Redis源码分析（七）---zipmap压缩图

如果有看过之前我分析的ziplist压缩列表的分析的话，理解这个我觉得不是什么特别的难题。ziplist压缩列表和zipmap都采用了动态分配字节的做法表示长度，比如通过固定的字节表示节省了不少的空间。同样带来的问题就是复杂的指针移动，和字符位置移动。但总的

如果有看过之前我分析的ziplist压缩列表的分析的话，理解这个我觉得不是什么特别的难题。ziplist压缩列表和zipmap都采用了动态分配字节的做法表示长度，比如通过固定的字节表示节省了不少的空间。同样带来的问题就是复杂的指针移动，和字符位置移动。但总的来说，一定是利大于弊了，要不然设计者也不会这么做。ziplist保存的使用一个列表，zipmap就保存的则是一个个键值对，通过key:value key:value的形式连着。下面我给出zipmap的结构构成，zipmap其实也就是一个超级长的字符串。

"foo""bar""hello""world" 

里面涉及了几个变量zmlen，len,free，下面给出完整的解释：

/* String -> String Map data structure optimized for size.
 * This file implements a data structure mapping strings to other strings
 * implementing an O(n) lookup data structure designed to be very memory
 * efficient.
 *
 * The Redis Hash type uses this data structure for hashes composed of a small
 * number of elements, to switch to a hash table once a given number of
 * elements is reached.
 *
 * Given that many times Redis Hashes are used to represent objects composed
 * of few fields, this is a very big win in terms of used memory.
 *
 * zipmap压缩表和ziplist十分类似，都做到了内存操作效率比较高的
 * --------------------------------------------------------------------------
 *
 * Copyright (c) 2009-2010, Salvatore Sanfilippo 
 * All rights reserved.
 *
 * Redistribution and use in source and binary forms, with or without
 * modification, are permitted provided that the following conditions are met:
 *
 *   * Redistributions of source code must retain the above copyright notice,
 *     this list of conditions and the following disclaimer.
 *   * Redistributions in binary form must reproduce the above copyright
 *     notice, this list of conditions and the following disclaimer in the
 *     documentation and/or other materials provided with the distribution.
 *   * Neither the name of Redis nor the names of its contributors may be used
 *     to endorse or promote products derived from this software without
 *     specific prior written permission.
 *
 * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
 * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
 * ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE
 * LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
 * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
 * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
 * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
 * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
 * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
 * POSSIBILITY OF SUCH DAMAGE.
 */

/* Memory layout of a zipmap, for the map "foo" => "bar", "hello" => "world":
 *
 * "foo""bar""hello""world"
 *
 *  is 1 byte length that holds the current size of the zipmap.
 * When the zipmap length is greater than or equal to 254, this value
 * is not used and the zipmap needs to be traversed to find out the length.
 * 占有着1个字节，所以他的最多可代表的数量是254,当zipmap中的元素记录超过这个数时，
 * 那只能从前往后后遍历算大小了，和ziplist是不一样的。
 *
 *  is the length of the following string (key or value).
 *  lengths are encoded in a single value or in a 5 bytes value.
 * If the first byte value (as an unsigned 8 bit value) is between 0 and
 * 252, it's a single-byte length. If it is 253 then a four bytes unsigned
 * integer follows (in the host byte ordering). A value of 255 is used to
 * signal the end of the hash. The special value 254 is used to mark
 * empty space that can be used to add new key/value pairs.
 * 代表了后面字符串key 或 value的值的长度，长度一般被编码1个字节或5个字节表示，这个和ziplist类似
 * 如果后面的字符串长度小于等于252个，可与用单字节表示，其他253,254等长度被用来表示其他作用了，当超过这个数时候
 * 则直接按5字节的方式存储长度。
 *
 *  is the number of free unused bytes after the string, resulting
 * from modification of values associated to a key. For instance if "foo"
 * is set to "bar", and later "foo" will be set to "hi", it will have a
 * free byte to use if the value will enlarge again later, or even in
 * order to add a key/value pair if it fits.
 * 一般来表示后面的value长度的空闲值，当key：value=“foo”:"bar",后来被改为“foo”:"hi"，空闲长度就为1了
 *
 *  is always an unsigned 8 bit number, because if after an
 * update operation there are more than a few free bytes, the zipmap will be
 * reallocated to make sure it is as small as possible.
 * 的数字一般比较小，如果空闲太大，zipmap会进行调整大小使map整体变得尽可能小
 *
 * The most compact representation of the above two elements hash is actually:
 * 这是一个例子：
 * "foo""bar""hello""world" 
 * key字符后面同前
 * "\x02\x03foo\x03\x00bar\x05hello\x05\x00world\xff"
 *
 * Note that because keys and values are prefixed length "objects",
 * the lookup will take O(N) where N is the number of elements
 * in the zipmap and *not* the number of bytes needed to represent the zipmap.
 * This lowers the constant times considerably.
 */

