/* Convert the set to specified encoding. The resulting dict (when converting
* to a hash table) is presized to hold the number of elements in the original
* set. */
void setTypeConvert(robj *setobj, int enc) {
setTypeIterator *si;
redisAssertWithInfo(NULL,setobj,setobj->type == REDIS_SET &&
setobj->encoding == REDIS_ENCODING_INTSET);
if (enc == REDIS_ENCODING_HT) {
int64_t intele;
dict *d = dictCreate(&setDictType,NULL);
robj *element;
/* Presize the dict to avoid rehashing */
dictExpand(d,intsetLen(setobj->ptr));
/* To add the elements we extract integers and create redis objects */
si = setTypeInitIterator(setobj);
while (setTypeNext(si,NULL,&intele) != -1) {
element = createStringObjectFromLongLong(intele);
redisAssertWithInfo(NULL,element,dictAdd(d,element,NULL) == DICT_OK);
}
setTypeReleaseIterator(si);
setobj->encoding = REDIS_ENCODING_HT;
zfree(setobj->ptr);
setobj->ptr = d;
} else {
redisPanic("Unsupported set conversion");
}
}
void lindexCommand(client *c) {
robj *o = lookupKeyReadOrReply(c,c->argv[1],shared.nullbulk);
if (o == NULL || checkType(c,o,OBJ_LIST)) return;
long index;
robj *value = NULL;
if ((getLongFromObjectOrReply(c, c->argv[2], &index, NULL) != C_OK))
return;
if (o->encoding == OBJ_ENCODING_QUICKLIST) {
quicklistEntry entry;
if (quicklistIndex(o->ptr, index, &entry)) {
if (entry.value) {
value = createStringObject((char*)entry.value,entry.sz);
} else {
value = createStringObjectFromLongLong(entry.longval);
}
addReplyBulk(c,value);
decrRefCount(value);
} else {
addReply(c,shared.nullbulk);
}
} else {
serverPanic("Unknown list encoding");
}
}
void spopCommand(redisClient *c) {
robj *set, *ele, *aux;
int64_t llele;
int encoding;
if ((set = lookupKeyWriteOrReply(c,c->argv[1],shared.nullbulk)) == NULL ||
checkType(c,set,REDIS_SET)) return;
encoding = setTypeRandomElement(set,&ele,&llele);
if (encoding == REDIS_ENCODING_INTSET) {
ele = createStringObjectFromLongLong(llele);
set->ptr = intsetRemove(set->ptr,llele,NULL);
} else {
incrRefCount(ele);
setTypeRemove(set,ele);
}
notifyKeyspaceEvent(REDIS_NOTIFY_SET,"spop",c->argv[1],c->db->id);
/* Replicate/AOF this command as an SREM operation */
aux = createStringObject("SREM",4);
rewriteClientCommandVector(c,3,aux,c->argv[1],ele);
decrRefCount(ele);
decrRefCount(aux);
addReplyBulk(c,ele);
if (setTypeSize(set) == 0) {
dbDelete(c->db,c->argv[1]);
notifyKeyspaceEvent(REDIS_NOTIFY_GENERIC,"del",c->argv[1],c->db->id);
}
signalModifiedKey(c->db,c->argv[1]);
server.dirty++;
}
void lindexCommand(redisClient *c) {
robj *o = lookupKeyReadOrReply(c,c->argv[1],shared.nullbulk);
if (o == NULL || checkType(c,o,REDIS_LIST)) return;
int index = atoi(c->argv[2]->ptr);
robj *value = NULL;
if (o->encoding == REDIS_ENCODING_ZIPLIST) {
unsigned char *p;
unsigned char *vstr;
unsigned int vlen;
long long vlong;
p = ziplistIndex(o->ptr,index);
if (ziplistGet(p,&vstr,&vlen,&vlong)) {
if (vstr) {
value = createStringObject((char*)vstr,vlen);
} else {
value = createStringObjectFromLongLong(vlong);
}
addReplyBulk(c,value);
decrRefCount(value);
} else {
addReply(c,shared.nullbulk);
}
} else if (o->encoding == REDIS_ENCODING_LINKEDLIST) {
listNode *ln = listIndex(o->ptr,index);
if (ln != NULL) {
value = listNodeValue(ln);
addReplyBulk(c,value);
} else {
addReply(c,shared.nullbulk);
}
} else {
redisPanic("Unknown list encoding");
}
}
/* Create the sds representation of an PEXPIREAT command, using
* 'seconds' as time to live and 'cmd' to understand what command
* we are translating into a PEXPIREAT.
