mirrors
/
sqlite
mirror of https://github.com/sqlite/sqlite
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

Enhance the ANALYZE command so that it gathers statistics in the sqlite_stat1 

table even for tables that are empty or have no indices.

FossilOrigin-Name: a7645d293801da64a7579737d0a8b48117af2e2c
This commit is contained in:
drh 2010-09-25 22:32:56 +00:00
parent 42825cd215
commit 155640552b
11 changed files with 114 additions and 67 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

40
manifest
View File

@ -1,5 +1,8 @@
C Add\snew\sfile\se_createtable.test.
D 2010-09-25T17:29:58
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
C Enhance\sthe\sANALYZE\scommand\sso\sthat\sit\sgathers\sstatistics\sin\sthe\ssqlite_stat1\ntable\seven\sfor\stables\sthat\sare\sempty\sor\shave\sno\sindices.
D 2010-09-25T22:32:56
F Makefile.arm-wince-mingw32ce-gcc d6df77f1f48d690bd73162294bbba7f59507c72f
F Makefile.in c599a15d268b1db2aeadea19df2adc3bf2eb6bee
F Makefile.linux-gcc 91d710bdc4998cb015f39edf3cb314ec4f4d7e23
@ -112,7 +115,7 @@ F sqlite.pc.in 42b7bf0d02e08b9e77734a47798d1a55a9e0716b
F sqlite3.1 6be1ad09113570e1fc8dcaff84c9b0b337db5ffc
F sqlite3.pc.in ae6f59a76e862f5c561eb32a380228a02afc3cad
F src/alter.c 8dc27638e7e2553e80b2b621f232be5eb1e85ef3
F src/analyze.c da65ce99bb159b10e85a1e460adbe53a88062500
F src/analyze.c df69779a9bad368de5f5a32b8c733282d8dcbbbe
F src/attach.c c689d516ee8cc52bf11bef2067d76eb8b716228a
F src/auth.c 523da7fb4979469955d822ff9298352d6b31de34
F src/backup.c d5b0137bc20327af08c14772227cc35134839c30
@ -121,7 +124,7 @@ F src/btmutex.c 96a12f50f7a17475155971a241d85ec5171573ff
F src/btree.c d878577184112d982d00ea05afcc7487cd9f06f5
F src/btree.h 2d1a83ad509047e8cc314fda7e054f99ff52414d
F src/btreeInt.h c424f2f131cc61ddf130f9bd736b3df12c8a51f0
F src/build.c 5acc8a7d79ca81102a5d020fbafb7a4162f96d1d
F src/build.c 99894b89b7943b6f8337757facc4e61cef1d9508
F src/callback.c a1d1b1c9c85415dff013af033e2fed9c8382d33b
F src/complete.c dc1d136c0feee03c2f7550bafc0d29075e36deac
F src/ctime.c 4f3aadad62c6c9f0d4e5a96718516ac4e3c598df
@ -177,7 +180,7 @@ F src/select.c b0b124781474e4e0c8f64022875e5e2009e13443
F src/shell.c 8517fc1f9c59ae4007e6cc8b9af91ab231ea2056
F src/sqlite.h.in dae3f74d7b2b516967ede39b8e503718b571d9da
F src/sqlite3ext.h 69dfb8116af51b84a029cddb3b35062354270c89
F src/sqliteInt.h ef7ed8746759c09edd87e5550de973a4da3e4ca7
F src/sqliteInt.h d8ae67308fec3f2abdce817a630aceae42ad8043
F src/sqliteLimit.h a17dcd3fb775d63b64a43a55c54cb282f9726f44
F src/status.c 496913d4e8441195f6f2a75b1c95993a45b9b30b
F src/table.c 2cd62736f845d82200acfa1287e33feb3c15d62e
@ -237,7 +240,7 @@ F src/vtab.c 0e8e0cb30dffb078367e843e84e37ef99236c7e4
F src/wal.c 7081f148cb52b0cf2280e6384196402dc58130a3
F src/wal.h 96669b645e27cd5a111ba59f0cae7743a207bc3c
F src/walker.c 3112bb3afe1d85dc52317cb1d752055e9a781f8f
F src/where.c a5040c004496d456761e8f10750f648bbd84e982
F src/where.c 34c733f9e8ca5e8f611958422e7cc236bc9cf739
F test/aggerror.test a867e273ef9e3d7919f03ef4f0e8c0d2767944f2
F test/alias.test 4529fbc152f190268a15f9384a5651bbbabc9d87
F test/all.test 6745008c144bd2956d58864d21f7b304689c1cce
@ -246,7 +249,7 @@ F test/alter2.test 75f731508f1bf27ba09a6075c66cd02216ba464b
F test/alter3.test 8677e48d95536f7a6ed86a1a774744dadcc22b07
F test/alter4.test 1e5dd6b951e9f65ca66422edff02e56df82dd403
