/* ACKJOB jobid_1 jobid_2 ... jobid_N
*
* Set job state as acknowledged, if the job does not exist creates a
* fake job just to hold the acknowledge.
*
* As a result of a job being acknowledged, the system tries to garbage
* collect it, that is, to remove the job from every node of the system
* in order to both avoid multiple deliveries of the same job, and to
* release resources.
*
* If a job was already acknowledged, the ACKJOB command still has the
* effect of forcing a GC attempt ASAP.
*
* The command returns the number of jobs already known and that were
* already not in the ACKED state.
*/
void ackjobCommand(client *c) {
int j, known = 0;
if (validateJobIDs(c,c->argv+1,c->argc-1) == C_ERR) return;
/* Perform the appropriate action for each job. */
for (j = 1; j < c->argc; j++) {
job *job = lookupJob(c->argv[j]->ptr);
/* Case 1: No such job. Create one just to hold the ACK. However
* if the cluster is composed by a single node we are sure the job
* does not exist in the whole cluster, so do this only if the
* cluster size is greater than one. */
if (job == NULL && server.cluster->size > 1 && !myselfLeaving()) {
char *id = c->argv[j]->ptr;
int ttl = getRawTTLFromJobID(id);
/* TTL is even for "at most once" jobs. In this case we
* don't need to create a dummy hack. */
if (ttl & 1) {
job = createJob(id,JOB_STATE_ACKED,0,0);
setJobTTLFromID(job);
serverAssert(registerJob(job) == C_OK);
}
}
/* Case 2: Job exists and is not acknowledged. Change state. */
else if (job && job->state != JOB_STATE_ACKED) {
dequeueJob(job); /* Safe to call if job is not queued. */
acknowledgeJob(job);
known++;
}
/* Anyway... start a GC attempt on the acked job. */
if (job) tryJobGC(job);
}
addReplyLongLong(c,known);
}
//如果有client因为等待一个key被push而被阻塞，那么将这个key放入ready_keys,key哈希表中
void signalListAsReady(redisDb *db, robj *key) {
readyList *rl;
/* No clients blocking for this key? No need to queue it. */
//如果在key不是正处于阻塞状态的键则返回
if (dictFind(db->blocking_keys,key) == NULL) return;
/* Key was already signaled? No need to queue it again. */
//key已经是ready_keys链表里的键，则返回
if (dictFind(db->ready_keys,key) != NULL) return;
/* Ok, we need to queue this key into server.ready_keys. */
//接下来需要将key添加到ready_keys中
//分配一个readyList结构的空间，该结构记录要解除client的阻塞状态的键
rl = zmalloc(sizeof(*rl));
rl->key = key; //设置要解除的键
rl->db = db; //设置所在数据库
incrRefCount(key);
//将rl添加到server.ready_keys的末尾
listAddNodeTail(server.ready_keys,rl);
/* We also add the key in the db->ready_keys dictionary in order
* to avoid adding it multiple times into a list with a simple O(1)
* check. */
//再讲key添加到ready_keys哈希表中，防止重复添加
incrRefCount(key);
serverAssert(dictAdd(db->ready_keys,key,NULL) == DICT_OK);
}
/* Add the specified value into a set.
