redict/src/ziplist.c

1322 lines
42 KiB
C
Raw Normal View History

/* The ziplist is a specially encoded dually linked list that is designed
* to be very memory efficient. It stores both strings and integer values,
* where integers are encoded as actual integers instead of a series of
* characters. It allows push and pop operations on either side of the list
* in O(1) time. However, because every operation requires a reallocation of
* the memory used by the ziplist, the actual complexity is related to the
* amount of memory used by the ziplist.
*
* ----------------------------------------------------------------------------
*
* ZIPLIST OVERALL LAYOUT:
* The general layout of the ziplist is as follows:
* <zlbytes><zltail><zllen><entry><entry><zlend>
*
* <zlbytes> is an unsigned integer to hold the number of bytes that the
* ziplist occupies. This value needs to be stored to be able to resize the
* entire structure without the need to traverse it first.
*
* <zltail> is the offset to the last entry in the list. This allows a pop
* operation on the far side of the list without the need for full traversal.
*
* <zllen> is the number of entries.When this value is larger than 2**16-2,
* we need to traverse the entire list to know how many items it holds.
*
* <zlend> is a single byte special value, equal to 255, which indicates the
* end of the list.
*
* ZIPLIST ENTRIES:
* Every entry in the ziplist is prefixed by a header that contains two pieces
* of information. First, the length of the previous entry is stored to be
* able to traverse the list from back to front. Second, the encoding with an
* optional string length of the entry itself is stored.
*
* The length of the previous entry is encoded in the following way:
* If this length is smaller than 254 bytes, it will only consume a single
* byte that takes the length as value. When the length is greater than or
* equal to 254, it will consume 5 bytes. The first byte is set to 254 to
* indicate a larger value is following. The remaining 4 bytes take the
* length of the previous entry as value.
*
* The other header field of the entry itself depends on the contents of the
* entry. When the entry is a string, the first 2 bits of this header will hold
* the type of encoding used to store the length of the string, followed by the
* actual length of the string. When the entry is an integer the first 2 bits
* are both set to 1. The following 2 bits are used to specify what kind of
* integer will be stored after this header. An overview of the different
* types and encodings is as follows:
*
* |00pppppp| - 1 byte
* String value with length less than or equal to 63 bytes (6 bits).
* |01pppppp|qqqqqqqq| - 2 bytes
* String value with length less than or equal to 16383 bytes (14 bits).
* |10______|qqqqqqqq|rrrrrrrr|ssssssss|tttttttt| - 5 bytes
* String value with length greater than or equal to 16384 bytes.
* |1100____| - 1 byte
* Integer encoded as int16_t (2 bytes).
* |1101____| - 1 byte
* Integer encoded as int32_t (4 bytes).
* |1110____| - 1 byte
* Integer encoded as int64_t (8 bytes).
*/
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <stdint.h>
#include <assert.h>
#include <limits.h>
#include "zmalloc.h"
2011-04-27 07:30:32 -04:00
#include "util.h"
#include "ziplist.h"
2011-03-09 12:49:59 -05:00
#include "endian.h"
2010-05-22 09:18:30 -04:00
#define ZIP_END 255
2010-05-22 13:35:26 -04:00
#define ZIP_BIGLEN 254
2010-05-22 09:18:30 -04:00
/* Different encoding/length possibilities */
#define ZIP_STR_06B (0 << 6)
#define ZIP_STR_14B (1 << 6)
#define ZIP_STR_32B (2 << 6)
#define ZIP_INT_16B (0xc0 | 0<<4)
#define ZIP_INT_32B (0xc0 | 1<<4)
#define ZIP_INT_64B (0xc0 | 2<<4)
2010-05-22 09:18:30 -04:00
/* Macro's to determine type */
#define ZIP_IS_STR(enc) (((enc) & 0xc0) < 0xc0)
#define ZIP_IS_INT(enc) (!ZIP_IS_STR(enc) && ((enc) & 0x30) < 0x30)
2010-05-22 09:18:30 -04:00
/* Utility macros */
#define ZIPLIST_BYTES(zl) (*((uint32_t*)(zl)))
#define ZIPLIST_TAIL_OFFSET(zl) (*((uint32_t*)((zl)+sizeof(uint32_t))))
#define ZIPLIST_LENGTH(zl) (*((uint16_t*)((zl)+sizeof(uint32_t)*2)))
#define ZIPLIST_HEADER_SIZE (sizeof(uint32_t)*2+sizeof(uint16_t))
#define ZIPLIST_ENTRY_HEAD(zl) ((zl)+ZIPLIST_HEADER_SIZE)
#define ZIPLIST_ENTRY_TAIL(zl) ((zl)+ZIPLIST_TAIL_OFFSET(zl))
#define ZIPLIST_ENTRY_END(zl) ((zl)+ZIPLIST_BYTES(zl)-1)
/* We know a positive increment can only be 1 because entries can only be
* pushed one at a time. */
2010-05-22 08:58:58 -04:00
#define ZIPLIST_INCR_LENGTH(zl,incr) { \