F test/altermalloc.test e81ac9657ed25c6c5bb09bebfa5a047cd8e4acfc
F test/analyze.test bf692e7db414f268a136bade16c03a1bdbb9240c
F test/analyze.test c1eb87067fc16ece7c07e823d6395fd831b270c5
F test/analyze2.test 59dac6c399c0c5d1a90a11ee7cc606743fb6db93
F test/analyze3.test 6d4f4b0929545a9d1af803a0608a0c51b92a3537
F test/async.test ad4ba51b77cd118911a3fe1356b0809da9c108c3
@ -258,7 +261,7 @@ F test/attach.test ce9660e51768fab93cf129787be886c5d6c4fd81
F test/attach2.test a295d2d7061adcee5884ef4a93c7c96a82765437
F test/attach3.test bd9830bc3a0d22ed1310c9bff6896927937017dc
F test/attachmalloc.test 38d2da5fdaf09ba0add57296967a3061e5842584
F test/auth.test 8f21c160a4562f54f27618e85bac869efcecbcaf
F test/auth.test 26cc6f219580191539bf335abe03e55e49310846
F test/auth2.test 270baddc8b9c273682760cffba6739d907bd2882
F test/auth3.test a4755e6a2a2fea547ffe63c874eb569e60a28eb5
F test/autoinc.test 85ef3180a737e6580086a018c09c6f1a52759b46
@ -543,7 +546,7 @@ F test/minmax3.test 66a60eb0f20281b0753249d347c5de0766954cee
F test/misc1.test e56baf44656dd68d6475a4b44521045a60241e9b
F test/misc2.test a628db7b03e18973e5d446c67696b03de718c9fd
F test/misc3.test 72c5dc87a78e7865c5ec7a969fc572913dbe96b6
F test/misc4.test 91e8ed25c092c2bb4e0bb01864631e2930f8d7de
F test/misc4.test 9c078510fbfff05a9869a0b6d8b86a623ad2c4f6
F test/misc5.test 45b2e3ed5f79af2b4f38ae362eaf4c49674575bd
F test/misc6.test 953cc693924d88e6117aeba16f46f0bf5abede91
F test/misc7.test c5f4e6a82e04e71820c0f9f64f6733f04c8ae0ae
@ -600,7 +603,7 @@ F test/select2.test 352480e0e9c66eda9c3044e412abdf5be0215b56
F test/select3.test 2ce595f8fb8e2ac10071d3b4e424cadd4634a054
F test/select4.test 44aa6e7110592e18110b0b9cf5c024d37d23be17
F test/select5.test e758b8ef94f69b111df4cb819008856655dcd535
F test/select6.test 2b5e8500d8ec3dd4c8e0c99eb1431b3d11fcc24c
F test/select6.test cc25a8650cf9a4d4f74e586c45a75f9836516b18
F test/select7.test dad6f00f0d49728a879d6eb6451d4752db0b0abe
F test/select8.test 391de11bdd52339c30580dabbbbe97e3e9a3c79d
F test/select9.test 74c0fb2c6eecb0219cbed0cbe3df136f8fbf9343
@ -736,7 +739,7 @@ F test/tkt35xx.test ed9721bd9eb1693b3b4d3cf2a093fa7f92af0c93
F test/tkt3630.test 929f64852103054125200bc825c316d5f75d42f7
F test/tkt3718.test 3b59dcb5c4e7754dacd91e7fd353a61492cc402a
F test/tkt3731.test 0c5f4cbffe102d43c3b2188af91a9e36348f974b
F test/tkt3757.test 8f2208930655bbd4f92c14e19e72303a43e098ef
F test/tkt3757.test 10cd679a88675c880533083fc79ac04324525595
F test/tkt3761.test b95ea9c98f21cf91325f18a984887e62caceab33
F test/tkt3762.test 2a9f3b03df44ec49ec0cfa8d5da6574c2a7853df
F test/tkt3773.test 430b06567ce40285dfd2c4834a2a61816403efeb
@ -867,7 +870,14 @@ F tool/speedtest2.tcl ee2149167303ba8e95af97873c575c3e0fab58ff
F tool/speedtest8.c 2902c46588c40b55661e471d7a86e4dd71a18224
F tool/speedtest8inst1.c 293327bc76823f473684d589a8160bde1f52c14e
F tool/vdbe-compress.tcl d70ea6d8a19e3571d7ab8c9b75cba86d1173ff0f
P 44deaaefeeb95827daeaf84aa5e205b456e75b40
R f9cdaa74878a8325422c228d4f8c7dbc
U dan
Z 39756bc936cfb3dbb51046b9bdefa610
P 20e16fef55c355a1d7e97d0c390769b941e83fdb
R 96b44d864f75b35f18ba4c7ab3226d4c
U drh
Z e1547e0aba5840882b569e70505d3edd
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.6 (GNU/Linux)
iD8DBQFMnnieoxKgR168RlERAs7CAJ9XFT8WTA9gYf6noiCftlq/XsWqugCfaGoh
P4TBzZ9lm4AFS3sK2f1UkgA=
=phK5
-----END PGP SIGNATURE-----