*
* If the value was already member of the set, nothing is done and 0 is
* returned, otherwise the new element is added and 1 is returned. */
int setTypeAdd(robj *subject, sds value) {
long long llval;
if (subject->encoding == OBJ_ENCODING_HT) {
dict *ht = subject->ptr;
dictEntry *de = dictAddRaw(ht,value);
if (de) {
dictSetKey(ht,de,sdsdup(value));
dictSetVal(ht,de,NULL);
return 1;
}
} else if (subject->encoding == OBJ_ENCODING_INTSET) {
if (isSdsRepresentableAsLongLong(value,&llval) == C_OK) {
uint8_t success = 0;
subject->ptr = intsetAdd(subject->ptr,llval,&success);
if (success) {
/* Convert to regular set when the intset contains
* too many entries. */
if (intsetLen(subject->ptr) > server.set_max_intset_entries)
setTypeConvert(subject,OBJ_ENCODING_HT);
return 1;
}
} else {
/* Failed to get integer from object, convert to regular set. */
setTypeConvert(subject,OBJ_ENCODING_HT);
/* The set *was* an intset and this value is not integer
* encodable, so dictAdd should always work. */
serverAssert(dictAdd(subject->ptr,sdsdup(value),NULL) == DICT_OK);
return 1;
}
} else {
serverPanic("Unknown set encoding");
}
return 0;
}
/* 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;
serverAssertWithInfo(NULL,setobj,setobj->type == OBJ_SET &&
setobj->encoding == OBJ_ENCODING_INTSET);
if (enc == OBJ_ENCODING_HT) {
int64_t intele;
dict *d = dictCreate(&setDictType,NULL);
sds 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,&element,&intele) != -1) {
element = sdsfromlonglong(intele);
serverAssert(dictAdd(d,element,NULL) == DICT_OK);
}
setTypeReleaseIterator(si);
setobj->encoding = OBJ_ENCODING_HT;
zfree(setobj->ptr);
setobj->ptr = d;
} else {
serverPanic("Unsupported set conversion");
}
}
void freeClient(client *c) {
listNode *ln;
if (c->flags & CLIENT_BLOCKED) {
if (c->bfree) c->bfree(c->bpop.data);
}
/* If this is marked as current client unset it */
if (server.current_client == c) server.current_client = NULL;
/* Free the query buffer */
sdsfree(c->querybuf);
c->querybuf = NULL;
/* Close socket, unregister events, and remove list of replies and
* accumulated arguments. */
if (c->fd != -1) {
aeDeleteFileEvent(server.el,c->fd,AE_READABLE);
aeDeleteFileEvent(server.el,c->fd,AE_WRITABLE);
close(c->fd);
}
freeClientArgv(c);
/* Remove from the list of clients */
if (c->fd != -1) {
ln = listSearchKey(server.clients,c);
serverAssert(ln != NULL);
listDelNode(server.clients,ln);
}
/* If this client was scheduled for async freeing we need to remove it
* from the queue. */
if (c->flags & CLIENT_CLOSE_ASAP) {
ln = listSearchKey(server.clients_to_close,c);
serverAssert(ln != NULL);
listDelNode(server.clients_to_close,ln);
}
/* Release other dynamically allocated client structure fields,
* and finally release the client structure itself. */
free(c->argv);
free(c);
}
/* Prepare the string object stored at 'key' to be modified destructively
* to implement commands like SETBIT or APPEND.
*
* An object is usually ready to be modified unless one of the two conditions
* are true:
*
* 1) The object 'o' is shared (refcount > 1), we don't want to affect
* other users.
* 2) The object encoding is not "RAW".
*
* If the object is found in one of the above conditions (or both) by the
* function, an unshared / not-encoded copy of the string object is stored
* at 'key' in the specified 'db'. Otherwise the object 'o' itself is
* returned.
*
* USAGE:
*
* The object 'o' is what the caller already obtained by looking up 'key'
* in 'db', the usage pattern looks like this:
*
* o = lookupKeyWrite(db,key);
* if (checkType(c,o,OBJ_STRING)) return;
* o = dbUnshareStringValue(db,key,o);
*
* At this point the caller is ready to modify the object, for example
* using an sdscat() call to append some data, or anything else.
*/
robj *dbUnshareStringValue(redisDb *db, robj *key, robj *o) {
serverAssert(o->type == OBJ_STRING);
if (o->refcount != 1 || o->encoding != OBJ_ENCODING_RAW) {
robj *decoded = getDecodedObject(o);
o = createRawStringObject(decoded->ptr, sdslen(decoded->ptr));
decrRefCount(decoded);
dbOverwrite(db,key,o);
}
return o;
}
/* Blocking RPOP/LPOP */
