First code for the ANALYZE command. Mostly untested. The analysis is

not loaded into the symbol tables and is not used by the optimizer. (CVS 2560)

FossilOrigin-Name: a4886b114d2ccb3841d3d219f6b97f67745b13c2

manifest

@@ -1,5 +1,5 @@
-C Clarify\sthe\smeaning\sof\sthe\sTEMP\skeyword\son\sCREATE\sTABLE.\s(CVS\s2559)
+C First\scode\sfor\sthe\sANALYZE\scommand.\s\sMostly\suntested.\s\sThe\sanalysis\sis\nnot\sloaded\sinto\sthe\ssymbol\stables\sand\sis\snot\sused\sby\sthe\soptimizer.\s(CVS\s2560)
-D 2005-07-22T23:56:50
+D 2005-07-23T00:41:49
 F Makefile.in 22ea9c0fe748f591712d8fe3c6d972c6c173a165
 F Makefile.linux-gcc 06be33b2a9ad4f005a5f42b22c4a19dab3cbb5c7
 F README 9c4e2d6706bdcc3efdd773ce752a8cdab4f90028
@@ -28,12 +28,12 @@ F sqlite3.1 6be1ad09113570e1fc8dcaff84c9b0b337db5ffc
 F sqlite3.def c413e514217736884254739a105c8c942fdf0c2f
 F sqlite3.pc.in 985b9bf34192a549d7d370e0f0b6b34a4f61369a
 F src/alter.c 03041f2464e22532601254f87cb49997fa21dcdf
-F src/analyze.c 2f770b39f9d77353ac36db1861319e6323772f82
+F src/analyze.c 0eb48929dd1caa21438b067c507b0ee55ce17047
 F src/attach.c 3615dbe960cbee4aa5ea300b8a213dad36527b0f
 F src/auth.c 18c5a0befe20f3a58a41e3ddd78f372faeeefe1f
 F src/btree.c ec55bd70052cdd0958f3a0e79ad58d93561acb20
 F src/btree.h 41a71ce027db9ddee72cb43df2316bbe3a1d92af
-F src/build.c c2b9379e3b51775de01137a68d3a67359e29d3aa
+F src/build.c 50e1197f5f70d544dc85b8483766f6194375d04e
 F src/callback.c 0910b611e0c158f107ee3ff86f8a371654971e2b
 F src/date.c 7444b0900a28da77e57e3337a636873cff0ae940
 F src/delete.c be1fc25c9e109cd8cbab42a43ee696263da7c04b
@@ -64,7 +64,7 @@ F src/random.c 90adff4e73a3b249eb4f1fc2a6ff9cf78c7233a4
 F src/select.c c611471052773b94af771693686bd5bcdbbb0dba
 F src/shell.c 25b3217d7c64e6497225439d261a253a23efff26
 F src/sqlite.h.in 838382ed6b48d392366a55e07f49d9d71263e1fe
-F src/sqliteInt.h 810f2ccd49338944ec53c95182e5a8b56da42fa0
+F src/sqliteInt.h 2925510c0233bb24c550fc5912fcdf0dd3a4421d
 F src/table.c 25b3ff2b39b7d87e8d4a5da0713d68dfc06cbee9
 F src/tclsqlite.c cccaf6b78c290d824cf8ea089b8b27377e545830
 F src/test1.c 722c1444b5774705eb6eb11163343fc94ffe17f7
@@ -286,7 +286,7 @@ F www/tclsqlite.tcl 425be741b8ae664f55cb1ef2371aab0a75109cf9
 F www/vdbe.tcl 87a31ace769f20d3627a64fa1fade7fed47b90d0
 F www/version3.tcl a99cf5f6d8bd4d5537584a2b342f0fb9fa601d8b
 F www/whentouse.tcl 528299b8316726dbcc5548e9aa0648c8b1bd055b
-P fd1e013a14710ed2ecabddedc0d110562cfdbfc2
+P 1a573619f558d1d67775d17aabb9d704f82ad1a8
-R c3a6d2b6548be7b005874f5369987f47
+R a3fdedefcb771a47f367010f865883f0
 U drh
-Z 5560bd6b70c08f20b53ebb18f06f4772
+Z d92cfb5ebecb44f8fc7d777743031b10