2
manifest.uuid
View File

@ -1 +1 @@
20e16fef55c355a1d7e97d0c390769b941e83fdb
a7645d293801da64a7579737d0a8b48117af2e2c

79
src/analyze.c
View File

@ -114,6 +114,7 @@ static void analyzeOneTable(
int topOfLoop; /* The top of the loop */
int endOfLoop; /* The end of the loop */
int addr; /* The address of an instruction */
int jZeroRows = 0; /* Jump from here if number of rows is zero */
int iDb; /* Index of database containing pTab */
int regTabname = iMem++; /* Register containing table name */
int regIdxname = iMem++; /* Register containing index name */
@ -132,8 +133,15 @@ static void analyzeOneTable(
#endif
v = sqlite3GetVdbe(pParse);
if( v==0 || NEVER(pTab==0) || pTab->pIndex==0 ){
/* Do no analysis for tables that have no indices */
if( v==0 || NEVER(pTab==0) ){
return;
}
if( pTab->pSelect ){
/* Do not gather statistics on views */
return;
}
if( memcmp(pTab->zName, "sqlite_", 7)==0 ){
/* Do not gather statistics on system tables */
return;
}
assert( sqlite3BtreeHoldsAllMutexes(db) );
@ -150,6 +158,7 @@ static void analyzeOneTable(
sqlite3TableLock(pParse, iDb, pTab->tnum, 0, pTab->zName);
iIdxCur = pParse->nTab++;
sqlite3VdbeAddOp4(v, OP_String8, 0, regTabname, 0, pTab->zName, 0);
for(pIdx=pTab->pIndex; pIdx; pIdx=pIdx->pNext){
int nCol = pIdx->nColumn;
KeyInfo *pKey = sqlite3IndexKeyinfo(pParse, pIdx);
@ -164,10 +173,7 @@ static void analyzeOneTable(
(char *)pKey, P4_KEYINFO_HANDOFF);
VdbeComment((v, "%s", pIdx->zName));
/* Populate the registers containing the table and index names. */
if( pTab->pIndex==pIdx ){
sqlite3VdbeAddOp4(v, OP_String8, 0, regTabname, 0, pTab->zName, 0);
}
/* Populate the register containing the index name. */
sqlite3VdbeAddOp4(v, OP_String8, 0, regIdxname, 0, pIdx->zName, 0);
#ifdef SQLITE_ENABLE_STAT2
@ -302,8 +308,10 @@ static void analyzeOneTable(
** If K>0 then it is always the case the D>0 so division by zero
** is never possible.
*/
addr = sqlite3VdbeAddOp1(v, OP_IfNot, iMem);
sqlite3VdbeAddOp2(v, OP_SCopy, iMem, regSampleno);
if( jZeroRows==0 ){
jZeroRows = sqlite3VdbeAddOp1(v, OP_IfNot, iMem);
}
for(i=0; i<nCol; i++){
sqlite3VdbeAddOp4(v, OP_String8, 0, regTemp, 0, " ", 0);
sqlite3VdbeAddOp3(v, OP_Concat, regTemp, regSampleno, regSampleno);
@ -317,13 +325,35 @@ static void analyzeOneTable(
sqlite3VdbeAddOp2(v, OP_NewRowid, iStatCur, regRowid);
sqlite3VdbeAddOp3(v, OP_Insert, iStatCur, regRec, regRowid);
sqlite3VdbeChangeP5(v, OPFLAG_APPEND);
}
/* If the table has no indices, create a single sqlite_stat1 entry
** containing NULL as the index name and the row count as the content.
*/
if( pTab->pIndex==0 ){
sqlite3VdbeAddOp3(v, OP_OpenRead, iIdxCur, pTab->tnum, iDb);