void blockingPopGenericCommand(client *c, int where) {
robj *o;
mstime_t timeout;
int j;
if (getTimeoutFromObjectOrReply(c,c->argv[c->argc-1],&timeout,UNIT_SECONDS)
!= C_OK) return;
for (j = 1; j < c->argc-1; j++) {
o = lookupKeyWrite(c->db,c->argv[j]);
if (o != NULL) {
if (o->type != OBJ_LIST) {
addReply(c,shared.wrongtypeerr);
return;
} else {
if (listTypeLength(o) != 0) {
/* Non empty list, this is like a non normal [LR]POP. */
char *event = (where == LIST_HEAD) ? "lpop" : "rpop";
robj *value = listTypePop(o,where);
serverAssert(value != NULL);
addReplyMultiBulkLen(c,2);
addReplyBulk(c,c->argv[j]);
addReplyBulk(c,value);
decrRefCount(value);
notifyKeyspaceEvent(NOTIFY_LIST,event,
c->argv[j],c->db->id);
if (listTypeLength(o) == 0) {
dbDelete(c->db,c->argv[j]);
notifyKeyspaceEvent(NOTIFY_GENERIC,"del",
c->argv[j],c->db->id);
}
signalModifiedKey(c->db,c->argv[j]);
server.dirty++;
/* Replicate it as an [LR]POP instead of B[LR]POP. */
rewriteClientCommandVector(c,2,
(where == LIST_HEAD) ? shared.lpop : shared.rpop,
c->argv[j]);
return;
}
}
}
}
/* If we are inside a MULTI/EXEC and the list is empty the only thing
* we can do is treating it as a timeout (even with timeout 0). */
if (c->flags & CLIENT_MULTI) {
addReply(c,shared.nullmultibulk);
return;
}
/* If the list is empty or the key does not exists we must block */
blockForKeys(c, c->argv + 1, c->argc - 2, timeout, NULL);
}
/* node can be NULL so node_name is a seperate parameter. */
shard *counterAddShard(counter *cntr, clusterNode* node, const char *node_name) {
shard *shrd = zcalloc(sizeof(shard));
shrd->node = node;
memcpy(shrd->node_name,node_name,CLUSTER_NAMELEN);
if (node == myself) {
serverAssert(cntr->myshard == NULL);
cntr->myshard = shrd;
}
listAddNodeTail(cntr->shards, shrd);
return shrd;
}
/* Create a counter and add it to server.counters. */
counter *counterCreate(sds name) {
counter *cntr = zcalloc(sizeof(counter) + (sizeof(long double) * server.history_size));
cntr->name = sdsdup(name);
cntr->shards = listCreate();
cntr->history = (long double*)(cntr + 1);
cntr->precision = server.default_precision;
serverAssert(dictAdd(server.counters, cntr->name, cntr) == DICT_OK);
return cntr;
}
/* Stores pointer to current the entry in the provided entry structure
* and advances the position of the iterator. Returns 1 when the current
* entry is in fact an entry, 0 otherwise. */
int listTypeNext(listTypeIterator *li, listTypeEntry *entry) {
/* Protect from converting when iterating */
serverAssert(li->subject->encoding == li->encoding);
entry->li = li;
if (li->encoding == OBJ_ENCODING_QUICKLIST) {
return quicklistNext(li->iter, &entry->entry);
} else {
serverPanic("Unknown list encoding");
}
return 0;
}
//将列表类型的迭代器指向的entry保存在提供的listTypeEntry结构中，并且更新迭代器，1表示成功，0失败
int listTypeNext(listTypeIterator *li, listTypeEntry *entry) {
/* Protect from converting when iterating */
//确保对象编码类型和迭代器中encoding成员相等
serverAssert(li->subject->encoding == li->encoding);
//设置listTypeEntry的entry成员关联到当前列表类型的迭代器
entry->li = li;
//对列表对象编码为quicklist类型操作
if (li->encoding == OBJ_ENCODING_QUICKLIST) {
//保存当前的entry到listTypeEntry的entry成员，并更新迭代器
return quicklistNext(li->iter, &entry->entry);
} else {
serverPanic("Unknown list encoding");
}
return 0;
}
/* The base case is to use the keys position as given in the command table
* (firstkey, lastkey, step). */
int *getKeysUsingCommandTable(struct redisCommand *cmd,robj **argv, int argc, int *numkeys) {
int j, i = 0, last, *keys;
UNUSED(argv);
if (cmd->firstkey == 0) {
*numkeys = 0;
return NULL;
}
last = cmd->lastkey;
if (last < 0) last = argc+last;
keys = zmalloc(sizeof(int)*((last - cmd->firstkey)+1));
for (j = cmd->firstkey; j <= last; j += cmd->keystep) {
serverAssert(j < argc);
keys[i++] = j;
}
*numkeys = i;
return keys;
}
/* Duplicate a string object, with the guarantee that the returned object
* has the same encoding as the original one.
*
* This function also guarantees that duplicating a small integere object
* (or a string object that contains a representation of a small integer)
* will always result in a fresh object that is unshared (refcount == 1).