if (ZIPLIST_LENGTH(zl) < UINT16_MAX) ZIPLIST_LENGTH(zl)+=incr; }
typedef struct zlentry {
unsigned int prevrawlensize, prevrawlen;
unsigned int lensize, len;
unsigned int headersize;
unsigned char encoding;
unsigned char *p;
} zlentry;
/* Return the encoding pointer to by 'p'. */
static unsigned int zipEntryEncoding(unsigned char *p) {
/* String encoding: 2 MSBs */
unsigned char b = p[0] & 0xc0;
if (b < 0xc0) {
return b;
} else {
/* Integer encoding: 4 MSBs */
return p[0] & 0xf0;
}
assert(NULL);
2010-12-23 06:25:56 -05:00
return 0;
}
2010-05-22 09:18:30 -04:00
/* Return bytes needed to store integer encoded by 'encoding' */
static unsigned int zipIntSize(unsigned char encoding) {
switch(encoding) {
case ZIP_INT_16B: return sizeof(int16_t);
case ZIP_INT_32B: return sizeof(int32_t);
case ZIP_INT_64B: return sizeof(int64_t);
2010-05-22 09:18:30 -04:00
}
assert(NULL);
2010-12-23 06:25:56 -05:00
return 0;
2010-05-22 09:18:30 -04:00
}
/* Decode the encoded length pointed by 'p'. If a pointer to 'lensize' is
* provided, it is set to the number of bytes required to encode the length. */
static unsigned int zipDecodeLength(unsigned char *p, unsigned int *lensize) {
unsigned char encoding = zipEntryEncoding(p);
2010-12-23 06:25:56 -05:00
unsigned int len = 0;
2010-05-22 09:18:30 -04:00
if (ZIP_IS_STR(encoding)) {
switch(encoding) {
case ZIP_STR_06B:
len = p[0] & 0x3f;
2010-05-22 09:18:30 -04:00
if (lensize) *lensize = 1;
break;
case ZIP_STR_14B:
len = ((p[0] & 0x3f) << 8) | p[1];
if (lensize) *lensize = 2;
break;
case ZIP_STR_32B:
len = (p[1] << 24) | (p[2] << 16) | (p[3] << 8) | p[4];
2010-05-22 09:18:30 -04:00
if (lensize) *lensize = 5;
break;
default:
assert(NULL);
2010-05-22 09:18:30 -04:00
}
} else {
len = zipIntSize(encoding);
2010-05-22 09:18:30 -04:00
if (lensize) *lensize = 1;
}
return len;
}
/* Encode the length 'l' writing it in 'p'. If p is NULL it just returns
* the amount of bytes required to encode such a length. */
static unsigned int zipEncodeLength(unsigned char *p, unsigned char encoding, unsigned int rawlen) {
unsigned char len = 1, buf[5];
if (ZIP_IS_STR(encoding)) {
/* Although encoding is given it may not be set for strings,
* so we determine it here using the raw length. */
if (rawlen <= 0x3f) {
2010-05-22 09:18:30 -04:00
if (!p) return len;
buf[0] = ZIP_STR_06B | rawlen;
} else if (rawlen <= 0x3fff) {
len += 1;
2010-05-22 09:18:30 -04:00
if (!p) return len;
buf[0] = ZIP_STR_14B | ((rawlen >> 8) & 0x3f);
buf[1] = rawlen & 0xff;
2010-05-22 09:18:30 -04:00
} else {
len += 4;
if (!p) return len;
buf[0] = ZIP_STR_32B;
buf[1] = (rawlen >> 24) & 0xff;
buf[2] = (rawlen >> 16) & 0xff;
buf[3] = (rawlen >> 8) & 0xff;
buf[4] = rawlen & 0xff;
2010-05-22 09:18:30 -04:00
}
} else {
/* Implies integer encoding, so length is always 1. */
if (!p) return len;
buf[0] = encoding;
2010-05-22 09:18:30 -04:00
}
/* Store this length at p */
2010-05-22 09:18:30 -04:00
memcpy(p,buf,len);
return len;
}
/* Decode the length of the previous element stored at "p". */
static unsigned int zipPrevDecodeLength(unsigned char *p, unsigned int *lensize) {
unsigned int len = *p;
if (len < ZIP_BIGLEN) {
if (lensize) *lensize = 1;
} else {
if (lensize) *lensize = 1+sizeof(len);
memcpy(&len,p+1,sizeof(len));
2011-03-09 12:49:59 -05:00
memrev32ifbe(&len);
}
return len;
}
/* Encode the length of the previous entry and write it to "p". Return the
* number of bytes needed to encode this length if "p" is NULL. */
static unsigned int zipPrevEncodeLength(unsigned char *p, unsigned int len) {
if (p == NULL) {
return (len < ZIP_BIGLEN) ? 1 : sizeof(len)+1;
} else {
if (len < ZIP_BIGLEN) {
p[0] = len;
return 1;
} else {
p[0] = ZIP_BIGLEN;
memcpy(p+1,&len,sizeof(len));
2011-03-09 12:49:59 -05:00
memrev32ifbe(p+1);
return 1+sizeof(len);
}
}
}
/* Encode the length of the previous entry and write it to "p". This only
* uses the larger encoding (required in __ziplistCascadeUpdate). */
static void zipPrevEncodeLengthForceLarge(unsigned char *p, unsigned int len) {
if (p == NULL) return;
p[0] = ZIP_BIGLEN;
memcpy(p+1,&len,sizeof(len));
2011-03-09 12:49:59 -05:00
memrev32ifbe(p+1);
}
/* Return the difference in number of bytes needed to store the new length
* "len" on the entry pointed to by "p". */
static int zipPrevLenByteDiff(unsigned char *p, unsigned int len) {
unsigned int prevlensize;
zipPrevDecodeLength(p,&prevlensize);
return zipPrevEncodeLength(NULL,len)-prevlensize;
}
2010-05-22 09:18:30 -04:00
/* Check if string pointed to by 'entry' can be encoded as an integer.