*
* This command is used in order to translate EXPIRE and PEXPIRE commands
* into PEXPIREAT command so that we retain precision in the append only
* file, and the time is always absolute and not relative. */
sds catAppendOnlyExpireAtCommand(sds buf, struct redisCommand *cmd, robj *key, robj *seconds) {
long long when;
robj *argv[3];
/* Make sure we can use strtol */
seconds = getDecodedObject(seconds);
when = strtoll(seconds->ptr,NULL,10);
/* Convert argument into milliseconds for EXPIRE, SETEX, EXPIREAT */
if (cmd->proc == expireCommand || cmd->proc == setexCommand ||
cmd->proc == expireatCommand)
{
when *= 1000;
}
/* Convert into absolute time for EXPIRE, PEXPIRE, SETEX, PSETEX */
if (cmd->proc == expireCommand || cmd->proc == pexpireCommand ||
cmd->proc == setexCommand || cmd->proc == psetexCommand)
{
when += mstime();
}
decrRefCount(seconds);
argv[0] = createStringObject("PEXPIREAT",9);
argv[1] = key;
argv[2] = createStringObjectFromLongLong(when);
buf = catAppendOnlyGenericCommand(buf, 3, argv);
decrRefCount(argv[0]);
decrRefCount(argv[2]);
return buf;
}
void spopCommand(client *c) {
robj *set, *ele, *aux;
sds sdsele;
int64_t llele;
int encoding;
if (c->argc == 3) {
spopWithCountCommand(c);
return;
} else if (c->argc > 3) {
addReply(c,shared.syntaxerr);
return;
}
/* Make sure a key with the name inputted exists, and that it's type is
* indeed a set */
if ((set = lookupKeyWriteOrReply(c,c->argv[1],shared.nullbulk)) == NULL ||
checkType(c,set,OBJ_SET)) return;
/* Get a random element from the set */
encoding = setTypeRandomElement(set,&sdsele,&llele);
/* Remove the element from the set */
if (encoding == OBJ_ENCODING_INTSET) {
ele = createStringObjectFromLongLong(llele);
set->ptr = intsetRemove(set->ptr,llele,NULL);
} else {
ele = createStringObject(sdsele,sdslen(sdsele));
setTypeRemove(set,ele->ptr);
}
notifyKeyspaceEvent(NOTIFY_SET,"spop",c->argv[1],c->db->id);
/* Replicate/AOF this command as an SREM operation */
aux = createStringObject("SREM",4);
rewriteClientCommandVector(c,3,aux,c->argv[1],ele);
decrRefCount(aux);
/* Add the element to the reply */
addReplyBulk(c,ele);
decrRefCount(ele);
/* Delete the set if it's empty */
if (setTypeSize(set) == 0) {
dbDelete(c->db,c->argv[1]);
notifyKeyspaceEvent(NOTIFY_GENERIC,"del",c->argv[1],c->db->id);
}
/* Set has been modified */
signalModifiedKey(c->db,c->argv[1]);
server.dirty++;
}
void incrDecrCommand(redisClient *c, long long init_value, long long incr) {
c->returncode = REDIS_ERR;
long long value;
long long oldvalue;
robj *o;
oldvalue=0;
oldvalue++;
o = lookupKeyWriteWithVersion(c->db,c->argv[1],&(c->version));
if (o != NULL && checkType(c,o,REDIS_STRING)) {
c->returncode = REDIS_ERR_WRONG_TYPE_ERROR;
return;
}
robj* key = c->argv[1];
if(o != NULL) {
uint16_t version = sdsversion(key->ptr);
if(c->version_care && version != 0 && version != c->version) {
c->returncode = REDIS_ERR_VERSION_ERROR;
return;
} else {
sdsversion_change(key->ptr, c->version);
}
} else {
sdsversion_change(key->ptr, 0);
}
if(c->version_care) {
sdsversion_add(key->ptr, 1);
}
if (o == NULL) {
value = init_value;
} else if (getLongLongFromObject(o,&value) != REDIS_OK) {
c->returncode = REDIS_ERR_IS_NOT_INTEGER;
return;
}
oldvalue = value;
value += incr;
value = (int32_t)value;
o = createStringObjectFromLongLong(value);
dbSuperReplace(c->db,c->argv[1],o);
c->db->dirty++;
EXPIRE_OR_NOT
c->retvalue.llnum = value;
c->returncode = REDIS_OK;
}
/* set "OFFSET var" for next cursor iteration */
void incrOffsetVar(redisClient *c, cswc_t *w, long incr) {
robj *ovar = createStringObject(w->ovar, sdslen(w->ovar));
if (w->lim > incr) {
deleteKey(c->db, ovar);
} else {
lolo value = (w->ofst == -1) ? (lolo)incr :
(lolo)w->ofst + (lolo)incr;
robj *val = createStringObjectFromLongLong(value);
int ret = dictAdd(c->db->dict, ovar, val);
if (ret == DICT_ERR) dictReplace(c->db->dict, ovar, val);
}
server.dirty++;
}
/* Return entry or NULL at the current position of the iterator. */
robj *listTypeGet(listTypeEntry *entry) {
robj *value = NULL;
if (entry->li->encoding == OBJ_ENCODING_QUICKLIST) {
if (entry->entry.value) {
value = createStringObject((char *)entry->entry.value,
entry->entry.sz);
} else {
value = createStringObjectFromLongLong(entry->entry.longval);
}
} else {
serverPanic("Unknown list encoding");
}
return value;
}
//返回一个节点的value对象，根据当前的迭代器
robj *listTypeGet(listTypeEntry *entry) {
robj *value = NULL;
//对列表对象编码为quicklist类型操作
if (entry->li->encoding == OBJ_ENCODING_QUICKLIST) {
if (entry->entry.value) { //创建一个字符串对象保存列表类型的entry结构所指向的entry节点的字符串值
value = createStringObject((char *)entry->entry.value,
entry->entry.sz);
} else {
//创建一个字符串对象保存列表类型的entry结构所指向的entry节点的整型值
value = createStringObjectFromLongLong(entry->entry.longval);
}
} else {
serverPanic("Unknown list encoding");
}
return value;
}
robj *listTypePop(robj *subject, int where) {
long long vlong;
robj *value = NULL;