*
* The resulting object always has refcount set to 1. */
robj *dupStringObject(robj *o) {
robj *d;
serverAssert(o->type == OBJ_STRING);
switch(o->encoding) {
case OBJ_ENCODING_RAW:
return createRawStringObject(o->ptr,sdslen(o->ptr));
case OBJ_ENCODING_EMBSTR:
return createEmbeddedStringObject(o->ptr,sdslen(o->ptr));
case OBJ_ENCODING_INT:
d = createObject(OBJ_STRING, NULL);
d->encoding = OBJ_ENCODING_INT;
d->ptr = o->ptr;
return d;
default:
serverPanic("Wrong encoding.");
break;
}
}
//返回 复制的o对象的副本的地址，且创建的对象非共享
robj *dupStringObject(robj *o) {
robj *d;
serverAssert(o->type == OBJ_STRING); //一定是OBJ_STRING类型
switch(o->encoding) { //根据不同的编码类型
case OBJ_ENCODING_RAW:
return createRawStringObject(o->ptr,sdslen(o->ptr)); //创建的对象非共享
case OBJ_ENCODING_EMBSTR:
return createEmbeddedStringObject(o->ptr,sdslen(o->ptr)); //创建的对象非共享
case OBJ_ENCODING_INT: //整数编码类型
d = createObject(OBJ_STRING, NULL); //即使是共享整数范围内的整数，创建的对象也是非共享的
d->encoding = OBJ_ENCODING_INT;
d->ptr = o->ptr;
return d;
default:
serverPanic("Wrong encoding.");
break;
}
}
/* This function is called in the beforeSleep() function of the event loop
* in order to process the pending input buffer of clients that were
* unblocked after a blocking operation. */
void processUnblockedClients(void) {
listNode *ln;
client *c;
while (listLength(server.unblocked_clients)) {
ln = listFirst(server.unblocked_clients);
serverAssert(ln != NULL);
c = ln->value;
listDelNode(server.unblocked_clients,ln);
c->flags &= ~CLIENT_UNBLOCKED;
/* Process remaining data in the input buffer, unless the client
* is blocked again. Actually processInputBuffer() checks that the
* client is not blocked before to proceed, but things may change and
* the code is conceptually more correct this way. */
if (!(c->flags & CLIENT_BLOCKED)) {
if (c->querybuf && sdslen(c->querybuf) > 0) {
processInputBuffer(c);
}
}
}
}
/* If the specified key has clients blocked waiting for list pushes, this
* function will put the key reference into the server.ready_keys list.
* Note that db->ready_keys is a hash table that allows us to avoid putting
* the same key again and again in the list in case of multiple pushes
* made by a script or in the context of MULTI/EXEC.
*
* The list will be finally processed by handleClientsBlockedOnLists() */
void signalListAsReady(redisDb *db, robj *key) {
readyList *rl;
/* No clients blocking for this key? No need to queue it. */
if (dictFind(db->blocking_keys,key) == NULL) return;
/* Key was already signaled? No need to queue it again. */
if (dictFind(db->ready_keys,key) != NULL) return;
/* Ok, we need to queue this key into server.ready_keys. */
rl = zmalloc(sizeof(*rl));
rl->key = key;
rl->db = db;
incrRefCount(key);
listAddNodeTail(server.ready_keys,rl);
/* We also add the key in the db->ready_keys dictionary in order
* to avoid adding it multiple times into a list with a simple O(1)
* check. */
incrRefCount(key);
serverAssert(dictAdd(db->ready_keys,key,NULL) == DICT_OK);
}
/* This function is called in the beforeSleep() function of the event loop
* in order to process the pending input buffer of clients that were
* unblocked after a blocking operation. */
void processUnblockedClients(void) {
listNode *ln;
client *c;
while (listLength(server.unblocked_clients)) {
ln = listFirst(server.unblocked_clients);
serverAssert(ln != NULL);
c = ln->value;
listDelNode(server.unblocked_clients,ln);
c->flags &= ~CLIENT_UNBLOCKED;
/* Note that the client may be blocked again at this point, since
* a new blocking command was processed. In that case we just remove
* it from the unblocked clients list without actually processing
* its pending query buffer. */
if (!(c->flags & CLIENT_BLOCKED)) {
c->btype = BLOCKED_NONE;
/* Process remaining data in the input buffer. */
if (c->querybuf && sdslen(c->querybuf) > 0) {
processInputBuffer(c);
}
}
}
}
/* Set the file-based rio object to auto-fsync every 'bytes' file written.
* By default this is set to zero that means no automatic file sync is
* performed.