* Stores the integer value in 'v' and its encoding in 'encoding'. */
static int zipTryEncoding(unsigned char *entry, unsigned int entrylen, long long *v, unsigned char *encoding) {
2010-05-22 09:18:30 -04:00
long long value;
if (entrylen >= 32 || entrylen == 0) return 0;
2011-04-27 07:30:32 -04:00
if (string2ll((char*)entry,entrylen,&value)) {
/* Great, the string can be encoded. Check what's the smallest
* of our encoding types that can hold this value. */
if (value >= INT16_MIN && value <= INT16_MAX) {
*encoding = ZIP_INT_16B;
} else if (value >= INT32_MIN && value <= INT32_MAX) {
*encoding = ZIP_INT_32B;
2010-05-22 09:18:30 -04:00
} else {
*encoding = ZIP_INT_64B;
2010-05-22 09:18:30 -04:00
}
*v = value;
return 1;
}
return 0;
}
/* Store integer 'value' at 'p', encoded as 'encoding' */
static void zipSaveInteger(unsigned char *p, int64_t value, unsigned char encoding) {
int16_t i16;
int32_t i32;
int64_t i64;
if (encoding == ZIP_INT_16B) {
i16 = value;
memcpy(p,&i16,sizeof(i16));
2011-03-09 12:49:59 -05:00
memrev16ifbe(p);
} else if (encoding == ZIP_INT_32B) {
i32 = value;
memcpy(p,&i32,sizeof(i32));
2011-03-09 12:49:59 -05:00
memrev32ifbe(p);
} else if (encoding == ZIP_INT_64B) {
i64 = value;
memcpy(p,&i64,sizeof(i64));
2011-03-09 12:49:59 -05:00
memrev64ifbe(p);
2010-05-22 09:18:30 -04:00
} else {
assert(NULL);
}
}
/* Read integer encoded as 'encoding' from 'p' */
static int64_t zipLoadInteger(unsigned char *p, unsigned char encoding) {
int16_t i16;
int32_t i32;
2010-12-23 06:25:56 -05:00
int64_t i64, ret = 0;
if (encoding == ZIP_INT_16B) {
memcpy(&i16,p,sizeof(i16));
2011-03-09 12:49:59 -05:00
memrev16ifbe(&i16);
ret = i16;
} else if (encoding == ZIP_INT_32B) {
memcpy(&i32,p,sizeof(i32));
2011-03-09 12:49:59 -05:00
memrev16ifbe(&i32);
ret = i32;
} else if (encoding == ZIP_INT_64B) {
memcpy(&i64,p,sizeof(i64));
2011-03-09 12:49:59 -05:00
memrev16ifbe(&i64);
ret = i64;
2010-05-22 09:18:30 -04:00
} else {
assert(NULL);
}
return ret;
}
/* Return a struct with all information about an entry. */
static zlentry zipEntry(unsigned char *p) {
zlentry e;
e.prevrawlen = zipPrevDecodeLength(p,&e.prevrawlensize);
e.len = zipDecodeLength(p+e.prevrawlensize,&e.lensize);
e.headersize = e.prevrawlensize+e.lensize;
e.encoding = zipEntryEncoding(p+e.prevrawlensize);
e.p = p;
return e;
}
/* Return the total number of bytes used by the entry at "p". */
2010-05-22 09:18:30 -04:00
static unsigned int zipRawEntryLength(unsigned char *p) {
zlentry e = zipEntry(p);
return e.headersize + e.len;
2010-05-22 09:18:30 -04:00
}
/* Create a new empty ziplist. */
unsigned char *ziplistNew(void) {
unsigned int bytes = ZIPLIST_HEADER_SIZE+1;
unsigned char *zl = zmalloc(bytes);
ZIPLIST_BYTES(zl) = bytes;
ZIPLIST_TAIL_OFFSET(zl) = ZIPLIST_HEADER_SIZE;
ZIPLIST_LENGTH(zl) = 0;
zl[bytes-1] = ZIP_END;
return zl;
}
2010-05-22 09:18:30 -04:00
/* Resize the ziplist. */
static unsigned char *ziplistResize(unsigned char *zl, unsigned int len) {
2010-05-22 09:18:30 -04:00
zl = zrealloc(zl,len);
ZIPLIST_BYTES(zl) = len;
zl[len-1] = ZIP_END;
return zl;
}
/* When an entry is inserted, we need to set the prevlen field of the next
* entry to equal the length of the inserted entry. It can occur that this
* length cannot be encoded in 1 byte and the next entry needs to be grow
* a bit larger to hold the 5-byte encoded prevlen. This can be done for free,
* because this only happens when an entry is already being inserted (which
* causes a realloc and memmove). However, encoding the prevlen may require
* that this entry is grown as well. This effect may cascade throughout
* the ziplist when there are consecutive entries with a size close to
* ZIP_BIGLEN, so we need to check that the prevlen can be encoded in every
* consecutive entry.
*
* Note that this effect can also happen in reverse, where the bytes required
* to encode the prevlen field can shrink. This effect is deliberately ignored,
* because it can cause a "flapping" effect where a chain prevlen fields is
* first grown and then shrunk again after consecutive inserts. Rather, the
* field is allowed to stay larger than necessary, because a large prevlen
* field implies the ziplist is holding large entries anyway.
*
* The pointer "p" points to the first entry that does NOT need to be
* updated, i.e. consecutive fields MAY need an update. */
static unsigned char *__ziplistCascadeUpdate(unsigned char *zl, unsigned char *p) {
2011-03-13 13:15:57 -04:00
size_t curlen = ZIPLIST_BYTES(zl), rawlen, rawlensize;
size_t offset, noffset, extra;
unsigned char *np;
zlentry cur, next;
while (p[0] != ZIP_END) {
cur = zipEntry(p);
rawlen = cur.headersize + cur.len;
rawlensize = zipPrevEncodeLength(NULL,rawlen);
/* Abort if there is no next entry. */
if (p[rawlen] == ZIP_END) break;
next = zipEntry(p+rawlen);
/* Abort when "prevlen" has not changed. */
if (next.prevrawlen == rawlen) break;
if (next.prevrawlensize < rawlensize) {
/* The "prevlen" field of "next" needs more bytes to hold
* the raw length of "cur". */
offset = p-zl;
extra = rawlensize-next.prevrawlensize;
zl = ziplistResize(zl,curlen+extra);
p = zl+offset;
/* Current pointer and offset for next element. */
np = p+rawlen;
noffset = np-zl;
/* Update tail offset when next element is not the tail element. */
if ((zl+ZIPLIST_TAIL_OFFSET(zl)) != np)
ZIPLIST_TAIL_OFFSET(zl) += extra;
/* Move the tail to the back. */
memmove(np+rawlensize,
np+next.prevrawlensize,
curlen-noffset-next.prevrawlensize-1);
zipPrevEncodeLength(np,rawlen);
/* Advance the cursor */
p += rawlen;
curlen += extra;
} else {
if (next.prevrawlensize > rawlensize) {
/* This would result in shrinking, which we want to avoid.