manifest.uuid

@@ -1 +1 @@
-1a573619f558d1d67775d17aabb9d704f82ad1a8
+a4886b114d2ccb3841d3d219f6b97f67745b13c2

src/analyze.c

@@ -11,16 +11,232 @@
 *************************************************************************
 ** This file contains code associated with the ANALYZE command.
 **
-** @(#) $Id: analyze.c,v 1.1 2005/07/08 12:13:05 drh Exp $
+** @(#) $Id: analyze.c,v 1.2 2005/07/23 00:41:49 drh Exp $
 */
 #ifndef SQLITE_OMIT_ANALYZE
 #include "sqliteInt.h"
 
 /*
-** Generate code for the ANALYZE command
+** This routine generates code that opens the sqlite_stat1 table on cursor
+** iStatCur.
+**
+** If the sqlite_stat1 tables does not previously exist, it is created.
+** If it does previously exist, all entires associated with table zWhere
+** are removed.  If zWhere==0 then all entries are removed.
+*/
+static void openStatTable(
+  Parse *pParse,          /* Parsing context */
+  int iDb,                /* The database we are looking in */
+  int iStatCur,           /* Open the sqlite_stat1 table on this cursor */
+  const char *zWhere      /* Delete entries associated with this table */
+){
+  sqlite3 *db = pParse->db;
+  Db *pDb;
+  int iRootPage;
+  Table *pStat;
+  Vdbe *v = sqlite3GetVdbe(pParse);
+
+  pDb = &db->aDb[iDb];
+  if( (pStat = sqlite3FindTable(db, "sqlite_stat1", pDb->zName))==0 ){
+    /* The sqlite_stat1 tables does not exist.  Create it.  
+    ** Note that a side-effect of the CREATE TABLE statement is to leave
+    ** the rootpage of the new table on the top of the stack.  This is
+    ** important because the OpenWrite opcode below will be needing it. */
+    sqlite3NestedParse(pParse,
+      "CREATE TABLE %Q.sqlite_stat1(tbl,idx,stat)",
+      pDb->zName
+    );
+    iRootPage = 0;  /* Cause rootpage to be taken from top of stack */
+  }else if( zWhere ){
+    /* The sqlite_stat1 table exists.  Delete all entries associated with
+    ** the table zWhere. */
+    sqlite3NestedParse(pParse,
+       "DELETE FROM %Q.sqlite_stat1 WHERE tbl=%Q",
+       pDb->zName, zWhere
+    );
+    iRootPage = pStat->tnum;
+  }else{
+    /* The sqlite_stat1 table already exists.  Delete all rows. */
+    iRootPage = pStat->tnum;
+    sqlite3VdbeAddOp(v, OP_Clear, pStat->tnum, iDb);
+  }
+
+  /* Open the sqlite_stat1 table for writing.
+  */
+  sqlite3VdbeAddOp(v, OP_Integer, iDb, 0);
+  sqlite3VdbeAddOp(v, OP_OpenWrite, iStatCur, 0);
+  sqlite3VdbeAddOp(v, OP_SetNumColumns, iStatCur, 3);
+}
+
+/*
+** Generate code to do an analysis of all indices associated with
+** a single table.
+*/
+static void analyzeOneTable(
+  Parse *pParse,   /* Parser context */
+  Table *pTab,     /* Table whose indices are to be analyzed */
+  int iStatCur,    /* Cursor that writes to the sqlite_stat1 table */
+  int iMem         /* Available memory locations begin here */
+){
+  Index *pIdx;     /* An index to being analyzed */
+  int iIdxCur;     /* Cursor number for index being analyzed */
+  int nCol;        /* Number of columns in the index */
+  Vdbe *v;         /* The virtual machine being built up */
+  int i;           /* Loop counter */
+  int topOfLoop;   /* The top of the loop */
+  int endOfLoop;   /* The end of the loop */
+  int addr;        /* The address of an instruction */
+
+  v = sqlite3GetVdbe(pParse);
+  if( pTab==0 || pTab->pIndex==0 || pTab->pIndex->pNext==0 ){
+    /* Do no analysis for tables with fewer than 2 indices */