VdbeComment((v, "%s", pTab->zName));
sqlite3VdbeAddOp2(v, OP_Count, iIdxCur, regSampleno);
sqlite3VdbeAddOp1(v, OP_Close, iIdxCur);
}else{
assert( jZeroRows>0 );
addr = sqlite3VdbeAddOp0(v, OP_Goto);
sqlite3VdbeJumpHere(v, jZeroRows);
}
sqlite3VdbeAddOp2(v, OP_Null, 0, regIdxname);
sqlite3VdbeAddOp4(v, OP_MakeRecord, regTabname, 3, regRec, "aaa", 0);
sqlite3VdbeAddOp2(v, OP_NewRowid, iStatCur, regRowid);
sqlite3VdbeAddOp3(v, OP_Insert, iStatCur, regRec, regRowid);
sqlite3VdbeChangeP5(v, OPFLAG_APPEND);
if( pParse->nMem<regRec ) pParse->nMem = regRec;
if( jZeroRows ){
sqlite3VdbeJumpHere(v, addr);
}
}
/*
** Generate code that will cause the most recent index analysis to
** be laoded into internal hash tables where is can be used.
** be loaded into internal hash tables where is can be used.
*/
static void loadAnalysis(Parse *pParse, int iDb){
Vdbe *v = sqlite3GetVdbe(pParse);
@ -453,33 +483,46 @@ struct analysisInfo {
** This callback is invoked once for each index when reading the
** sqlite_stat1 table.
**
** argv[0] = name of the index
** argv[1] = results of analysis - on integer for each column
** argv[0] = name of the table
** argv[1] = name of the index (might be NULL)
** argv[2] = results of analysis - on integer for each column
**
** Entries for which argv[1]==NULL simply record the number of rows in
** the table.
*/
static int analysisLoader(void *pData, int argc, char **argv, char **NotUsed){
analysisInfo *pInfo = (analysisInfo*)pData;
Index *pIndex;
int i, c;
Table *pTable;
int i, c, n;
unsigned int v;
const char *z;
assert( argc==2 );
assert( argc==3 );
UNUSED_PARAMETER2(NotUsed, argc);
if( argv==0 || argv[0]==0 || argv[1]==0 ){
if( argv==0 || argv[0]==0 || argv[2]==0 ){
return 0;
}
pIndex = sqlite3FindIndex(pInfo->db, argv[0], pInfo->zDatabase);
if( pIndex==0 ){
pTable = sqlite3FindTable(pInfo->db, argv[0], pInfo->zDatabase);
if( pTable==0 ){
return 0;
}
z = argv[1];
for(i=0; *z && i<=pIndex->nColumn; i++){
if( argv[1] ){
pIndex = sqlite3FindIndex(pInfo->db, argv[1], pInfo->zDatabase);
}else{
pIndex = 0;
}
n = pIndex ? pIndex->nColumn : 0;
z = argv[2];
for(i=0; *z && i<=n; i++){
v = 0;
while( (c=z[0])>='0' && c<='9' ){
v = v*10 + c - '0';
z++;
}
if( i==0 ) pTable->nRowEst = v;
if( pIndex==0 ) break;
pIndex->aiRowEst[i] = v;
if( *z==' ' ) z++;
}
@ -555,7 +598,7 @@ int sqlite3AnalysisLoad(sqlite3 *db, int iDb){
/* Load new statistics out of the sqlite_stat1 table */
zSql = sqlite3MPrintf(db,
"SELECT idx, stat FROM %Q.sqlite_stat1", sInfo.zDatabase);
"SELECT tbl, idx, stat FROM %Q.sqlite_stat1", sInfo.zDatabase);
if( zSql==0 ){
rc = SQLITE_NOMEM;
}else{