* So, set "rawlen" in the available bytes. */
zipPrevEncodeLengthForceLarge(p+rawlen,rawlen);
} else {
zipPrevEncodeLength(p+rawlen,rawlen);
}
/* Stop here, as the raw length of "next" has not changed. */
break;
}
}
return zl;
}
/* Delete "num" entries, starting at "p". Returns pointer to the ziplist. */
static unsigned char *__ziplistDelete(unsigned char *zl, unsigned char *p, unsigned int num) {
unsigned int i, totlen, deleted = 0;
2011-03-13 13:15:57 -04:00
size_t offset;
int nextdiff = 0;
zlentry first, tail;
first = zipEntry(p);
for (i = 0; p[0] != ZIP_END && i < num; i++) {
p += zipRawEntryLength(p);
deleted++;
}
totlen = p-first.p;
if (totlen > 0) {
if (p[0] != ZIP_END) {
/* Tricky: storing the prevlen in this entry might reduce or
* increase the number of bytes needed, compared to the current
* prevlen. Note that we can always store this length because
* it was previously stored by an entry that is being deleted. */
nextdiff = zipPrevLenByteDiff(p,first.prevrawlen);
zipPrevEncodeLength(p-nextdiff,first.prevrawlen);
/* Update offset for tail */
ZIPLIST_TAIL_OFFSET(zl) -= totlen;
/* When the tail contains more than one entry, we need to take
* "nextdiff" in account as well. Otherwise, a change in the
* size of prevlen doesn't have an effect on the *tail* offset. */
tail = zipEntry(p);
if (p[tail.headersize+tail.len] != ZIP_END)
ZIPLIST_TAIL_OFFSET(zl) += nextdiff;
/* Move tail to the front of the ziplist */
memmove(first.p,p-nextdiff,ZIPLIST_BYTES(zl)-(p-zl)-1+nextdiff);
} else {
/* The entire tail was deleted. No need to move memory. */
ZIPLIST_TAIL_OFFSET(zl) = (first.p-zl)-first.prevrawlen;
}
/* Resize and update length */
offset = first.p-zl;
zl = ziplistResize(zl, ZIPLIST_BYTES(zl)-totlen+nextdiff);
ZIPLIST_INCR_LENGTH(zl,-deleted);
p = zl+offset;
/* When nextdiff != 0, the raw length of the next entry has changed, so
* we need to cascade the update throughout the ziplist */
if (nextdiff != 0)
zl = __ziplistCascadeUpdate(zl,p);
}
return zl;
}
/* Insert item at "p". */
static unsigned char *__ziplistInsert(unsigned char *zl, unsigned char *p, unsigned char *s, unsigned int slen) {
2011-03-13 13:15:57 -04:00
size_t curlen = ZIPLIST_BYTES(zl), reqlen, prevlen = 0;
size_t offset;
int nextdiff = 0;
unsigned char encoding = 0;
long long value;
zlentry entry, tail;
/* Find out prevlen for the entry that is inserted. */
if (p[0] != ZIP_END) {
entry = zipEntry(p);
prevlen = entry.prevrawlen;
} else {
unsigned char *ptail = ZIPLIST_ENTRY_TAIL(zl);
if (ptail[0] != ZIP_END) {
prevlen = zipRawEntryLength(ptail);
}
}
/* See if the entry can be encoded */
if (zipTryEncoding(s,slen,&value,&encoding)) {
/* 'encoding' is set to the appropriate integer encoding */
reqlen = zipIntSize(encoding);
} else {
/* 'encoding' is untouched, however zipEncodeLength will use the
* string length to figure out how to encode it. */
reqlen = slen;
}
/* We need space for both the length of the previous entry and
* the length of the payload. */
reqlen += zipPrevEncodeLength(NULL,prevlen);
reqlen += zipEncodeLength(NULL,encoding,slen);
/* When the insert position is not equal to the tail, we need to
* make sure that the next entry can hold this entry's length in
* its prevlen field. */
2010-05-29 19:37:03 -04:00
nextdiff = (p[0] != ZIP_END) ? zipPrevLenByteDiff(p,reqlen) : 0;
/* Store offset because a realloc may change the address of zl. */
offset = p-zl;
zl = ziplistResize(zl,curlen+reqlen+nextdiff);
p = zl+offset;
/* Apply memory move when necessary and update tail offset. */
if (p[0] != ZIP_END) {
/* Subtract one because of the ZIP_END bytes */
memmove(p+reqlen,p-nextdiff,curlen-offset-1+nextdiff);
/* Encode this entry's raw length in the next entry. */
zipPrevEncodeLength(p+reqlen,reqlen);
/* Update offset for tail */
ZIPLIST_TAIL_OFFSET(zl) += reqlen;
/* When the tail contains more than one entry, we need to take
* "nextdiff" in account as well. Otherwise, a change in the
* size of prevlen doesn't have an effect on the *tail* offset. */
tail = zipEntry(p+reqlen);
if (p[reqlen+tail.headersize+tail.len] != ZIP_END)
ZIPLIST_TAIL_OFFSET(zl) += nextdiff;
} else {
/* This element will be the new tail. */
ZIPLIST_TAIL_OFFSET(zl) = p-zl;
}
/* When nextdiff != 0, the raw length of the next entry has changed, so
* we need to cascade the update throughout the ziplist */
if (nextdiff != 0) {
offset = p-zl;
zl = __ziplistCascadeUpdate(zl,p+reqlen);
p = zl+offset;
}
/* Write the entry */
p += zipPrevEncodeLength(p,prevlen);
p += zipEncodeLength(p,encoding,slen);
if (ZIP_IS_STR(encoding)) {
memcpy(p,s,slen);
} else {
zipSaveInteger(p,value,encoding);
}
2010-05-22 08:58:58 -04:00
ZIPLIST_INCR_LENGTH(zl,1);
return zl;
}
unsigned char *ziplistPush(unsigned char *zl, unsigned char *s, unsigned int slen, int where) {
unsigned char *p;
p = (where == ZIPLIST_HEAD) ? ZIPLIST_ENTRY_HEAD(zl) : ZIPLIST_ENTRY_END(zl);
return __ziplistInsert(zl,p,s,slen);
}
/* Returns an offset to use for iterating with ziplistNext. When the given
* index is negative, the list is traversed back to front. When the list
* doesn't contain an element at the provided index, NULL is returned. */
unsigned char *ziplistIndex(unsigned char *zl, int index) {
unsigned char *p;
zlentry entry;
if (index < 0) {
index = (-index)-1;
p = ZIPLIST_ENTRY_TAIL(zl);
if (p[0] != ZIP_END) {
entry = zipEntry(p);
while (entry.prevrawlen > 0 && index--) {
p -= entry.prevrawlen;
entry = zipEntry(p);
}
}
} else {
p = ZIPLIST_ENTRY_HEAD(zl);
while (p[0] != ZIP_END && index--) {
p += zipRawEntryLength(p);
}
2010-05-21 07:16:08 -04:00
}
2010-05-29 19:37:03 -04:00
return (p[0] == ZIP_END || index > 0) ? NULL : p;
2010-05-21 07:16:08 -04:00
}
/* Return pointer to next entry in ziplist.