+    return;
+  }
+  iIdxCur = pParse->nTab;
+  for(pIdx=pTab->pIndex; pIdx; pIdx=pIdx->pNext){
+    /* Open a cursor to the index to be analyzed
+    */
+    sqlite3VdbeAddOp(v, OP_Integer, pIdx->iDb, 0);
+    VdbeComment((v, "# %s", pIdx->zName));
+    sqlite3VdbeOp3(v, OP_OpenRead, iIdxCur, pIdx->tnum,
+                     (char*)&pIdx->keyInfo, P3_KEYINFO);
+    nCol = pIdx->nColumn;
+    if( iMem+nCol*2>=pParse->nMem ){
+      pParse->nMem = iMem+nCol*2+1;
+    }
+    sqlite3VdbeAddOp(v, OP_SetNumColumns, iIdxCur, nCol+1);
+
+    /* Memory cells are used as follows:
+    **
+    **    mem[iMem]:             The total number of rows in the table.
+    **    mem[iMem+1]:           Number of distinct values in column 1
+    **    ...
+    **    mem[iMem+nCol]:        Number of distinct values in column N
+    **    mem[iMem+nCol+1]       Last observed value of column 1
+    **    ...
+    **    mem[iMem+nCol+nCol]:   Last observed value of column N
+    **
+    ** Cells iMem through iMem+nCol are initialized to 0.  The others
+    ** are initialized to NULL.
+    */
+    sqlite3VdbeAddOp(v, OP_Integer, 0, 0);
+    for(i=0; i<=nCol; i++){
+      sqlite3VdbeAddOp(v, OP_MemStore, iMem+i, i==nCol);
+    }
+    sqlite3VdbeAddOp(v, OP_Null, 0, 0);
+    for(i=0; i<nCol; i++){
+      sqlite3VdbeAddOp(v, OP_MemStore, iMem+nCol+i+1, i==nCol-1);
+    }
+
+    /* Do the analysis.
+    */
+    sqlite3VdbeAddOp(v, OP_Rewind, iIdxCur, 0);
+    topOfLoop = sqlite3VdbeCurrentAddr(v);
+    endOfLoop = sqlite3VdbeMakeLabel(v);
+    sqlite3VdbeAddOp(v, OP_MemIncr, iMem, 0);
+    for(i=0; i<nCol; i++){
+      sqlite3VdbeAddOp(v, OP_Column, iIdxCur, i);
+      sqlite3VdbeAddOp(v, OP_MemLoad, iMem+nCol+i+1, 0);
+      sqlite3VdbeAddOp(v, OP_Ne, 0x100, 0);
+    }
+    sqlite3VdbeAddOp(v, OP_Goto, 0, endOfLoop);
+    for(i=0; i<nCol; i++){
+      addr = sqlite3VdbeAddOp(v, OP_MemIncr, iMem+i+1, 0);
+      sqlite3VdbeChangeP2(v, topOfLoop + 3*i + 3, addr);
+      sqlite3VdbeAddOp(v, OP_Column, iIdxCur, i);
+      sqlite3VdbeAddOp(v, OP_MemStore, iMem+nCol+i+1, 1);
+    }
+    sqlite3VdbeResolveLabel(v, endOfLoop);
+    sqlite3VdbeAddOp(v, OP_Next, iIdxCur, topOfLoop);
+    sqlite3VdbeAddOp(v, OP_Close, iIdxCur, 0);
+
+    /* Store the results.  
+    **
+    ** The result is a single row of the sqlite_stmt1 table.  The first
+    ** two columns are the names of the table and index.  The third column
+    ** is a string composed of a list of integer statistics about the
+    ** index.  There is one integer in the list for each column of the table.
+    ** This integer is a guess of how many rows of the table the index will
+    ** select.  If D is the count of distinct values and K is the total
+    ** number of rows, then the integer is computed as:
+    **
+    **        I = (K+D-1)/D
+    **
+    ** If K==0 then no entry is made into the sqlite_stat1 table.  
+    ** If K>0 then it is always the case the D>0 so division by zero
+    ** is never possible.
+    */
+    sqlite3VdbeAddOp(v, OP_MemLoad, iMem, 0);
+    addr = sqlite3VdbeAddOp(v, OP_IfNot, 0, 0);
+    sqlite3VdbeAddOp(v, OP_NewRowid, iStatCur, 0);
+    sqlite3VdbeOp3(v, OP_String8, 0, 0, pTab->zName, 0);
+    sqlite3VdbeOp3(v, OP_String8, 0, 0, pIdx->zName, 0);
+    for(i=0; i<nCol; i++){
+      sqlite3VdbeAddOp(v, OP_MemLoad, iMem, 0);
+      sqlite3VdbeAddOp(v, OP_MemLoad, iMem+i+1, 0);
