Enhancements to PRAGMA optimize and ANALYZE. Add the 0x10000 flag to

PRAGMA optimize.  ANALYZE now records zero-size partial indexes in the
sqlite_stat1 table.  PRAGMA optimize looks for both growth and shrinkage
in table sizes, and uses tighter bounds (10x rather than 25x) to trigger
a re-analyze.  PRAGMA optimize automatically uses are reasonable
analysis_limit to prevent excessive runtimes.

FossilOrigin-Name: 63ef234e88857a653fa3541e80d59802ceccb806ac8296e8bae79a385b7086f7

1 files changed, 158 insertions, 38 deletions

diff --git a/src/pragma.c b/src/pragma.c
index 872177e13..877fb3f72 100644
--- a/src/pragma.c
+++ b/src/pragma.c
@@ -31,6 +31,34 @@
 #include "pragma.h"
 
 /*
+** When the 0x10 bit of PRAGMA optimize is set, any ANALYZE commands
+** will be run with an analysis_limit set to the lessor of the value of
+** the following macro or to the actual analysis_limit if it is non-zero,
+** in order to prevent PRAGMA optimize from running for too long.
+**
+** The value of 2000 is chosen emperically so that the worst-case run-time
+** for PRAGMA optimize does not exceed 100 milliseconds against a variety
+** of test databases on a RaspberryPI-4 compiled using -Os and without
+** -DSQLITE_DEBUG.  Of course, your mileage may vary.  For the purpose of
+** his paragraph, "worst-case" means that ANALYZE ends up being
+** run on every table in the database.  The worst case typically only
+** happens if PRAGMA optimize is run on a database file for which ANALYZE
+** has not been previously run and the 0x10000 flag is included so that
+** all tables are analyzed.  The usual case for PRAGMA optimize is that
+** no ANALYZE commands will be run at all, or if any ANALYZE happens it
+** will be against a single table, so that expected timing for PRAGMA
+** optimize on a PI-4 is more like 1 millisecond or less with the 0x10000
+** flag or less than 100 microseconds without the 0x10000 flag.
+**
+** An analysis limit of 2000 is almost always sufficient for the query
+** planner to fully characterize an index.  The additional accuracy from
+** a larger analysis is not usually helpful.
+*/
+#ifndef SQLITE_DEFAULT_OPTIMIZE_LIMIT
+# define SQLITE_DEFAULT_OPTIMIZE_LIMIT 2000
+#endif
+
+/*
 ** Interpret the given string as a safety level.  Return 0 for OFF,
 ** 1 for ON or NORMAL, 2 for FULL, and 3 for EXTRA.  Return 1 for an empty or
 ** unrecognized string argument.  The FULL and EXTRA option is disallowed
@@ -2383,44 +2411,66 @@ void sqlite3Pragma(
   **
   ** The optional argument is a bitmask of optimizations to perform:
   **
-  **    0x0001    Debugging mode.  Do not actually perform any optimizations
-  **              but instead return one line of text for each optimization
-  **              that would have been done.  Off by default.
+  **    0x00001    Debugging mode.  Do not actually perform any optimizations
+  **               but instead return one line of text for each optimization
+  **               that would have been done.  Off by default.
   **
-  **    0x0002    Run ANALYZE on tables that might benefit.  On by default.
-  **              See below for additional information.
+  **    0x00002    Run ANALYZE on tables that might benefit.  On by default.
+  **               See below for additional information.
   **
-  **    0x0004    (Not yet implemented) Record usage and performance
-  **              information from the current session in the
-  **              database file so that it will be available to "optimize"
-  **              pragmas run by future database connections.
+  **    0x00010    Run all ANALYZE operations using an analysis_limit that
+  **               is the lessor of the current analysis_limit and the
+  **               SQLITE_DEFAULT_OPTIMIZE_LIMIT compile-time option.
+  **               The default value of SQLITE_DEFAULT_OPTIMIZE_LIMIT is
+  **               currently (2024-02-19) set to 2000, which is such that
+  **               the worst case run-time for PRAGMA optimize on a 100MB
+  **               database will usually be less than 100 milliseconds on
+  **               a RaspberryPI-4 class machine.  On by default.
   **
-  **    0x0008    (Not yet implemented) Create indexes that might have
-  **              been helpful to recent queries
+  **    0x00020    Run ANALYZE on any table that has a index that lacks an
+  **               entry in the sqlite_stat1 table.  On by default.
   **
-  ** The default MASK is and always shall be 0xfffe.  0xfffe means perform all
-  ** of the optimizations listed above except Debug Mode, including new
-  ** optimizations that have not yet been invented.  If new optimizations are
-  ** ever added that should be off by default, those off-by-default
-  ** optimizations will have bitmasks of 0x10000 or larger.
+  **    0x10000    Look at tables to see if they need to be reanalyzed
+  **               due to growth or shrinkage even if they have not been
+  **               queried during the current connection.  Off by default.
+  **
+  ** The default MASK is and always shall be 0x0fffe.  In the current
+  ** implementation, the default mask only covers the 0x00002 optimization,
+  ** though additional optimizations that are covered by 0x0fffe might be
+  ** added in the future.  Optimizations that are off by default and must
+  ** be explicitly requested have masks of 0x10000 or greater.
   **
   ** DETERMINATION OF WHEN TO RUN ANALYZE
   **
   ** In the current implementation, a table is analyzed if only if all of
   ** the following are true:
   **
-  ** (1) MASK bit 0x02 is set.
+  ** (1) MASK bit 0x00002 is set.
+  **
+  ** (2) The table is an ordinary table, not a virtual table or view.
   **
-  ** (2) The query planner used sqlite_stat1-style statistics for one or
-  **     more indexes of the table at some point during the lifetime of
-  **     the current connection.
+  ** (3) The table name does not begin with "sqlite_".
   **
-  ** (3) One or more indexes of the table are currently unanalyzed OR
-  **     the number of rows in the table has increased by 25 times or more
-  **     since the last time ANALYZE was run.
+  ** (4) One or more of the following is true:
+  **      (4a) The 0x10000 MASK bit is set.
+  **      (4b) One or more indexes on the table lacks an entry
+  **           in the sqlite_stat1 table.
+  **      (4c) The query planner used sqlite_stat1-style statistics for one
+  **           or more indexes of the tableat some point during the lifetime
+  **           of the current connection.
+  **
+  ** (5) One or more of the following is true:
+  **      (5a) One or more indexes on the table lacks an entry
+  **           in the sqlite_stat1 table.  (Same as 4a)
+  **      (5b) The number of rows in the table has increased or decreased by
+  **           10-fold.  In other words, the current size of the table is
+  **           10 times larger than the size in sqlite_stat1 or else the
+  **           current size is less than 1/10th the size in sqlite_stat1.
   **
   ** The rules for when tables are analyzed are likely to change in
-  ** future releases.
+  ** future releases.  Future versions of SQLite might accept a string
+  ** literal argument to this pragma that contains a mnemonic description
+  ** of the options rather than a bitmap.
   */
   case PragTyp_OPTIMIZE: {
     int iDbLast;           /* Loop termination point for the schema loop */
@@ -2432,6 +2482,11 @@ void sqlite3Pragma(
     LogEst szThreshold;    /* Size threshold above which reanalysis needed */
     char *zSubSql;         /* SQL statement for the OP_SqlExec opcode */
     u32 opMask;            /* Mask of operations to perform */
+    int nLimit;            /* Analysis limit to use */
+    int nCheck = 0;        /* Number of tables to be optimized */
+    int nBtree = 0;        /* Number of btrees to scan */
+    int nIndex;            /* Number of indexes on the current table */
+    int hasStat1;          /* True if any STAT1 info available for the table */
 