*
* This feature is useful in a few contexts since when we rely on OS write
* buffers sometimes the OS buffers way too much, resulting in too many
* disk I/O concentrated in very little time. When we fsync in an explicit
* way instead the I/O pressure is more distributed across time. */
void rioSetAutoSync(rio *r, off_t bytes) {
serverAssert(r->read == rioFileIO.read);
r->io.file.autosync = bytes;
}
/* 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(client *c) {
long l;
unsigned long count, size;
int uniq = 1;
robj *set;
sds ele;
int64_t llele;
int encoding;
dict *d;
if (getLongFromObjectOrReply(c,c->argv[2],&l,NULL) != C_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,OBJ_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 == OBJ_ENCODING_INTSET) {
addReplyBulkLongLong(c,llele);
} else {
addReplyBulkCBuffer(c,ele,sdslen(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,SET_OP_UNION);
return;
}
/* For CASE 3 and CASE 4 we need an auxiliary dictionary. */
d = dictCreate(&objectKeyPointerValueDictType,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 == OBJ_ENCODING_INTSET) {
retval = dictAdd(d,createStringObjectFromLongLong(llele),NULL);
} else {
retval = dictAdd(d,createStringObject(ele,sdslen(ele)),NULL);
}
serverAssert(retval == DICT_OK);
}
setTypeReleaseIterator(si);
serverAssert(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;
robj *objele;
while(added < count) {
encoding = setTypeRandomElement(set,&ele,&llele);
if (encoding == OBJ_ENCODING_INTSET) {
objele = createStringObjectFromLongLong(llele);
} else {
objele = createStringObject(ele,sdslen(ele));
}
/* 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,objele,NULL) == DICT_OK)
added++;
else
decrRefCount(objele);
}
}
/* 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 counterWantAck(counter *cntr, const clusterNode *node) {
serverAssert(dictAdd(cntr->want_acks,(void*)node->name,NULL) == DICT_OK);
}
// BRPOP BLPOP 命令的底层实现
// BLPOP key [key ...] timeout
void blockingPopGenericCommand(client *c, int where) {
robj *o;
mstime_t timeout;
int j;
// 以秒为单位取出timeout值
if (getTimeoutFromObjectOrReply(c,c->argv[c->argc-1],&timeout,UNIT_SECONDS)
!= C_OK) return;
//遍历所有的key
for (j = 1; j < c->argc-1; j++) {
//以写操作取出当前key的值
o = lookupKeyWrite(c->db,c->argv[j]);
// value对象不为空
if (o != NULL) {
// 如果value对象的类型不是列表类型，发送类型错误信息
if (o->type != OBJ_LIST) {
addReply(c,shared.wrongtypeerr);
return;
} else {
// 列表长度不为0
if (listTypeLength(o) != 0) {
/* Non empty list, this is like a non normal [LR]POP. */
// 保存事件名称
char *event = (where == LIST_HEAD) ? "lpop" : "rpop";
// 保存弹出的value对象
robj *value = listTypePop(o,where);
serverAssert(value != NULL);
// 发送回复给client
addReplyMultiBulkLen(c,2);
addReplyBulk(c,c->argv[j]);
addReplyBulk(c,value);
// 释放value
decrRefCount(value);
// 发送事件通知
notifyKeyspaceEvent(NOTIFY_LIST,event,
c->argv[j],c->db->id);
//如果弹出元素后列表为空
if (listTypeLength(o) == 0) {
//从数据库中删除当前的key
dbDelete(c->db,c->argv[j]);
// 发送"del"的事件通知
notifyKeyspaceEvent(NOTIFY_GENERIC,"del",
c->argv[j],c->db->id);
}
//数据库的键被修改，发送信号
signalModifiedKey(c->db,c->argv[j]);
//更新脏键
server.dirty++;
/* Replicate it as an [LR]POP instead of B[LR]POP. */
// 传播一个[LR]POP 而不是B[LR]POP
rewriteClientCommandVector(c,2,
(where == LIST_HEAD) ? shared.lpop : shared.rpop,
c->argv[j]);
return;
}
}
}
}
/* If we are inside a MULTI/EXEC and the list is empty the only thing
* we can do is treating it as a timeout (even with timeout 0). */
// 如果命令在一个事务中执行，则发送一个空回复以避免死等待
if (c->flags & CLIENT_MULTI) {
addReply(c,shared.nullmultibulk);
return;
}
/* If the list is empty or the key does not exists we must block */
// 参数中的所有键都不存在，则阻塞这些键
blockForKeys(c, c->argv + 1, c->argc - 2, timeout, NULL);
}