int ql_where = where == LIST_HEAD ? QUICKLIST_HEAD : QUICKLIST_TAIL;
if (subject->encoding == OBJ_ENCODING_QUICKLIST) {
if (quicklistPopCustom(subject->ptr, ql_where, (unsigned char **)&value,
NULL, &vlong, listPopSaver)) {
if (!value)
value = createStringObjectFromLongLong(vlong);
}
} else {
serverPanic("Unknown list encoding");
}
return value;
}
void incrDecrCommand(redisClient *c, long long incr) {
long long value;
robj *o;
o = lookupKeyWrite(c->db,c->argv[1]);
if (o != NULL && checkType(c,o,REDIS_STRING)) return;
if (getLongLongFromObjectOrReply(c,o,&value,NULL) != REDIS_OK) return;
value += incr;
o = createStringObjectFromLongLong(value);
dbReplace(c->db,c->argv[1],o);
server.dirty++;
addReply(c,shared.colon);
addReply(c,o);
addReply(c,shared.crlf);
}
/* The not copy on write friendly version but easy to use version
* of setTypeNext() is setTypeNextObject(), returning new objects
* or incrementing the ref count of returned objects. So if you don't
* retain a pointer to this object you should call decrRefCount() against it.
*
* This function is the way to go for write operations where COW is not
* an issue as the result will be anyway of incrementing the ref count. */
robj *setTypeNextObject(setTypeIterator *si) {
int64_t intele;
robj *objele;
int encoding;
encoding = setTypeNext(si,&objele,&intele);
switch(encoding) {
case -1: return NULL;
case REDIS_ENCODING_INTSET:
return createStringObjectFromLongLong(intele);
case REDIS_ENCODING_HT:
incrRefCount(objele);
return objele;
default:
redisPanic("Unsupported encoding");
}
return NULL; /* just to suppress warnings */
}
//列表类型的从where弹出一个value，POP命令底层实现
robj *listTypePop(robj *subject, int where) {
long long vlong;
robj *value = NULL;
//获得POP的位置，quicklist的头部或尾部
int ql_where = where == LIST_HEAD ? QUICKLIST_HEAD : QUICKLIST_TAIL;
//对列表对象编码为quicklist类型操作
if (subject->encoding == OBJ_ENCODING_QUICKLIST) {
//从ql_where位置POP出一个entry节点，保存在value或vlong中
if (quicklistPopCustom(subject->ptr, ql_where, (unsigned char **)&value,
NULL, &vlong, listPopSaver)) {
if (!value) //如果弹出的entry节点是整型的
//则根据整型值创建一个字符串对象
value = createStringObjectFromLongLong(vlong);
}
} else {
serverPanic("Unknown list encoding");
}
return value; //返回弹出entry节点的value值
}
/* A non copy-on-write friendly but higher level version of hashTypeCurrent*()
* that returns an object with incremented refcount (or a new object). It is up
* to the caller to decrRefCount() the object if no reference is retained. */
robj *hashTypeCurrentObject(hashTypeIterator *hi, int what) {
robj *dst;
if (hi->encoding == REDIS_ENCODING_ZIPLIST) {
unsigned char *vstr = NULL;
unsigned int vlen = UINT_MAX;
long long vll = LLONG_MAX;
hashTypeCurrentFromZiplist(hi, what, &vstr, &vlen, &vll);
if (vstr) {
dst = createStringObject((char*)vstr, vlen);
} else {
dst = createStringObjectFromLongLong(vll);
}
} else if (hi->encoding == REDIS_ENCODING_HT) {
hashTypeCurrentFromHashTable(hi, what, &dst);
incrRefCount(dst);
} else {
logicError("Unknown hash encoding");
}
return dst;
}
robj *listTypePop(robj *subject, int where) {
robj *value = NULL;
if (subject->encoding == REDIS_ENCODING_ZIPLIST) {
unsigned char *p;
unsigned char *vstr;
unsigned int vlen;
long long vlong;
int pos = (where == REDIS_HEAD) ? 0 : -1;
p = ziplistIndex(subject->ptr,pos);
if (ziplistGet(p,&vstr,&vlen,&vlong)) {
if (vstr) {
value = createStringObject((char*)vstr,vlen);
} else {
value = createStringObjectFromLongLong(vlong);
}
/* We only need to delete an element when it exists */
subject->ptr = ziplistDelete(subject->ptr,&p);
}
} else if (subject->encoding == REDIS_ENCODING_LINKEDLIST) {
list *list = subject->ptr;
listNode *ln;
if (where == REDIS_HEAD) {
ln = listFirst(list);
} else {
ln = listLast(list);
}
if (ln != NULL) {
value = listNodeValue(ln);
incrRefCount(value);
listDelNode(list,ln);
}
} else {
redisPanic("Unknown list encoding");
}
return value;
}
/* Return entry or NULL at the current position of the iterator. */
robj *listTypeGet(listTypeEntry *entry) {
listTypeIterator *li = entry->li;
robj *value = NULL;
if (li->encoding == REDIS_ENCODING_ZIPLIST) {
unsigned char *vstr;
unsigned int vlen;
long long vlong;
redisAssert(entry->zi != NULL);
if (ziplistGet(entry->zi,&vstr,&vlen,&vlong)) {
if (vstr) {
value = createStringObject((char*)vstr,vlen);
} else {
value = createStringObjectFromLongLong(vlong);
}
}
} else if (li->encoding == REDIS_ENCODING_LINKEDLIST) {
redisAssert(entry->ln != NULL);
value = listNodeValue(entry->ln);
incrRefCount(value);
} else {
redisPanic("Unknown list encoding");
}
return value;
}
void sinterGenericCommand(redisClient *c, robj **setkeys, unsigned long setnum, robj *dstkey) {
robj **sets = zmalloc(sizeof(robj*)*setnum);
setTypeIterator *si;//迭代器
robj *eleobj, *dstset = NULL;
int64_t intobj;
void *replylen = NULL;
unsigned long j, cardinality = 0;
int encoding;
for (j = 0; j < setnum; j++) {//得到所有的集合
robj *setobj = dstkey ?
