From: dan <dan@noemail.net>
Date: Mon, 30 Sep 2013 18:14:45 +0000 (+0000)
Subject: Fix a performance problem in the FTS4 auxiliary functions triggered by an OR clause... 
X-Git-Tag: version-3.8.1~57
X-Git-Url: http://git.ipfire.org/?a=commitdiff_plain;h=1f03e609ff2e02493bfa274872837d47e47ee4a8;p=thirdparty%2Fsqlite.git

Fix a performance problem in the FTS4 auxiliary functions triggered by an OR clause in the full-text query.

FossilOrigin-Name: fa0f2f0e3e79ae653118b901e1cca7725dfaf249
---

diff --git a/ext/fts3/fts3.c b/ext/fts3/fts3.c
index 0f7c38e571..4bd80db8b1 100644
--- a/ext/fts3/fts3.c
+++ b/ext/fts3/fts3.c
@@ -5268,15 +5268,23 @@ int sqlite3Fts3EvalPhrasePoslist(
   pIter = pPhrase->doclist.pList;
   if( iDocid!=pCsr->iPrevId || pExpr->bEof ){
     int bDescDoclist = pTab->bDescIdx;      /* For DOCID_CMP macro */
+    int iMul;                     /* +1 if csr dir matches index dir, else -1 */
     int bOr = 0;
     u8 bEof = 0;
-    Fts3Expr *p;
+    u8 bTreeEof = 0;
+    Fts3Expr *p;                  /* Used to iterate from pExpr to root */
+    Fts3Expr *pNear;              /* Most senior NEAR ancestor (or pExpr) */
 
     /* Check if this phrase descends from an OR expression node. If not, 
     ** return NULL. Otherwise, the entry that corresponds to docid 
-    ** pCsr->iPrevId may lie earlier in the doclist buffer. */
+    ** pCsr->iPrevId may lie earlier in the doclist buffer. Or, if the
+    ** tree that the node is part of has been marked as EOF, but the node
+    ** itself is not EOF, then it may point to an earlier entry. */
+    pNear = pExpr;
     for(p=pExpr->pParent; p; p=p->pParent){
       if( p->eType==FTSQUERY_OR ) bOr = 1;
+      if( p->eType==FTSQUERY_NEAR ) pNear = p;
+      if( p->bEof ) bTreeEof = 1;
     }
     if( bOr==0 ) return SQLITE_OK;
 
@@ -5295,29 +5303,59 @@ int sqlite3Fts3EvalPhrasePoslist(
       assert( rc!=SQLITE_OK || pPhrase->bIncr==0 );
       if( rc!=SQLITE_OK ) return rc;
     }
-
-    if( pExpr->bEof ){
-      pIter = 0;
-      iDocid = 0;
+    
+    iMul = ((pCsr->bDesc==bDescDoclist) ? 1 : -1);
+    while( bTreeEof==1 
+        && pNear->bEof==0
+        && (DOCID_CMP(pNear->iDocid, pCsr->iPrevId) * iMul)<0
+    ){
+      int rc = SQLITE_OK;
+      fts3EvalNextRow(pCsr, pExpr, &rc);
+      if( rc!=SQLITE_OK ) return rc;
+      iDocid = pExpr->iDocid;
+      pIter = pPhrase->doclist.pList;
     }
+
     bEof = (pPhrase->doclist.nAll==0);
     assert( bDescDoclist==0 || bDescDoclist==1 );
     assert( pCsr->bDesc==0 || pCsr->bDesc==1 );
 