说到键值对，里面最最重要的方法当然是根据key ，setValue的方法了，方法如下：

/* Set key to value, creating the key if it does not already exist.
 * If 'update' is not NULL, *update is set to 1 if the key was
 * already preset, otherwise to 0. */
unsigned char *zipmapSet(unsigned char *zm, unsigned char *key, unsigned int klen, unsigned char *val, unsigned int vlen, int *update) {
    unsigned int zmlen, offset;
    unsigned int freelen, reqlen = zipmapRequiredLength(klen,vlen);
    unsigned int empty, vempty;
    unsigned char *p;

    freelen = reqlen;
    if (update) *update = 0;
    //寻找key的位置
    p = zipmapLookupRaw(zm,key,klen,&zmlen);
    if (p == NULL) {
        /* Key not found: enlarge */
        //key的位置没有找到，调整zipmap的大小，准备添加操作
        zm = zipmapResize(zm, zmlen+reqlen);
        p = zm+zmlen-1;
        zmlen = zmlen+reqlen;

        /* Increase zipmap length (this is an insert) */
        //如果头字节还没有达到最大值，则递增
        if (zm[0] = ZIPMAP_VALUE_MAX_FREE) {
        /* First, move the tail  bytes to the front, then resize
         * the zipmap to be  bytes smaller. */
        offset = p-zm;
        memmove(p+reqlen, p+freelen, zmlen-(offset+freelen+1));
        zmlen -= empty;
        zm = zipmapResize(zm, zmlen);
        p = zm+offset;
        vempty = 0;
    } else {
        vempty = empty;
    }

    /* Just write the key + value and we are done. */
    /* Key: */
    //定位到插入的位置，首先写入key值
    p += zipmapEncodeLength(p,klen);
    memcpy(p,key,klen);
    p += klen;
    /* Value: */
    //key值后面是value值，再次写入
    p += zipmapEncodeLength(p,vlen);
    *p++ = vempty;
    memcpy(p,val,vlen);
    return zm;
}

map里返回长度的方法有点特别，就直接定位了就用一个字节存储长度：

/* Return the number of entries inside a zipmap */
/* 返回map的长度 */
unsigned int zipmapLen(unsigned char *zm) {
    unsigned int len = 0;
    //如果第一个长度小于最大值，则直接返回
    if (zm[0] 
平常我们在redis客户端执行set key "value"命令的时候，调用的其实就是set方法，如下： 
    zm = zipmapSet(zm,(unsigned char*) "name",4, (unsigned char*) "foo",3,NULL);
    zm = zipmapSet(zm,(unsigned char*) "surname",7, (unsigned char*) "foo",3,NULL);
    zm = zipmapSet(zm,(unsigned char*) "age",3, (unsigned char*) "foo",3,NULL);
比ziplist方法简单许多了，最后给出头文件 
/* String -> String Map data structure optimized for size.
 *
 * See zipmap.c for more info.
 *
 * --------------------------------------------------------------------------
 *
 * Copyright (c) 2009-2010, Salvatore Sanfilippo 
 * All rights reserved.
 *
 * Redistribution and use in source and binary forms, with or without
 * modification, are permitted provided that the following conditions are met:
 *
 *   * Redistributions of source code must retain the above copyright notice,
 *     this list of conditions and the following disclaimer.
 *   * Redistributions in binary form must reproduce the above copyright
 *     notice, this list of conditions and the following disclaimer in the
 *     documentation and/or other materials provided with the distribution.
 *   * Neither the name of Redis nor the names of its contributors may be used
 *     to endorse or promote products derived from this software without
 *     specific prior written permission.
 *
 * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
 * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
 * ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE
 * LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
 * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
 * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
 * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
 * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
 * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
 * POSSIBILITY OF SUCH DAMAGE.
 */

#ifndef _ZIPMAP_H
#define _ZIPMAP_H

unsigned char *zipmapNew(void);  //创建一个新的压缩图
unsigned char *zipmapSet(unsigned char *zm, unsigned char *key, unsigned int klen, unsigned char *val, unsigned int vlen, int *update); //设置压缩图中的某个键值对
unsigned char *zipmapDel(unsigned char *zm, unsigned char *key, unsigned int klen, int *deleted);  //删除压缩图上的某个键值对
unsigned char *zipmapRewind(unsigned char *zm);   //将在zipmapNext中被调用到
unsigned char *zipmapNext(unsigned char *zm, unsigned char **key, unsigned int *klen, unsigned char **value, unsigned int *vlen); //取得此键值对的下一个键值对
int zipmapGet(unsigned char *zm, unsigned char *key, unsigned int klen, unsigned char **value, unsigned int *vlen); //获取某个键值对
int zipmapExists(unsigned char *zm, unsigned char *key, unsigned int klen); //某个key值在zipmap中是否存在
unsigned int zipmapLen(unsigned char *zm); //zipmap压缩图的总键值对数
size_t zipmapBlobLen(unsigned char *zm); //压缩图的序列化到文件中所需大小
void zipmapRepr(unsigned char *p);  //输出的压缩图的具体信息，用于测试

#endif

最后，基于本人对redis源代码分析有一段时间了，我把分析好的代码，同步到了我的个人github上了，放上地址大家可以一起学习：
github：https://github.com/linyiqun/Redis-Code