+      sqlite3VdbeAddOp(v, OP_Add, 0, 0);
+      sqlite3VdbeAddOp(v, OP_AddImm, -1, 0);
+      sqlite3VdbeAddOp(v, OP_MemLoad, iMem+i+1, 0);
+      sqlite3VdbeAddOp(v, OP_Divide, 0, 0);
+      if( i==nCol-1 ){
+        if( i>0 ){
+          sqlite3VdbeAddOp(v, OP_Concat, nCol*2-3, 0);
+        }
+      }else{
+        if( i==0 ){
+          sqlite3VdbeOp3(v, OP_String8, 0, 0, " ", 0);
+        }else{
+          sqlite3VdbeAddOp(v, OP_Dup, 1, 0);
+        }
+      }
+    }
+    sqlite3VdbeOp3(v, OP_MakeRecord, 3, 0, "ttt", 0);
+    sqlite3VdbeAddOp(v, OP_Insert, iStatCur, 0);
+    sqlite3VdbeChangeP2(v, addr, sqlite3VdbeCurrentAddr(v));
+  }
+}
+
+/*
+** Generate code that will do an analysis of an entire database
+*/
+static void analyzeDatabase(Parse *pParse, int iDb){
+  sqlite3 *db = pParse->db;
+  HashElem *k;
+  int iStatCur;
+  int iMem;
+
+  sqlite3BeginWriteOperation(pParse, 0, iDb);
+  iStatCur = pParse->nTab++;
+  openStatTable(pParse, iDb, iStatCur, 0);
+  iMem = pParse->nMem;
+  for(k=sqliteHashFirst(&db->aDb[iDb].tblHash);  k; k=sqliteHashNext(k)){
+    Table *pTab = (Table*)sqliteHashData(k);
+    analyzeOneTable(pParse, pTab, iStatCur, iMem);
+  }
+}
+
+/*
+** Generate code that will do an analysis of a single table in
+** a database.
+*/
+static void analyzeTable(Parse *pParse, Table *pTab){
+  int iDb;
+  int iStatCur;
+
+  assert( pTab!=0 );
+  iDb = pTab->iDb;
+  sqlite3BeginWriteOperation(pParse, 0, iDb);
+  iStatCur = pParse->nTab++;
+  openStatTable(pParse, iDb, iStatCur, pTab->zName);
+  analyzeOneTable(pParse, pTab, iStatCur, pParse->nMem);
+}
+
+/*
+** Generate code for the ANALYZE command.  The parser calls this routine
+** when it recognizes an ANALYZE command.
 **
 **        ANALYZE                            -- 1
-**        ANALYZE  <database >               -- 2
+**        ANALYZE  <database>                -- 2
 **        ANALYZE  ?<database>.?<tablename>  -- 3
 **
 ** Form 1 causes all indices in all attached databases to be analyzed.
@@ -28,7 +244,54 @@
 ** Form 3 analyzes all indices associated with the named table.
 */
 void sqlite3Analyze(Parse *pParse, Token *pName1, Token *pName2){
+  sqlite3 *db = pParse->db;
+  int iDb;
+  int i;
+  char *z, *zDb;
+  Table *pTab;
+  Token *pTableName;
+
+  /* Read the database schema. If an error occurs, leave an error message
+  ** and code in pParse and return NULL. */
+  if( SQLITE_OK!=sqlite3ReadSchema(pParse) ){
+    return;
+  }
+
+  if( pName1==0 ){
+    /* Form 1:  Analyze everything */
+    for(i=0; i<db->nDb; i++){
+      if( i==1 ) continue;  /* Do not analyze the TEMP database */
+      analyzeDatabase(pParse, i);
+    }
+  }else if( pName2==0 ){
+    /* Form 2:  Analyze the database or table named */
+    iDb = sqlite3FindDb(db, pName1);
+    if( iDb>=0 ){
+      analyzeDatabase(pParse, iDb);
+      return;
+    }
+    z = sqlite3NameFromToken(pName1);
+    pTab = sqlite3LocateTable(pParse, z, 0);
+    sqliteFree(z);
+    if( pTab ){
+      analyzeTable(pParse, pTab);
+    }
+    return;
+  }else{
+    /* Form 3: Analyze the fully qualified table name */
+    iDb = sqlite3TwoPartName(pParse, pName1, pName2, &pTableName);
+    if( iDb>=0 ){
+      zDb = db->aDb[iDb].zName;
+      z = sqlite3NameFromToken(pTableName);
+      pTab = sqlite3LocateTable(pParse, z, zDb);
+      sqliteFree(z);
+      if( pTab ){
+        analyzeTable(pParse, pTab);
+      }
+    }   
+  }
 }
 