-    if( pCsr->bDesc==bDescDoclist ){
-      int dummy;
-      while( (pIter==0 || DOCID_CMP(iDocid, pCsr->iPrevId)>0 ) && bEof==0 ){
-        sqlite3Fts3DoclistPrev(
-            bDescDoclist, pPhrase->doclist.aAll, pPhrase->doclist.nAll, 
-            &pIter, &iDocid, &dummy, &bEof
-        );
-      }
-    }else{
-      while( (pIter==0 || DOCID_CMP(iDocid, pCsr->iPrevId)<0 ) && bEof==0 ){
-        sqlite3Fts3DoclistNext(
-            bDescDoclist, pPhrase->doclist.aAll, pPhrase->doclist.nAll, 
-            &pIter, &iDocid, &bEof
-        );
+    if( bEof==0 ){
+      if( pCsr->bDesc==bDescDoclist ){
+        int dummy;
+        if( pNear->bEof ){
+          /* This expression is already at EOF. So position it to point to the
+          ** last entry in the doclist at pPhrase->doclist.aAll[]. Variable
+          ** iDocid is already set for this entry, so all that is required is
+          ** to set pIter to point to the first byte of the last position-list
+          ** in the doclist. 
+          **
+          ** It would also be correct to set pIter and iDocid to zero. In
+          ** this case, the first call to sqltie3Fts4DoclistPrev() below
+          ** would also move the iterator to point to the last entry in the 
+          ** doclist. However, this is expensive, as to do so it has to 
+          ** iterate through the entire doclist from start to finish (since
+          ** it does not know the docid for the last entry).  */
+          pIter = &pPhrase->doclist.aAll[pPhrase->doclist.nAll-1];
+          fts3ReversePoslist(pPhrase->doclist.aAll, &pIter);
+        }
+        while( (pIter==0 || DOCID_CMP(iDocid, pCsr->iPrevId)>0 ) && bEof==0 ){
+          sqlite3Fts3DoclistPrev(
+              bDescDoclist, pPhrase->doclist.aAll, pPhrase->doclist.nAll, 
+              &pIter, &iDocid, &dummy, &bEof
+          );
+        }
+      }else{
+        if( pNear->bEof ){
+          pIter = 0;
+          iDocid = 0;
+        }
+        while( (pIter==0 || DOCID_CMP(iDocid, pCsr->iPrevId)<0 ) && bEof==0 ){
+          sqlite3Fts3DoclistNext(
+              bDescDoclist, pPhrase->doclist.aAll, pPhrase->doclist.nAll, 
+              &pIter, &iDocid, &bEof
+          );
+        }
       }
     }
 
diff --git a/manifest b/manifest
index eb27458602..1acf1cc13a 100644
--- a/manifest
+++ b/manifest
@@ -1,5 +1,5 @@
-C Fix\smemory\sleaks\sin\sthe\samatch\sextension.\s\sAdd\sa\sfew\ssimple\stest\scases.
-D 2013-09-30T17:37:15.455
+C Fix\sa\sperformance\sproblem\sin\sthe\sFTS4\sauxiliary\sfunctions\striggered\sby\san\sOR\sclause\sin\sthe\sfull-text\squery.
+D 2013-09-30T18:14:45.059
 F Makefile.arm-wince-mingw32ce-gcc d6df77f1f48d690bd73162294bbba7f59507c72f
 F Makefile.in 5e41da95d92656a5004b03d3576e8b226858a28e
 F Makefile.linux-gcc 91d710bdc4998cb015f39edf3cb314ec4f4d7e23
@@ -78,7 +78,7 @@ F ext/fts3/README.content fdc666a70d5257a64fee209f97cf89e0e6e32b51
 F ext/fts3/README.syntax a19711dc5458c20734b8e485e75fb1981ec2427a
 F ext/fts3/README.tokenizers e0a8b81383ea60d0334d274fadf305ea14a8c314
 F ext/fts3/README.txt 8c18f41574404623b76917b9da66fcb0ab38328d
-F ext/fts3/fts3.c 0d6311cd433ea30c9e941b93bfeac2f9e6937980
+F ext/fts3/fts3.c e1240ab6f5999174309a41ffac63b94ed1233098
 F ext/fts3/fts3.h 3a10a0af180d502cecc50df77b1b22df142817fe
 F ext/fts3/fts3Int.h c7a451661c2d9b2440b2008c3f63ce06f13181d6
 F ext/fts3/fts3_aux.c b02632f6dd0e375ce97870206d914ea6d8df5ccd
@@ -548,7 +548,7 @@ F test/fts3prefix2.test e1f0a822ca661dced7f12ce392e14eaf65609dce
 F test/fts3query.test 4fefd43ff24993bc2c9b2778f2bec0cc7629e7ed
 F test/fts3rnd.test 1320d8826a845e38a96e769562bf83d7a92a15d0
 F test/fts3shared.test 57e26a801f21027b7530da77db54286a6fe4997e
-F test/fts3snippet.test 24d6ff1920a70fd970c401a8525834b4ad12cece
+F test/fts3snippet.test d524af6bcef4714e059ef559113dbdc924cd33d1
 F test/fts3sort.test ed34c716a11cc2009a35210e84ad5f9c102362ca
 F test/fts3tok1.test b10d0a12a0ab5f905cea1200b745de233f37443f
 F test/fts3tok_err.test 52273cd193b9036282f7bacb43da78c6be87418d
@@ -1117,7 +1117,7 @@ F tool/warnings-clang.sh f6aa929dc20ef1f856af04a730772f59283631d4
 F tool/warnings.sh fbc018d67fd7395f440c28f33ef0f94420226381
 F tool/wherecosttest.c f407dc4c79786982a475261866a161cd007947ae
 F tool/win/sqlite.vsix 030f3eeaf2cb811a3692ab9c14d021a75ce41fff
-P 94c914e3fa632f88a0d0c14537f81aa46759e2be
-R 365ba09c87d852154921d1b9e7f8c208
-U drh
-Z 32444bec405923a830b81223af98656e
+P 604134732e309a738e21bcb4240b9ff285d2bfe4
+R 6a5e216c5c51e185221e1c2d8a375388
+U dan
+Z 83a6f701584057aa6ad5500f631da9c2
diff --git a/manifest.uuid b/manifest.uuid
index 3aed65bfb3..3f571c4238 100644
--- a/manifest.uuid
+++ b/manifest.uuid
@@ -1 +1 @@
-604134732e309a738e21bcb4240b9ff285d2bfe4
\ No newline at end of file
+fa0f2f0e3e79ae653118b901e1cca7725dfaf249
\ No newline at end of file
diff --git a/test/fts3snippet.test b/test/fts3snippet.test
index f19db925ba..415251dcce 100644
--- a/test/fts3snippet.test
+++ b/test/fts3snippet.test
@@ -16,6 +16,7 @@
 