*
* zl is the pointer to the ziplist
* p is the pointer to the current element
*
* The element after 'p' is returned, otherwise NULL if we are at the end. */
unsigned char *ziplistNext(unsigned char *zl, unsigned char *p) {
((void) zl);
/* "p" could be equal to ZIP_END, caused by ziplistDelete,
* and we should return NULL. Otherwise, we should return NULL
* when the *next* element is ZIP_END (there is no next entry). */
if (p[0] == ZIP_END) {
return NULL;
} else {
p = p+zipRawEntryLength(p);
return (p[0] == ZIP_END) ? NULL : p;
}
}
/* Return pointer to previous entry in ziplist. */
unsigned char *ziplistPrev(unsigned char *zl, unsigned char *p) {
zlentry entry;
/* Iterating backwards from ZIP_END should return the tail. When "p" is
* equal to the first element of the list, we're already at the head,
* and should return NULL. */
if (p[0] == ZIP_END) {
p = ZIPLIST_ENTRY_TAIL(zl);
return (p[0] == ZIP_END) ? NULL : p;
} else if (p == ZIPLIST_ENTRY_HEAD(zl)) {
return NULL;
} else {
entry = zipEntry(p);
assert(entry.prevrawlen > 0);
return p-entry.prevrawlen;
}
}
/* Get entry pointer to by 'p' and store in either 'e' or 'v' depending
* on the encoding of the entry. 'e' is always set to NULL to be able
* to find out whether the string pointer or the integer value was set.
* Return 0 if 'p' points to the end of the zipmap, 1 otherwise. */
unsigned int ziplistGet(unsigned char *p, unsigned char **sstr, unsigned int *slen, long long *sval) {
zlentry entry;
if (p == NULL || p[0] == ZIP_END) return 0;
if (sstr) *sstr = NULL;
entry = zipEntry(p);
if (ZIP_IS_STR(entry.encoding)) {
if (sstr) {
*slen = entry.len;
*sstr = p+entry.headersize;
}
} else {
if (sval) {
*sval = zipLoadInteger(p+entry.headersize,entry.encoding);
}
2010-05-21 07:16:08 -04:00
}
return 1;
2010-05-21 07:16:08 -04:00
}
/* Insert an entry at "p". */
unsigned char *ziplistInsert(unsigned char *zl, unsigned char *p, unsigned char *s, unsigned int slen) {
return __ziplistInsert(zl,p,s,slen);
}
/* Delete a single entry from the ziplist, pointed to by *p.
* Also update *p in place, to be able to iterate over the
* ziplist, while deleting entries. */
unsigned char *ziplistDelete(unsigned char *zl, unsigned char **p) {
2011-03-13 13:15:57 -04:00
size_t offset = *p-zl;
zl = __ziplistDelete(zl,*p,1);
/* Store pointer to current element in p, because ziplistDelete will
* do a realloc which might result in a different "zl"-pointer.