15
src/build.c
View File

@ -802,6 +802,7 @@ void sqlite3StartTable(
pTable->iPKey = -1;
pTable->pSchema = db->aDb[iDb].pSchema;
pTable->nRef = 1;
pTable->nRowEst = 1000000;
assert( pParse->pNewTable==0 );
pParse->pNewTable = pTable;
@ -2830,14 +2831,14 @@ exit_create_index:
void sqlite3DefaultRowEst(Index *pIdx){
unsigned *a = pIdx->aiRowEst;
int i;
unsigned n;
assert( a!=0 );
a[0] = 1000000;
for(i=pIdx->nColumn; i>=5; i--){
a[i] = 5;
}
while( i>=1 ){
a[i] = 11 - i;
i--;
a[0] = pIdx->pTable->nRowEst;
if( a[0]<10 ) a[0] = 10;
n = 10;
for(i=1; i<=pIdx->nColumn; i++){
a[i] = n;
if( n>5 ) n--;
}
if( pIdx->onError!=OE_None ){
a[pIdx->nColumn] = 1;

1
src/sqliteInt.h
View File

@ -1252,6 +1252,7 @@ struct Table {
Column *aCol; /* Information about each column */
Index *pIndex; /* List of SQL indexes on this table. */
int tnum; /* Root BTree node for this table (see note above) */
unsigned nRowEst; /* Estimated rows in table - from sqlite_stat1 table */
Select *pSelect; /* NULL for tables. Points to definition if a view. */
u16 nRef; /* Number of pointers to this Table */
u8 tabFlags; /* Mask of TF_* values */

17
src/where.c
View File

@ -1706,7 +1706,7 @@ static void bestAutomaticIndex(
assert( pParse->nQueryLoop >= (double)1 );
pTable = pSrc->pTab;
nTableRow = pTable->pIndex ? pTable->pIndex->aiRowEst[0] : 1000000;
nTableRow = pTable->nRowEst;
logN = estLog(nTableRow);
costTempIdx = 2*logN*(nTableRow/pParse->nQueryLoop + 1);
if( costTempIdx>=pCost->rCost ){
@ -2513,23 +2513,14 @@ static void bestBtreeIndex(
sPk.nColumn = 1;
sPk.aiColumn = &aiColumnPk;
sPk.aiRowEst = aiRowEstPk;
aiRowEstPk[1] = 1;
sPk.onError = OE_Replace;
sPk.pTable = pSrc->pTab;
aiRowEstPk[0] = pSrc->pTab->nRowEst;
aiRowEstPk[1] = 1;
pFirst = pSrc->pTab->pIndex;
if( pSrc->notIndexed==0 ){
sPk.pNext = pFirst;
}
/* The aiRowEstPk[0] is an estimate of the total number of rows in the
** table. Get this information from the ANALYZE information if it is
** available. If not available, assume the table 1 million rows in size.
*/
if( pFirst ){
assert( pFirst->aiRowEst!=0 ); /* Allocated together with pFirst */
aiRowEstPk[0] = pFirst->aiRowEst[0];
}else{
aiRowEstPk[0] = 1000000;
}
pProbe = &sPk;
wsFlagMask = ~(
WHERE_COLUMN_IN|WHERE_COLUMN_EQ|WHERE_COLUMN_NULL|WHERE_COLUMN_RANGE
@ -4103,7 +4094,7 @@ WhereInfo *sqlite3WhereBegin(
**
** The best strategy is to iterate through table t1 first. However it
** is not possible to determine this with a simple greedy algorithm.
** However, since the cost of a linear scan through table t2 is the same
** Since the cost of a linear scan through table t2 is the same
** as the cost of a linear scan through table t1, a simple greedy
** algorithm may choose to use t2 for the outer loop, which is a much
** costlier approach.