 
+
 #endif /* SQLITE_OMIT_ANALYZE */

src/build.c

@@ -22,7 +22,7 @@
 **     COMMIT
 **     ROLLBACK
 **
-** $Id: build.c,v 1.332 2005/07/21 18:23:20 drh Exp $
+** $Id: build.c,v 1.333 2005/07/23 00:41:49 drh Exp $
 */
 #include "sqliteInt.h"
 #include <ctype.h>
@@ -535,7 +535,7 @@ void sqlite3OpenMasterTable(Vdbe *v, int iDb){
 ** index of the named database in db->aDb[], or -1 if the named db 
 ** does not exist.
 */
-static int findDb(sqlite3 *db, Token *pName){
+int sqlite3FindDb(sqlite3 *db, Token *pName){
   int i = -1;    /* Database number */
   int n;         /* Number of characters in the name */
   Db *pDb;       /* A database whose name space is being searched */
@@ -583,7 +583,7 @@ int sqlite3TwoPartName(
   if( pName2 && pName2->n>0 ){
     assert( !db->init.busy );
     *pUnqual = pName2;
-    iDb = findDb(db, pName1);
+    iDb = sqlite3FindDb(db, pName1);
     if( iDb<0 ){
       sqlite3ErrorMsg(pParse, "unknown database %T", pName1);
       pParse->nErr++;
@@ -2813,7 +2813,7 @@ void reindexDatabases(Parse *pParse, CollSeq *pColl){
 
   for(iDb=0, pDb=db->aDb; iDb<db->nDb; iDb++, pDb++){
     if( pDb==0 ) continue;
-      for(k=sqliteHashFirst(&pDb->tblHash);  k; k=sqliteHashNext(k)){
+    for(k=sqliteHashFirst(&pDb->tblHash);  k; k=sqliteHashNext(k)){
       pTab = (Table*)sqliteHashData(k);
       reindexTable(pParse, pTab, pColl);
     }

src/sqliteInt.h

@@ -11,7 +11,7 @@
 *************************************************************************
 ** Internal interface definitions for SQLite.
 **
-** @(#) $Id: sqliteInt.h,v 1.397 2005/07/22 00:31:40 drh Exp $
+** @(#) $Id: sqliteInt.h,v 1.398 2005/07/23 00:41:49 drh Exp $
 */
 #ifndef _SQLITEINT_H_
 #define _SQLITEINT_H_
@@ -1569,6 +1569,7 @@ CollSeq *sqlite3GetCollSeq(sqlite3*, CollSeq *, const char *, int);
 char sqlite3AffinityType(const Token*);
 void sqlite3Analyze(Parse*, Token*, Token*);
 int sqlite3InvokeBusyHandler(BusyHandler*);
+int sqlite3FindDb(sqlite3*, Token*);
 
 #ifdef SQLITE_SSE
 #include "sseInt.h"