     if( zRight ){
       opMask = (u32)sqlite3Atoi(zRight);
@@ -2439,6 +2494,14 @@ void sqlite3Pragma(
     }else{
       opMask = 0xfffe;
     }
+    if( (opMask & 0x10)==0 ){
+      nLimit = 0;
+    }else if( db->nAnalysisLimit>0
+           && db->nAnalysisLimit<SQLITE_DEFAULT_OPTIMIZE_LIMIT ){
+      nLimit = 0;
+    }else{
+      nLimit = SQLITE_DEFAULT_OPTIMIZE_LIMIT;
+    }
     iTabCur = pParse->nTab++;
     for(iDbLast = zDb?iDb:db->nDb-1; iDb<=iDbLast; iDb++){
       if( iDb==1 ) continue;
@@ -2447,23 +2510,64 @@ void sqlite3Pragma(
       for(k=sqliteHashFirst(&pSchema->tblHash); k; k=sqliteHashNext(k)){
         pTab = (Table*)sqliteHashData(k);
 
-        /* If table pTab has not been used in a way that would benefit from
-        ** having analysis statistics during the current session, then skip it.
-        ** This also has the effect of skipping virtual tables and views */
-        if( (pTab->tabFlags & TF_MaybeReanalyze)==0 ) continue;
+        /* This only works for ordinary tables */
+        if( !IsOrdinaryTable(pTab) ) continue;
+
+        /* Do not scan system tables */
+        if( 0==sqlite3StrNICmp(pTab->zName, "sqlite_", 7) ) continue;
 