lookupKeyWrite(c->db,setkeys[j]) :
lookupKeyRead(c->db,setkeys[j]);
if (!setobj) {//任何一个集合不存在，那么总的交集就为空
zfree(sets);
if (dstkey) {
if (dbDelete(c->db,dstkey)) {
signalModifiedKey(c->db,dstkey);
server.dirty++;
}
addReply(c,shared.czero);
} else {
addReply(c,shared.emptymultibulk);
}
return;
}
if (checkType(c,setobj,REDIS_SET)) {
zfree(sets);
return;
}
sets[j] = setobj;
}
/* Sort sets from the smallest to largest, this will improve our
* algorithm's performance */
//按照集合元素个数从小到大排序
qsort(sets,setnum,sizeof(robj*),qsortCompareSetsByCardinality);
/* The first thing we should output is the total number of elements...
* since this is a multi-bulk write, but at this stage we don't know
* the intersection set size, so we use a trick, append an empty object
* to the output list and save the pointer to later modify it with the
* right length */
if (!dstkey) {
replylen = addDeferredMultiBulkLength(c);
} else {
/* If we have a target key where to store the resulting set
* create this key with an empty set inside */
dstset = createIntsetObject();
}
/* Iterate all the elements of the first (smallest) set, and test
* the element against all the other sets, if at least one set does
* not include the element it is discarded */
/**
求多个集合交集的算法思想：
首先按照集合元素个数对集合进行qsort，然后遍历排序后的第一个集合中的元素，查看该元素在
其他集合中是否存在，如果在其他集合中都存在，那么该元素为一个结果
*/
si = setTypeInitIterator(sets[0]);
while((encoding = setTypeNext(si,&eleobj,&intobj)) != -1) {
for (j = 1; j < setnum; j++) {
if (sets[j] == sets[0]) continue;//这段代码没意义啊
if (encoding == REDIS_ENCODING_INTSET) {//intset
/* intset with intset is simple... and fast */
//集合sets[j]编码为intset
if (sets[j]->encoding == REDIS_ENCODING_INTSET &&
!intsetFind((intset*)sets[j]->ptr,intobj))//在集合sets[j]中没有找到集合sets[0]的intobj
{
break;
/* in order to compare an integer with an object we
* have to use the generic function, creating an object
* for this */
} else if (sets[j]->encoding == REDIS_ENCODING_HT) {//集合sets[j]编码为HT，sets[0]为INTSET
eleobj = createStringObjectFromLongLong(intobj);//将sets[0]中的intobj转换为sds
if (!setTypeIsMember(sets[j],eleobj)) {//如果eleobj不在集合sets[j]中
decrRefCount(eleobj);
break;
}
decrRefCount(eleobj);
}
} else if (encoding == REDIS_ENCODING_HT) {//HT
/* Optimization... if the source object is integer
* encoded AND the target set is an intset, we can get
* a much faster path. */
if (eleobj->encoding == REDIS_ENCODING_INT &&
sets[j]->encoding == REDIS_ENCODING_INTSET &&
!intsetFind((intset*)sets[j]->ptr,(long)eleobj->ptr))
{
break;
/* else... object to object check is easy as we use the
* type agnostic API here. */
} else if (!setTypeIsMember(sets[j],eleobj)) {
break;
}
}
}
/* Only take action when all sets contain the member */
if (j == setnum) {
if (!dstkey) {
if (encoding == REDIS_ENCODING_HT)
addReplyBulk(c,eleobj);
else
addReplyBulkLongLong(c,intobj);
cardinality++;
} else {//添加到临时目标集合
if (encoding == REDIS_ENCODING_INTSET) {
eleobj = createStringObjectFromLongLong(intobj);
setTypeAdd(dstset,eleobj);
decrRefCount(eleobj);
} else {
setTypeAdd(dstset,eleobj);
}
}
}
}
setTypeReleaseIterator(si);
if (dstkey) {
/* Store the resulting set into the target, if the intersection
* is not an empty set. */
int deleted = dbDelete(c->db,dstkey);//覆盖原来的目标集合
if (setTypeSize(dstset) > 0) {
dbAdd(c->db,dstkey,dstset);
addReplyLongLong(c,setTypeSize(dstset));
notifyKeyspaceEvent(REDIS_NOTIFY_SET,"sinterstore",
dstkey,c->db->id);
} else {//空集
decrRefCount(dstset);
addReply(c,shared.czero);
if (deleted)
notifyKeyspaceEvent(REDIS_NOTIFY_GENERIC,"del",
dstkey,c->db->id);
}
signalModifiedKey(c->db,dstkey);
server.dirty++;
} else {
setDeferredMultiBulkLength(c,replylen,cardinality);
}
zfree(sets);
}
void sinterGenericCommand(redisClient *c, robj **setkeys, unsigned long setnum, robj *dstkey) {
robj **sets = zmalloc(sizeof(robj*)*setnum);
setTypeIterator *si;
robj *eleobj, *dstset = NULL;
int64_t intobj;
void *replylen = NULL;
unsigned long j, cardinality = 0;
int encoding;
for (j = 0; j < setnum; j++) {
robj *setobj = dstkey ?