14
test/analyze.test
View File

@ -73,7 +73,7 @@ do_test analyze-1.6.3 {
} {1 {table sqlite_stat1 may not be indexed}}
do_test analyze-1.7 {
execsql {
SELECT * FROM sqlite_stat1
SELECT * FROM sqlite_stat1 WHERE idx NOT NULL
}
} {}
do_test analyze-1.8 {
@ -83,7 +83,7 @@ do_test analyze-1.8 {
} {0 {}}
do_test analyze-1.9 {
execsql {
SELECT * FROM sqlite_stat1
SELECT * FROM sqlite_stat1 WHERE idx NOT NULL
}
} {}
do_test analyze-1.10 {
@ -96,7 +96,7 @@ do_test analyze-1.11 {
execsql {
SELECT * FROM sqlite_stat1
}
} {}
} {t1 {} 0}
do_test analyze-1.12 {
catchsql {
ANALYZE t1;
@ -106,7 +106,7 @@ do_test analyze-1.13 {
execsql {
SELECT * FROM sqlite_stat1
}
} {}
} {t1 {} 0}
# Create some indices that can be analyzed. But do not yet add
# data. Without data in the tables, no analysis is done.
@ -117,21 +117,21 @@ do_test analyze-2.1 {
ANALYZE main.t1;
SELECT * FROM sqlite_stat1 ORDER BY idx;
}
} {}
} {t1 {} 0}
do_test analyze-2.2 {
execsql {
CREATE INDEX t1i2 ON t1(b);
ANALYZE t1;
SELECT * FROM sqlite_stat1 ORDER BY idx;
}
} {}
} {t1 {} 0}
do_test analyze-2.3 {
execsql {
CREATE INDEX t1i3 ON t1(a,b);
ANALYZE main;
SELECT * FROM sqlite_stat1 ORDER BY idx;
}
} {}
} {t1 {} 0}
# Start adding data to the table. Verify that the analysis
# is done correctly.

6
test/auth.test
View File

@ -1976,12 +1976,12 @@ ifcapable analyze {
ANALYZE;
}
set ::authargs
} {t4 {} main {}}
} {t4 {} main {} t2 {} main {}}
do_test auth-1.295 {
execsql {
SELECT count(*) FROM sqlite_stat1;
}
} 2
} 3
proc auth {code args} {
if {$code=="SQLITE_ANALYZE"} {
set ::authargs [concat $::authargs $args]
@ -1999,7 +1999,7 @@ ifcapable analyze {
execsql {
SELECT count(*) FROM sqlite_stat1;
}
} 2
} 3
} ;# ifcapable analyze

2
test/misc4.test
View File

@ -151,7 +151,7 @@ ifcapable subquery {
select a.*, x.*
from a, (select key,sum(period) from b group by key) as x
where a.key=x.key;
where a.key=x.key order by 1 desc;
}
} {01 data01 01 3 +1 data+1 +1 7}

1
test/select6.test
View File

@ -458,6 +458,7 @@ do_test select6-8.6 {
do_test select6-9.1 {
execsql {
SELECT a.x, b.x FROM t1 AS a, (SELECT x FROM t1 LIMIT 2) AS b
ORDER BY 1, 2
}
} {1 1 1 2 2 1 2 2 3 1 3 2 4 1 4 2}
do_test select6-9.2 {

4
test/tkt3757.test
View File

@ -35,9 +35,9 @@ do_test tkt3757-1.1 {
CREATE TABLE t2(a INTEGER, b TEXT);
INSERT INTO t2 VALUES(2, 'two');
ANALYZE;
SELECT * FROM sqlite_stat1;
SELECT * FROM sqlite_stat1 ORDER BY 1, 2;
}
} {t1 t1i1 {1 1 1}}
} {t1 t1i1 {1 1 1} t2 {} 1}
# Modify statistics in order to make the optimizer then that:
#