* When the delete direction is back to front, we might delete the last
* entry and end up with "p" pointing to ZIP_END, so check this. */
*p = zl+offset;
return zl;
}
/* Delete a range of entries from the ziplist. */
unsigned char *ziplistDeleteRange(unsigned char *zl, unsigned int index, unsigned int num) {
unsigned char *p = ziplistIndex(zl,index);
return (p == NULL) ? zl : __ziplistDelete(zl,p,num);
}
/* Compare entry pointer to by 'p' with 'entry'. Return 1 if equal. */
unsigned int ziplistCompare(unsigned char *p, unsigned char *sstr, unsigned int slen) {
zlentry entry;
unsigned char sencoding;
long long zval, sval;
2010-05-29 19:37:03 -04:00
if (p[0] == ZIP_END) return 0;
entry = zipEntry(p);
if (ZIP_IS_STR(entry.encoding)) {
/* Raw compare */
if (entry.len == slen) {
return memcmp(p+entry.headersize,sstr,slen) == 0;
} else {
return 0;
}
} else {
/* Try to compare encoded values */
if (zipTryEncoding(sstr,slen,&sval,&sencoding)) {
if (entry.encoding == sencoding) {
zval = zipLoadInteger(p+entry.headersize,entry.encoding);
return zval == sval;
}
}
}
return 0;
}
/* Return length of ziplist. */
unsigned int ziplistLen(unsigned char *zl) {
unsigned int len = 0;
if (ZIPLIST_LENGTH(zl) < UINT16_MAX) {
len = ZIPLIST_LENGTH(zl);
} else {
unsigned char *p = zl+ZIPLIST_HEADER_SIZE;
while (*p != ZIP_END) {
p += zipRawEntryLength(p);
len++;
}
/* Re-store length if small enough */
if (len < UINT16_MAX) ZIPLIST_LENGTH(zl) = len;
}
return len;
}
2011-02-28 08:48:49 -05:00
/* Return ziplist blob size in bytes. */
size_t ziplistBlobLen(unsigned char *zl) {
return ZIPLIST_BYTES(zl);
}
void ziplistRepr(unsigned char *zl) {
unsigned char *p;
int index = 0;
zlentry entry;
printf(
"{total bytes %d} "
"{length %u}\n"
"{tail offset %u}\n",
ZIPLIST_BYTES(zl),
ZIPLIST_LENGTH(zl),
ZIPLIST_TAIL_OFFSET(zl));
p = ZIPLIST_ENTRY_HEAD(zl);
while(*p != ZIP_END) {
entry = zipEntry(p);
printf(
"{"
"addr 0x%08lx, "
"index %2d, "
"offset %5ld, "
"rl: %5u, "
"hs %2u, "
"pl: %5u, "
"pls: %2u, "
"payload %5u"
"} ",
(long unsigned)p,
index,
(unsigned long) (p-zl),
entry.headersize+entry.len,
entry.headersize,
entry.prevrawlen,
entry.prevrawlensize,
entry.len);
p += entry.headersize;
if (ZIP_IS_STR(entry.encoding)) {
if (entry.len > 40) {
if (fwrite(p,40,1,stdout) == 0) perror("fwrite");
printf("...");
} else {
if (entry.len &&
fwrite(p,entry.len,1,stdout) == 0) perror("fwrite");
}
} else {
printf("%lld", (long long) zipLoadInteger(p,entry.encoding));
}
printf("\n");
p += entry.len;
index++;
}
printf("{end}\n\n");
}
#ifdef ZIPLIST_TEST_MAIN
#include <sys/time.h>
#include "adlist.h"
#include "sds.h"
#define debug(f, ...) { if (DEBUG) printf(f, __VA_ARGS__); }
2010-05-21 07:16:08 -04:00
unsigned char *createList() {
unsigned char *zl = ziplistNew();
zl = ziplistPush(zl, (unsigned char*)"foo", 3, ZIPLIST_TAIL);
zl = ziplistPush(zl, (unsigned char*)"quux", 4, ZIPLIST_TAIL);
zl = ziplistPush(zl, (unsigned char*)"hello", 5, ZIPLIST_HEAD);
zl = ziplistPush(zl, (unsigned char*)"1024", 4, ZIPLIST_TAIL);
2010-05-21 07:16:08 -04:00
return zl;
}
unsigned char *createIntList() {
unsigned char *zl = ziplistNew();
char buf[32];
sprintf(buf, "100");
zl = ziplistPush(zl, (unsigned char*)buf, strlen(buf), ZIPLIST_TAIL);
sprintf(buf, "128000");
zl = ziplistPush(zl, (unsigned char*)buf, strlen(buf), ZIPLIST_TAIL);
sprintf(buf, "-100");
zl = ziplistPush(zl, (unsigned char*)buf, strlen(buf), ZIPLIST_HEAD);
sprintf(buf, "4294967296");
zl = ziplistPush(zl, (unsigned char*)buf, strlen(buf), ZIPLIST_HEAD);
sprintf(buf, "non integer");
zl = ziplistPush(zl, (unsigned char*)buf, strlen(buf), ZIPLIST_TAIL);
sprintf(buf, "much much longer non integer");
zl = ziplistPush(zl, (unsigned char*)buf, strlen(buf), ZIPLIST_TAIL);
return zl;
}
long long usec(void) {
struct timeval tv;
gettimeofday(&tv,NULL);
return (((long long)tv.tv_sec)*1000000)+tv.tv_usec;
}
void stress(int pos, int num, int maxsize, int dnum) {
int i,j,k;
unsigned char *zl;
char posstr[2][5] = { "HEAD", "TAIL" };
long long start;
for (i = 0; i < maxsize; i+=dnum) {
zl = ziplistNew();
for (j = 0; j < i; j++) {
zl = ziplistPush(zl,(unsigned char*)"quux",4,ZIPLIST_TAIL);
}
/* Do num times a push+pop from pos */
start = usec();
for (k = 0; k < num; k++) {
zl = ziplistPush(zl,(unsigned char*)"quux",4,pos);
zl = ziplistDeleteRange(zl,0,1);
}
printf("List size: %8d, bytes: %8d, %dx push+pop (%s): %6lld usec\n",
i,ZIPLIST_BYTES(zl),num,posstr[pos],usec()-start);
zfree(zl);
}
}
void pop(unsigned char *zl, int where) {
unsigned char *p, *vstr;
unsigned int vlen;
long long vlong;
p = ziplistIndex(zl,where == ZIPLIST_HEAD ? 0 : -1);