lookupKeyWrite(c->db,setkeys[j]) :
lookupKeyRead(c->db,setkeys[j]);
if (!setobj) {
zfree(sets);
if (dstkey) {
if (dbDelete(c->db,dstkey)) {
touchWatchedKey(c->db,dstkey);
server.dirty++;
}
addReply(c,shared.czero);
} else {
addReply(c,shared.emptymultibulk);
}
return;
}
if (checkType(c,setobj,REDIS_SET)) {
zfree(sets);
return;
}
sets[j] = setobj;
}
/* Sort sets from the smallest to largest, this will improve our
* algorithm's performace */
qsort(sets,setnum,sizeof(robj*),qsortCompareSetsByCardinality);
/* The first thing we should output is the total number of elements...
* since this is a multi-bulk write, but at this stage we don't know
* the intersection set size, so we use a trick, append an empty object
* to the output list and save the pointer to later modify it with the
* right length */
if (!dstkey) {
replylen = addDeferredMultiBulkLength(c);
} else {
/* If we have a target key where to store the resulting set
* create this key with an empty set inside */
dstset = createIntsetObject();
}
/* Iterate all the elements of the first (smallest) set, and test
* the element against all the other sets, if at least one set does
* not include the element it is discarded */
si = setTypeInitIterator(sets[0]);
while((encoding = setTypeNext(si,&eleobj,&intobj)) != -1) {
for (j = 1; j < setnum; j++) {
if (encoding == REDIS_ENCODING_INTSET) {
/* intset with intset is simple... and fast */
if (sets[j]->encoding == REDIS_ENCODING_INTSET &&
!intsetFind((intset*)sets[j]->ptr,intobj))
{
break;
/* in order to compare an integer with an object we
* have to use the generic function, creating an object
* for this */
} else if (sets[j]->encoding == REDIS_ENCODING_HT) {
eleobj = createStringObjectFromLongLong(intobj);
if (!setTypeIsMember(sets[j],eleobj)) {
decrRefCount(eleobj);
break;
}
decrRefCount(eleobj);
}
} else if (encoding == REDIS_ENCODING_HT) {
/* Optimization... if the source object is integer
* encoded AND the target set is an intset, we can get
* a much faster path. */
if (eleobj->encoding == REDIS_ENCODING_INT &&
sets[j]->encoding == REDIS_ENCODING_INTSET &&
!intsetFind((intset*)sets[j]->ptr,(long)eleobj->ptr))
{
break;
/* else... object to object check is easy as we use the
* type agnostic API here. */
} else if (!setTypeIsMember(sets[j],eleobj)) {
break;
}
}
}
/* Only take action when all sets contain the member */
if (j == setnum) {
if (!dstkey) {
if (encoding == REDIS_ENCODING_HT)
addReplyBulk(c,eleobj);
else
addReplyBulkLongLong(c,intobj);
cardinality++;
} else {
if (encoding == REDIS_ENCODING_INTSET) {
eleobj = createStringObjectFromLongLong(intobj);
setTypeAdd(dstset,eleobj);
decrRefCount(eleobj);
} else {
setTypeAdd(dstset,eleobj);
}
}
}
}
setTypeReleaseIterator(si);
if (dstkey) {
/* Store the resulting set into the target, if the intersection
* is not an empty set. */
dbDelete(c->db,dstkey);
if (setTypeSize(dstset) > 0) {
dbAdd(c->db,dstkey,dstset);
addReplyLongLong(c,setTypeSize(dstset));
} else {
decrRefCount(dstset);
addReply(c,shared.czero);
}
touchWatchedKey(c->db,dstkey);
server.dirty++;
} else {
setDeferredMultiBulkLength(c,replylen,cardinality);
}
zfree(sets);
}
/* This command implements SCAN, HSCAN and SSCAN commands.
* If object 'o' is passed, then it must be a Hash or Set object, otherwise
* if 'o' is NULL the command will operate on the dictionary associated with
* the current database.