 set testdir [file dirname $argv0]
 source $testdir/tester.tcl
+set testprefix fts3snippet
 
 # If SQLITE_ENABLE_FTS3 is not defined, omit this file.
 ifcapable !fts3 { finish_test ; return }
@@ -138,7 +139,7 @@ foreach {DO_MALLOC_TEST enc} {
 
   # Set variable $T to the test name prefix for this iteration of the loop.
   #
-  set T "fts3snippet-$enc"
+  set T "fts3snippet-1.$enc"
 
   ##########################################################################
   # Test the offset function.
@@ -459,5 +460,65 @@ foreach {DO_MALLOC_TEST enc} {
   } {0 blob}
 }
 
+#-------------------------------------------------------------------------
+# Test an interaction between the snippet() function and OR clauses.
+#
+do_execsql_test 2.1 {
+  CREATE VIRTUAL TABLE t2 USING fts4;
+  INSERT INTO t2 VALUES('one two three four five');
+  INSERT INTO t2 VALUES('two three four five one');
+  INSERT INTO t2 VALUES('three four five one two');
+  INSERT INTO t2 VALUES('four five one two three');
+  INSERT INTO t2 VALUES('five one two three four');
+}
+
+do_execsql_test 2.2 {
+  SELECT snippet(t2, '[', ']') FROM t2 WHERE t2 MATCH 'one OR (four AND six)'
+} {
+  {[one] two three [four] five}
+  {two three [four] five [one]}
+  {three [four] five [one] two}
+  {[four] five [one] two three}
+  {five [one] two three [four]}
+}
+
+do_execsql_test 2.3 {
+  SELECT snippet(t2, '[', ']') FROM t2 
+  WHERE t2 MATCH 'one OR (four AND six)' 
+  ORDER BY docid DESC
+} {
+  {five [one] two three [four]}
+  {[four] five [one] two three}
+  {three [four] five [one] two}
+  {two three [four] five [one]}
+  {[one] two three [four] five}
+}
+
+do_execsql_test 2.4 {
+  INSERT INTO t2 VALUES('six');
+}
+
+do_execsql_test 2.5 {
+  SELECT snippet(t2, '[', ']') FROM t2 WHERE t2 MATCH 'one OR (four AND six)'
+} {
+  {[one] two three [four] five}
+  {two three [four] five [one]}
+  {three [four] five [one] two}
+  {[four] five [one] two three}
+  {five [one] two three [four]}
+}
+
+do_execsql_test 2.6 {
+  SELECT snippet(t2, '[', ']') FROM t2 
+  WHERE t2 MATCH 'one OR (four AND six)' 
+  ORDER BY docid DESC
+} {
+  {five [one] two three [four]}
+  {[four] five [one] two three}
+  {three [four] five [one] two}
+  {two three [four] five [one]}
+  {[one] two three [four] five}
+}
+
 set sqlite_fts3_enable_parentheses 0
 finish_test