if (ziplistGet(p,&vstr,&vlen,&vlong)) {
if (where == ZIPLIST_HEAD)
printf("Pop head: ");
else
printf("Pop tail: ");
if (vstr)
if (vlen && fwrite(vstr,vlen,1,stdout) == 0) perror("fwrite");
else
printf("%lld", vlong);
printf("\n");
ziplistDeleteRange(zl,-1,1);
} else {
printf("ERROR: Could not pop\n");
exit(1);
}
}
int randstring(char *target, unsigned int min, unsigned int max) {
int p, len = min+rand()%(max-min+1);
int minval, maxval;
switch(rand() % 3) {
case 0:
minval = 0;
maxval = 255;
break;
case 1:
minval = 48;
maxval = 122;
break;
case 2:
minval = 48;
maxval = 52;
break;
default:
assert(NULL);
}
while(p < len)
target[p++] = minval+rand()%(maxval-minval+1);
return len;
}
2010-05-21 07:16:08 -04:00
int main(int argc, char **argv) {
2010-05-29 13:34:14 -04:00
unsigned char *zl, *p;
unsigned char *entry;
unsigned int elen;
long long value;
2010-05-21 07:16:08 -04:00
/* If an argument is given, use it as the random seed. */
if (argc == 2)
srand(atoi(argv[1]));
zl = createIntList();
ziplistRepr(zl);
2010-05-21 07:16:08 -04:00
zl = createList();
ziplistRepr(zl);
pop(zl,ZIPLIST_TAIL);
ziplistRepr(zl);
pop(zl,ZIPLIST_HEAD);
ziplistRepr(zl);
pop(zl,ZIPLIST_TAIL);
ziplistRepr(zl);
pop(zl,ZIPLIST_TAIL);
ziplistRepr(zl);
printf("Get element at index 3:\n");
{
zl = createList();
p = ziplistIndex(zl, 3);
if (!ziplistGet(p, &entry, &elen, &value)) {
printf("ERROR: Could not access index 3\n");
return 1;
}
if (entry) {
if (elen && fwrite(entry,elen,1,stdout) == 0) perror("fwrite");
printf("\n");
} else {
printf("%lld\n", value);
}
printf("\n");
}
printf("Get element at index 4 (out of range):\n");
{
zl = createList();
p = ziplistIndex(zl, 4);
if (p == NULL) {
printf("No entry\n");
} else {
printf("ERROR: Out of range index should return NULL, returned offset: %ld\n", p-zl);
return 1;
}
printf("\n");
}
printf("Get element at index -1 (last element):\n");
{
zl = createList();
p = ziplistIndex(zl, -1);
if (!ziplistGet(p, &entry, &elen, &value)) {
printf("ERROR: Could not access index -1\n");
return 1;
}
if (entry) {
if (elen && fwrite(entry,elen,1,stdout) == 0) perror("fwrite");
printf("\n");
} else {
printf("%lld\n", value);
}
printf("\n");
}
printf("Get element at index -4 (first element):\n");
{
zl = createList();
p = ziplistIndex(zl, -4);
if (!ziplistGet(p, &entry, &elen, &value)) {
printf("ERROR: Could not access index -4\n");
return 1;
}
if (entry) {
if (elen && fwrite(entry,elen,1,stdout) == 0) perror("fwrite");
printf("\n");
} else {
printf("%lld\n", value);
}
printf("\n");
}
printf("Get element at index -5 (reverse out of range):\n");
{
zl = createList();
p = ziplistIndex(zl, -5);
if (p == NULL) {
printf("No entry\n");
} else {
printf("ERROR: Out of range index should return NULL, returned offset: %ld\n", p-zl);
return 1;
}
printf("\n");
}
2010-05-21 07:16:08 -04:00
printf("Iterate list from 0 to end:\n");
{
zl = createList();
p = ziplistIndex(zl, 0);
while (ziplistGet(p, &entry, &elen, &value)) {
printf("Entry: ");
if (entry) {
if (elen && fwrite(entry,elen,1,stdout) == 0) perror("fwrite");
} else {
printf("%lld", value);
}
p = ziplistNext(zl,p);
printf("\n");
2010-05-21 07:16:08 -04:00
}
printf("\n");
}
printf("Iterate list from 1 to end:\n");
{
zl = createList();
p = ziplistIndex(zl, 1);
while (ziplistGet(p, &entry, &elen, &value)) {
printf("Entry: ");
if (entry) {
if (elen && fwrite(entry,elen,1,stdout) == 0) perror("fwrite");
} else {
printf("%lld", value);
}
p = ziplistNext(zl,p);
printf("\n");
2010-05-21 07:16:08 -04:00
}
printf("\n");
}
printf("Iterate list from 2 to end:\n");
{
zl = createList();
p = ziplistIndex(zl, 2);
while (ziplistGet(p, &entry, &elen, &value)) {
printf("Entry: ");
if (entry) {
if (elen && fwrite(entry,elen,1,stdout) == 0) perror("fwrite");
} else {
printf("%lld", value);
}
p = ziplistNext(zl,p);
printf("\n");
2010-05-21 07:16:08 -04:00
}
printf("\n");
}
printf("Iterate starting out of range:\n");
{
zl = createList();
p = ziplistIndex(zl, 4);
if (!ziplistGet(p, &entry, &elen, &value)) {
2010-05-21 07:16:08 -04:00
printf("No entry\n");
} else {
printf("ERROR\n");
}
printf("\n");
}
printf("Iterate from back to front:\n");
{
zl = createList();
p = ziplistIndex(zl, -1);
while (ziplistGet(p, &entry, &elen, &value)) {
printf("Entry: ");
if (entry) {
if (elen && fwrite(entry,elen,1,stdout) == 0) perror("fwrite");
} else {
printf("%lld", value);
}
p = ziplistPrev(zl,p);
printf("\n");
}
printf("\n");
}
printf("Iterate from back to front, deleting all items:\n");
{
zl = createList();
p = ziplistIndex(zl, -1);
while (ziplistGet(p, &entry, &elen, &value)) {
printf("Entry: ");
if (entry) {