*
* When 'o' is not NULL the function assumes that the first argument in
* the client arguments vector is a key so it skips it before iterating
* in order to parse options.
*
* In the case of a Hash object the function returns both the field and value
* of every element on the Hash. */
void scanGenericCommand(client *c, robj *o, unsigned long cursor) {
int i, j;
list *keys = listCreate();
listNode *node, *nextnode;
long count = 10;
sds pat = NULL;
int patlen = 0, use_pattern = 0;
dict *ht;
/* Object must be NULL (to iterate keys names), or the type of the object
* must be Set, Sorted Set, or Hash. */
serverAssert(o == NULL || o->type == OBJ_SET || o->type == OBJ_HASH ||
o->type == OBJ_ZSET);
/* Set i to the first option argument. The previous one is the cursor. */
i = (o == NULL) ? 2 : 3; /* Skip the key argument if needed. */
/* Step 1: Parse options. */
while (i < c->argc) {
j = c->argc - i;
if (!strcasecmp(c->argv[i]->ptr, "count") && j >= 2) {
if (getLongFromObjectOrReply(c, c->argv[i+1], &count, NULL)
!= C_OK)
{
goto cleanup;
}
if (count < 1) {
addReply(c,shared.syntaxerr);
goto cleanup;
}
i += 2;
} else if (!strcasecmp(c->argv[i]->ptr, "match") && j >= 2) {
pat = c->argv[i+1]->ptr;
patlen = sdslen(pat);
/* The pattern always matches if it is exactly "*", so it is
* equivalent to disabling it. */
use_pattern = !(pat[0] == '*' && patlen == 1);
i += 2;
} else {
addReply(c,shared.syntaxerr);
goto cleanup;
}
}
/* Step 2: Iterate the collection.
*
* Note that if the object is encoded with a ziplist, intset, or any other
* representation that is not a hash table, we are sure that it is also
* composed of a small number of elements. So to avoid taking state we
* just return everything inside the object in a single call, setting the
* cursor to zero to signal the end of the iteration. */
/* Handle the case of a hash table. */
ht = NULL;
if (o == NULL) {
ht = c->db->dict;
} else if (o->type == OBJ_SET && o->encoding == OBJ_ENCODING_HT) {
ht = o->ptr;
} else if (o->type == OBJ_HASH && o->encoding == OBJ_ENCODING_HT) {
ht = o->ptr;
count *= 2; /* We return key / value for this type. */
} else if (o->type == OBJ_ZSET && o->encoding == OBJ_ENCODING_SKIPLIST) {
zset *zs = o->ptr;
ht = zs->dict;
count *= 2; /* We return key / value for this type. */
}
if (ht) {
void *privdata[2];
/* We set the max number of iterations to ten times the specified
* COUNT, so if the hash table is in a pathological state (very
* sparsely populated) we avoid to block too much time at the cost
* of returning no or very few elements. */
long maxiterations = count*10;
/* We pass two pointers to the callback: the list to which it will
* add new elements, and the object containing the dictionary so that
* it is possible to fetch more data in a type-dependent way. */
privdata[0] = keys;
privdata[1] = o;
do {
cursor = dictScan(ht, cursor, scanCallback, privdata);
} while (cursor &&
maxiterations-- &&
listLength(keys) < (unsigned long)count);
} else if (o->type == OBJ_SET) {
int pos = 0;
int64_t ll;
while(intsetGet(o->ptr,pos++,&ll))
listAddNodeTail(keys,createStringObjectFromLongLong(ll));
cursor = 0;
} else if (o->type == OBJ_HASH || o->type == OBJ_ZSET) {
unsigned char *p = ziplistIndex(o->ptr,0);
unsigned char *vstr;
unsigned int vlen;
long long vll;
while(p) {
ziplistGet(p,&vstr,&vlen,&vll);
listAddNodeTail(keys,
(vstr != NULL) ? createStringObject((char*)vstr,vlen) :
createStringObjectFromLongLong(vll));
p = ziplistNext(o->ptr,p);
}
cursor = 0;
} else {
serverPanic("Not handled encoding in SCAN.");
}
/* Step 3: Filter elements. */
node = listFirst(keys);
while (node) {
robj *kobj = listNodeValue(node);
nextnode = listNextNode(node);
int filter = 0;
/* Filter element if it does not match the pattern. */
if (!filter && use_pattern) {
if (sdsEncodedObject(kobj)) {
if (!stringmatchlen(pat, patlen, kobj->ptr, sdslen(kobj->ptr), 0))
filter = 1;
} else {
char buf[LONG_STR_SIZE];
int len;
serverAssert(kobj->encoding == OBJ_ENCODING_INT);