-        /* Reanalyze if the table is 25 times larger than the last analysis */
-        szThreshold = pTab->nRowLogEst + 46; assert( sqlite3LogEst(25)==46 );
+        /* Find the size of the table as last recorded in sqlite_stat1.
+        ** If any index is unanalyzed, then the threshold is -1 to
+        ** indicate a new, unanalyzed index
+        */
+        szThreshold = pTab->nRowLogEst;
+        hasStat1 = (pTab->tabFlags & TF_HasStat1)!=0;
+        nIndex = 0;
         for(pIdx=pTab->pIndex; pIdx; pIdx=pIdx->pNext){
-          if( !pIdx->hasStat1 ){
-            szThreshold = 0; /* Always analyze if any index lacks statistics */
-            break;
+          nIndex++;
+          if( pIdx->hasStat1 ){
+            hasStat1 = 1;
+          }else{
+            szThreshold = -1; /* Always analyze if any index lacks statistics */
           }
         }
-        if( szThreshold ){
-          sqlite3OpenTable(pParse, iTabCur, iDb, pTab, OP_OpenRead);
-          sqlite3VdbeAddOp3(v, OP_IfSmaller, iTabCur,
-                         sqlite3VdbeCurrentAddr(v)+2+(opMask&1), szThreshold);
+
+        /* If table pTab has not been used in a way that would benefit from
+        ** having analysis statistics during the current session, then skip it,
+        ** unless the 0x10000 MASK bit is set. */
+        if( (pTab->tabFlags & TF_MaybeReanalyze)!=0 ){
+          /* Check for size change if stat1 has been used for a query */
+        }else if( opMask & 0x10000 ){
+          /* Check for size change if 0x10000 is set */
+        }else if( pTab->pIndex!=0 && szThreshold<0 ){
+          /* Do analysis if unanalyzed indexes exists */
+        }else{
+          /* Otherwise, we can skip this table */
+          continue;
+        }
+
+        nCheck++;
+        if( nCheck==2 ){
+          /* If ANALYZE might be invoked two or more times, hold a write
+          ** transaction for efficiency */
+          sqlite3BeginWriteOperation(pParse, 0, iDb);
+        }
+        nBtree += nIndex+1;
+
+        /* Reanalyze if the table is 10 times larger or smaller than
+        ** the last analysis.  Unconditional reanalysis if there are
+        ** unanalyzed indexes. */
+        sqlite3OpenTable(pParse, iTabCur, iDb, pTab, OP_OpenRead);
+        if( szThreshold>=0 ){
+          const LogEst iRange = 33;   /* 10x size change */
+          sqlite3VdbeAddOp4Int(v, OP_IfSizeBetween, iTabCur,
+                         sqlite3VdbeCurrentAddr(v)+2+(opMask&1),
+                         szThreshold>=iRange ? szThreshold-iRange : -1,
+                         szThreshold+iRange);
+          VdbeCoverage(v);
+        }else{
+          sqlite3VdbeAddOp2(v, OP_Rewind, iTabCur,
+                         sqlite3VdbeCurrentAddr(v)+2+(opMask&1));
           VdbeCoverage(v);
         }
         zSubSql = sqlite3MPrintf(db, "ANALYZE \"%w\".\"%w\"",
@@ -2473,11 +2577,27 @@ void sqlite3Pragma(
           sqlite3VdbeAddOp4(v, OP_String8, 0, r1, 0, zSubSql, P4_DYNAMIC);
           sqlite3VdbeAddOp2(v, OP_ResultRow, r1, 1);
         }else{
-          sqlite3VdbeAddOp4(v, OP_SqlExec, 0, 0, 0, zSubSql, P4_DYNAMIC);
+          sqlite3VdbeAddOp4(v, OP_SqlExec, nLimit ? 0x02 : 00, nLimit, 0,
+                            zSubSql, P4_DYNAMIC);
         }
       }
     }
     sqlite3VdbeAddOp0(v, OP_Expire);
+
+    /* In a schema with a large number of tables and indexes, scale back
+    ** the analysis_limit to avoid excess run-time in the worst case.
+    */
+    if( !db->mallocFailed && nLimit>0 && nBtree>100 ){
+      int iAddr, iEnd;
+      VdbeOp *aOp;
+      nLimit = 100*nLimit/nBtree;
+      if( nLimit<100 ) nLimit = 100;
+      aOp = sqlite3VdbeGetOp(v, 0);
+      iEnd = sqlite3VdbeCurrentAddr(v);
+      for(iAddr=0; iAddr<iEnd; iAddr++){
+        if( aOp[iAddr].opcode==OP_SqlExec ) aOp[iAddr].p2 = nLimit;
+      }
+    }
     break;
   }