if (elen && fwrite(entry,elen,1,stdout) == 0) perror("fwrite");
} else {
printf("%lld", value);
}
zl = ziplistDelete(zl,&p);
p = ziplistPrev(zl,p);
printf("\n");
}
printf("\n");
}
printf("Delete inclusive range 0,0:\n");
{
zl = createList();
zl = ziplistDeleteRange(zl, 0, 1);
ziplistRepr(zl);
}
printf("Delete inclusive range 0,1:\n");
{
zl = createList();
zl = ziplistDeleteRange(zl, 0, 2);
ziplistRepr(zl);
}
printf("Delete inclusive range 1,2:\n");
{
zl = createList();
zl = ziplistDeleteRange(zl, 1, 2);
ziplistRepr(zl);
}
printf("Delete with start index out of range:\n");
{
zl = createList();
zl = ziplistDeleteRange(zl, 5, 1);
ziplistRepr(zl);
}
printf("Delete with num overflow:\n");
{
zl = createList();
zl = ziplistDeleteRange(zl, 1, 5);
ziplistRepr(zl);
2010-05-21 07:16:08 -04:00
}
printf("Delete foo while iterating:\n");
{
zl = createList();
p = ziplistIndex(zl,0);
while (ziplistGet(p,&entry,&elen,&value)) {
if (entry && strncmp("foo",(char*)entry,elen) == 0) {
printf("Delete foo\n");
zl = ziplistDelete(zl,&p);
} else {
printf("Entry: ");
if (entry) {
if (elen && fwrite(entry,elen,1,stdout) == 0)
perror("fwrite");
} else {
printf("%lld",value);
}
p = ziplistNext(zl,p);
printf("\n");
}
}
printf("\n");
ziplistRepr(zl);
}
printf("Regression test for >255 byte strings:\n");
{
char v1[257],v2[257];
memset(v1,'x',256);
memset(v2,'y',256);
zl = ziplistNew();
zl = ziplistPush(zl,(unsigned char*)v1,strlen(v1),ZIPLIST_TAIL);
zl = ziplistPush(zl,(unsigned char*)v2,strlen(v2),ZIPLIST_TAIL);
/* Pop values again and compare their value. */
p = ziplistIndex(zl,0);
assert(ziplistGet(p,&entry,&elen,&value));
assert(strncmp(v1,(char*)entry,elen) == 0);
p = ziplistIndex(zl,1);
assert(ziplistGet(p,&entry,&elen,&value));
assert(strncmp(v2,(char*)entry,elen) == 0);
printf("SUCCESS\n\n");
}
printf("Create long list and check indices:\n");
{
zl = ziplistNew();
char buf[32];
int i,len;
for (i = 0; i < 1000; i++) {
len = sprintf(buf,"%d",i);
zl = ziplistPush(zl,(unsigned char*)buf,len,ZIPLIST_TAIL);
}
for (i = 0; i < 1000; i++) {
p = ziplistIndex(zl,i);
assert(ziplistGet(p,NULL,NULL,&value));
assert(i == value);
p = ziplistIndex(zl,-i-1);
assert(ziplistGet(p,NULL,NULL,&value));
assert(999-i == value);
}
printf("SUCCESS\n\n");
}
printf("Compare strings with ziplist entries:\n");
{
zl = createList();
p = ziplistIndex(zl,0);
if (!ziplistCompare(p,(unsigned char*)"hello",5)) {
printf("ERROR: not \"hello\"\n");
2010-05-29 13:34:14 -04:00
return 1;
}
if (ziplistCompare(p,(unsigned char*)"hella",5)) {
printf("ERROR: \"hella\"\n");
2010-05-29 13:34:14 -04:00
return 1;
}
p = ziplistIndex(zl,3);
if (!ziplistCompare(p,(unsigned char*)"1024",4)) {
printf("ERROR: not \"1024\"\n");
2010-05-29 13:34:14 -04:00
return 1;
}
if (ziplistCompare(p,(unsigned char*)"1025",4)) {
printf("ERROR: \"1025\"\n");
2010-05-29 13:34:14 -04:00
return 1;
}
printf("SUCCESS\n\n");
}
printf("Stress with random payloads of different encoding:\n");
{
int i,j,len,where;
unsigned char *p;
char buf[1024];
int buflen;
list *ref;
listNode *refnode;
/* Hold temp vars from ziplist */
unsigned char *sstr;
unsigned int slen;
long long sval;
for (i = 0; i < 20000; i++) {
zl = ziplistNew();
ref = listCreate();
listSetFreeMethod(ref,sdsfree);
len = rand() % 256;
/* Create lists */
for (j = 0; j < len; j++) {
where = (rand() & 1) ? ZIPLIST_HEAD : ZIPLIST_TAIL;
if (rand() % 2) {
buflen = randstring(buf,1,sizeof(buf)-1);
} else {
switch(rand() % 3) {
case 0:
buflen = sprintf(buf,"%lld",(0LL + rand()) >> 20);
break;
case 1:
buflen = sprintf(buf,"%lld",(0LL + rand()));
break;
case 2:
buflen = sprintf(buf,"%lld",(0LL + rand()) << 20);
break;
default:
assert(NULL);
}
}
/* Add to ziplist */
zl = ziplistPush(zl, (unsigned char*)buf, buflen, where);
/* Add to reference list */
if (where == ZIPLIST_HEAD) {
listAddNodeHead(ref,sdsnewlen(buf, buflen));
} else if (where == ZIPLIST_TAIL) {
listAddNodeTail(ref,sdsnewlen(buf, buflen));
} else {
assert(NULL);
}
}
assert(listLength(ref) == ziplistLen(zl));
for (j = 0; j < len; j++) {
/* Naive way to get elements, but similar to the stresser
* executed from the Tcl test suite. */
p = ziplistIndex(zl,j);
refnode = listIndex(ref,j);
assert(ziplistGet(p,&sstr,&slen,&sval));
if (sstr == NULL) {
buflen = sprintf(buf,"%lld",sval);
} else {
buflen = slen;
memcpy(buf,sstr,buflen);
buf[buflen] = '\0';
}
assert(memcmp(buf,listNodeValue(refnode),buflen) == 0);
}
zfree(zl);
listRelease(ref);
}
printf("SUCCESS\n\n");
}
printf("Stress with variable ziplist size:\n");
{
stress(ZIPLIST_HEAD,100000,16384,256);
stress(ZIPLIST_TAIL,100000,16384,256);
}
return 0;
}
#endif