len = ll2string(buf,sizeof(buf),(long)kobj->ptr);
if (!stringmatchlen(pat, patlen, buf, len, 0)) filter = 1;
}
}
/* Filter element if it is an expired key. */
if (!filter && o == NULL && expireIfNeeded(c->db, kobj)) filter = 1;
/* Remove the element and its associted value if needed. */
if (filter) {
decrRefCount(kobj);
listDelNode(keys, node);
}
/* If this is a hash or a sorted set, we have a flat list of
* key-value elements, so if this element was filtered, remove the
* value, or skip it if it was not filtered: we only match keys. */
if (o && (o->type == OBJ_ZSET || o->type == OBJ_HASH)) {
node = nextnode;
nextnode = listNextNode(node);
if (filter) {
kobj = listNodeValue(node);
decrRefCount(kobj);
listDelNode(keys, node);
}
}
node = nextnode;
}
/* Step 4: Reply to the client. */
addReplyMultiBulkLen(c, 2);
addReplyBulkLongLong(c,cursor);
addReplyMultiBulkLen(c, listLength(keys));
while ((node = listFirst(keys)) != NULL) {
robj *kobj = listNodeValue(node);
addReplyBulk(c, kobj);
decrRefCount(kobj);
listDelNode(keys, node);
}
cleanup:
listSetFreeMethod(keys,decrRefCountVoid);
listRelease(keys);
}
void srandmemberWithCountCommand(redisClient *c) {
long l;
unsigned long count, size;
int uniq = 1;
robj *set, *ele;
int64_t llele;
int encoding;
dict *d;
if (getLongFromObjectOrReply(c,c->argv[2],&l,NULL) != REDIS_OK) return;
if (l >= 0) {
count = (unsigned) l;
} else {
/* A negative count means: return the same elements multiple times
* (i.e. don't remove the extracted element after every extraction). */
count = -l;
uniq = 0;
}
if ((set = lookupKeyReadOrReply(c,c->argv[1],shared.emptymultibulk))
== NULL || checkType(c,set,REDIS_SET)) return;
size = setTypeSize(set);
/* If count is zero, serve it ASAP to avoid special cases later. */
if (count == 0) {
addReply(c,shared.emptymultibulk);
return;
}
/* CASE 1: The count was negative, so the extraction method is just:
* "return N random elements" sampling the whole set every time.
* This case is trivial and can be served without auxiliary data
* structures. */
if (!uniq) {
addReplyMultiBulkLen(c,count);
while(count--) {
encoding = setTypeRandomElement(set,&ele,&llele);
if (encoding == REDIS_ENCODING_INTSET) {
addReplyBulkLongLong(c,llele);
} else {
addReplyBulk(c,ele);
}
}
return;
}
/* CASE 2:
* The number of requested elements is greater than the number of
* elements inside the set: simply return the whole set. */
if (count >= size) {
sunionDiffGenericCommand(c,c->argv+1,1,NULL,REDIS_OP_UNION);
return;
}
/* For CASE 3 and CASE 4 we need an auxiliary dictionary. */
d = dictCreate(&setDictType,NULL);
/* CASE 3:
* The number of elements inside the set is not greater than
* SRANDMEMBER_SUB_STRATEGY_MUL times the number of requested elements.
* In this case we create a set from scratch with all the elements, and
* subtract random elements to reach the requested number of elements.
*
* This is done because if the number of requsted elements is just
* a bit less than the number of elements in the set, the natural approach
* used into CASE 3 is highly inefficient. */
if (count*SRANDMEMBER_SUB_STRATEGY_MUL > size) {
setTypeIterator *si;
/* Add all the elements into the temporary dictionary. */
si = setTypeInitIterator(set);
while((encoding = setTypeNext(si,&ele,&llele)) != -1) {
int retval = DICT_ERR;
if (encoding == REDIS_ENCODING_INTSET) {
retval = dictAdd(d,createStringObjectFromLongLong(llele),NULL);
} else if (ele->encoding == REDIS_ENCODING_RAW) {
retval = dictAdd(d,dupStringObject(ele),NULL);
} else if (ele->encoding == REDIS_ENCODING_INT) {
retval = dictAdd(d,
createStringObjectFromLongLong((long)ele->ptr),NULL);
}
redisAssert(retval == DICT_OK);
}
setTypeReleaseIterator(si);
redisAssert(dictSize(d) == size);
/* Remove random elements to reach the right count. */
while(size > count) {
dictEntry *de;
de = dictGetRandomKey(d);
dictDelete(d,dictGetKey(de));
size--;
}
}
/* CASE 4: We have a big set compared to the requested number of elements.
