs/java/collection/images/java7_concurrenthashmap.png similarity index 100% rename from docs/java/collection/images/image-20200405151029416.png rename to docs/java/collection/images/java7_concurrenthashmap.png diff --git a/docs/java/collection/images/jdk1.8之前的内部结构-HashMap.png b/docs/java/collection/images/jdk1.7_hashmap.png similarity index 100% rename from docs/java/collection/images/jdk1.8之前的内部结构-HashMap.png rename to docs/java/collection/images/jdk1.7_hashmap.png diff --git a/docs/java/collection/images/jdk1.8之后的内部结构-HashMap.png b/docs/java/collection/images/jdk1.8_hashmap.png similarity index 100% rename from docs/java/collection/images/jdk1.8之后的内部结构-HashMap.png rename to docs/java/collection/images/jdk1.8_hashmap.png diff --git a/docs/java/collection/images/TreeMap继承结构.png b/docs/java/collection/images/treemap_hierarchy.png similarity index 100% rename from docs/java/collection/images/TreeMap继承结构.png rename to docs/java/collection/images/treemap_hierarchy.png diff --git a/docs/java/collection/java-collection-questions-02.md b/docs/java/collection/java-collection-questions-02.md index 46b08903..5508d641 100644 --- a/docs/java/collection/java-collection-questions-02.md +++ b/docs/java/collection/java-collection-questions-02.md @@ -67,7 +67,7 @@ tag: `TreeMap` 和`HashMap` 都继承自`AbstractMap` ，但是需要注意的是`TreeMap`它还实现了`NavigableMap`接口和`SortedMap` 接口。 -![](./images/TreeMap继承结构.png) +![](./images/treemap_hierarchy.png) 实现 `NavigableMap` 接口让 `TreeMap` 有了对集合内元素的搜索的能力。 @@ -213,13 +213,13 @@ static int hash(int h) { 所谓 **“拉链法”** 就是：将链表和数组相结合。也就是说创建一个链表数组，数组中每一格就是一个链表。若遇到哈希冲突，则将冲突的值加到链表中即可。 -![jdk1.8之前的内部结构-HashMap](./images/jdk1.8之前的内部结构-HashMap.png) +![jdk1.8 之前的内部结构-HashMap](./images/jdk1.7_hashmap.png) #### JDK1.8 之后 相比于之前的版本， JDK1.8 之后在解决哈希冲突时有了较大的变化，当链表长度大于阈值（默认为 8）（将链表转换成红黑树前会判断，如果当前数组的长度小于 64，那么会选择先进行数组扩容，而不是转换为红黑树）时，将链表转化为红黑树，以减少搜索时间。 -![jdk1.8之后的内部结构-HashMap](./images/jdk1.8之后的内部结构-HashMap.png) +![jdk1.8之后的内部结构-HashMap](./images/jdk1.8_hashmap.png) > TreeMap、TreeSet 以及 JDK1.8 之后的 HashMap 底层都用到了红黑树。红黑树就是为了解决二叉查找树的缺陷，因为二叉查找树在某些情况下会退化成一个线性结构。 @@ -258,9 +258,7 @@ static int hash(int h) { **JDK1.7 的 ConcurrentHashMap：** -![JDK1.7的ConcurrentHashMap](https://my-blog-to-use.oss-cn-beijing.aliyuncs.com/2019-6/ConcurrentHashMap分段锁.jpg) - -

https://www.cnblogs.com/chengxiao/p/6842045.html>

+![Java 7 ConcurrentHashMap 存储结构](./images/java7_concurrenthashmap.png) **JDK1.8 的 ConcurrentHashMap：**