* In this case we can simply get random elements from the set and add
* to the temporary set, trying to eventually get enough unique elements
* to reach the specified count. */
else {
unsigned long added = 0;
while(added < count) {
encoding = setTypeRandomElement(set,&ele,&llele);
if (encoding == REDIS_ENCODING_INTSET) {
ele = createStringObjectFromLongLong(llele);
} else if (ele->encoding == REDIS_ENCODING_RAW) {
ele = dupStringObject(ele);
} else if (ele->encoding == REDIS_ENCODING_INT) {
ele = createStringObjectFromLongLong((long)ele->ptr);
}
/* Try to add the object to the dictionary. If it already exists
* free it, otherwise increment the number of objects we have
* in the result dictionary. */
if (dictAdd(d,ele,NULL) == DICT_OK)
added++;
else
decrRefCount(ele);
}
}
/* CASE 3 & 4: send the result to the user. */
{
dictIterator *di;
dictEntry *de;
addReplyMultiBulkLen(c,count);
di = dictGetIterator(d);
while((de = dictNext(di)) != NULL)
addReplyBulk(c,dictGetKey(de));
dictReleaseIterator(di);
dictRelease(d);
}
}
void spopWithCountCommand(redisClient *c) {
long l;
unsigned long count, size;
robj *set;
/* Get the count argument */
if (getLongFromObjectOrReply(c,c->argv[2],&l,NULL) != REDIS_OK) return;
if (l >= 0) {
count = (unsigned) l;
} else {
addReply(c,shared.outofrangeerr);
return;
}
/* Make sure a key with the name inputted exists, and that it's type is
* indeed a set. Otherwise, return nil */
if ((set = lookupKeyReadOrReply(c,c->argv[1],shared.emptymultibulk))
== NULL || checkType(c,set,REDIS_SET)) return;
/* If count is zero, serve an empty multibulk ASAP to avoid special
* cases later. */
if (count == 0) {
addReply(c,shared.emptymultibulk);
return;
}
size = setTypeSize(set);
/* Generate an SPOP keyspace notification */
notifyKeyspaceEvent(REDIS_NOTIFY_SET,"spop",c->argv[1],c->db->id);
server.dirty += count;
/* CASE 1:
* The number of requested elements is greater than or equal to
* the number of elements inside the set: simply return the whole set. */
if (count >= size) {
/* We just return the entire set */
sunionDiffGenericCommand(c,c->argv+1,1,NULL,REDIS_OP_UNION);
/* Delete the set as it is now empty */
dbDelete(c->db,c->argv[1]);
notifyKeyspaceEvent(REDIS_NOTIFY_GENERIC,"del",c->argv[1],c->db->id);
/* Propagate this command as an DEL operation */
rewriteClientCommandVector(c,2,shared.del,c->argv[1]);
signalModifiedKey(c->db,c->argv[1]);
server.dirty++;
return;
}
/* Case 2 and 3 require to replicate SPOP as a set of SERM commands.
* Prepare our replication argument vector. Also send the array length
* which is common to both the code paths. */
robj *propargv[3];
propargv[0] = createStringObject("SREM",4);
propargv[1] = c->argv[1];
addReplyMultiBulkLen(c,count);
/* Common iteration vars. */
robj *objele;
int encoding;
int64_t llele;
unsigned long remaining = size-count; /* Elements left after SPOP. */
/* If we are here, the number of requested elements is less than the
* number of elements inside the set. Also we are sure that count < size.
* Use two different strategies.
*
* CASE 2: The number of elements to return is small compared to the
* set size. We can just extract random elements and return them to
* the set. */
if (remaining*SPOP_MOVE_STRATEGY_MUL > count) {
while(count--) {
encoding = setTypeRandomElement(set,&objele,&llele);
if (encoding == REDIS_ENCODING_INTSET) {
objele = createStringObjectFromLongLong(llele);
} else {
incrRefCount(objele);
}
/* Return the element to the client and remove from the set. */
addReplyBulk(c,objele);
setTypeRemove(set,objele);
/* Replicate/AOF this command as an SREM operation */
propargv[2] = objele;
alsoPropagate(server.sremCommand,c->db->id,propargv,3,
REDIS_PROPAGATE_AOF|REDIS_PROPAGATE_REPL);
decrRefCount(objele);
}
} else {
/* CASE 3: The number of elements to return is very big, approaching
* the size of the set itself. After some time extracting random elements
* from such a set becomes computationally expensive, so we use
* a different strategy, we extract random elements that we don't
* want to return (the elements that will remain part of the set),
* creating a new set as we do this (that will be stored as the original
* set). Then we return the elements left in the original set and
* release it. */
robj *newset = NULL;
/* Create a new set with just the remaining elements. */
while(remaining--) {
encoding = setTypeRandomElement(set,&objele,&llele);
if (encoding == REDIS_ENCODING_INTSET) {
objele = createStringObjectFromLongLong(llele);
} else {
incrRefCount(objele);
}
if (!newset) newset = setTypeCreate(objele);
setTypeAdd(newset,objele);
setTypeRemove(set,objele);
decrRefCount(objele);
}
/* Assign the new set as the key value. */
incrRefCount(set); /* Protect the old set value. */
dbOverwrite(c->db,c->argv[1],newset);
/* Tranfer the old set to the client and release it. */
setTypeIterator *si;
si = setTypeInitIterator(set);
while((encoding = setTypeNext(si,&objele,&llele)) != -1) {
if (encoding == REDIS_ENCODING_INTSET) {
objele = createStringObjectFromLongLong(llele);
} else {
incrRefCount(objele);
}
addReplyBulk(c,objele);
/* Replicate/AOF this command as an SREM operation */
propargv[2] = objele;
alsoPropagate(server.sremCommand,c->db->id,propargv,3,
REDIS_PROPAGATE_AOF|REDIS_PROPAGATE_REPL);
decrRefCount(objele);
}
setTypeReleaseIterator(si);
decrRefCount(set);
}
/* Don't propagate the command itself even if we incremented the
* dirty counter. We don't want to propagate an SPOP command since
* we propagated the command as a set of SREMs operations using
* the alsoPropagate() API. */
decrRefCount(propargv[0]);
preventCommandPropagation(c);
}
