Fix some fts5 problems with very large position lists.

FossilOrigin-Name: 2ea8f9cbe67dac60c1a0a661c95a03ecfa9a0b9a

4 files changed, 88 insertions, 13 deletions

diff --git a/ext/fts5/fts5_index.c b/ext/fts5/fts5_index.c
index d407411a2..7d48d074d 100644
--- a/ext/fts5/fts5_index.c
+++ b/ext/fts5/fts5_index.c
@@ -1901,13 +1901,20 @@ static void fts5SegIterNext(
 ** the doclist.
 */
 static void fts5SegIterReverse(Fts5Index *p, int iIdx, Fts5SegIter *pIter){
+  Fts5DlidxIter *pDlidx = pIter->pDlidx;
   Fts5Data *pLast = 0;
   int pgnoLast = 0;
 
-  if( pIter->pDlidx ){
-    int iSegid = pIter->pSeg->iSegid;
-    pgnoLast = pIter->pDlidx->iLeafPgno;
-    pLast = fts5DataRead(p, FTS5_SEGMENT_ROWID(iIdx, iSegid, 0, pgnoLast));
+  if( pDlidx ){
+    /* If the doclist-iterator is already at EOF, then the current doclist
+    ** contains no entries except those on the current page. */
+    if( fts5DlidxIterEof(p, pDlidx)==0 ){
+      int iSegid = pIter->pSeg->iSegid;
+      pgnoLast = pDlidx->iLeafPgno;
+      pLast = fts5DataRead(p, FTS5_SEGMENT_ROWID(iIdx, iSegid, 0, pgnoLast));
+    }else{
+      pIter->iLeafOffset -= sqlite3Fts5GetVarintLen(pIter->nPos*2+pIter->bDel);
+    }
   }else{
     int iOff;                               /* Byte offset within pLeaf */
     Fts5Data *pLeaf = pIter->pLeaf;         /* Current leaf data */
@@ -1915,7 +1922,7 @@ static void fts5SegIterReverse(Fts5Index *p, int iIdx, Fts5SegIter *pIter){
     /* Currently, Fts5SegIter.iLeafOffset (and iOff) points to the first 
     ** byte of position-list content for the current rowid. Back it up
     ** so that it points to the start of the position-list size field. */
-    pIter->iLeafOffset -= sqlite3Fts5GetVarintLen(pIter->nPos*2 + pIter->bDel);
+    pIter->iLeafOffset -= sqlite3Fts5GetVarintLen(pIter->nPos*2+pIter->bDel);
     iOff = pIter->iLeafOffset;
     assert( iOff>=4 );
 
@@ -3285,6 +3292,9 @@ static void fts5TrimSegments(Fts5Index *p, Fts5MultiSegIter *pIter){
     if( pSeg->pSeg==0 ){
       /* no-op */
     }else if( pSeg->pLeaf==0 ){
+      /* All keys from this input segment have been transfered to the output.
+      ** Set both the first and last page-numbers to 0 to indicate that the
+      ** segment is now empty. */
       pSeg->pSeg->pgnoLast = 0;
       pSeg->pSeg->pgnoFirst = 0;
     }else{
@@ -4092,7 +4102,13 @@ static void fts5IndexIntegrityCheckSegment(
     }
   }
 
-  if( p->rc==SQLITE_OK && iter.iLeaf!=pSeg->pgnoLast ){
+  /* Either iter.iLeaf must be the rightmost leaf-page in the segment, or 
+  ** else the segment has been completely emptied by an ongoing merge
+  ** operation. */
+  if( p->rc==SQLITE_OK 
+   && iter.iLeaf!=pSeg->pgnoLast 
+   && (pSeg->pgnoFirst || pSeg->pgnoLast) 
+  ){
     p->rc = FTS5_CORRUPT;
   }
 
diff --git a/ext/fts5/test/fts5bigpl.test b/ext/fts5/test/fts5bigpl.test
new file mode 100644
index 000000000..172c0396b
--- /dev/null
+++ b/ext/fts5/test/fts5bigpl.test
@@ -0,0 +1,58 @@
+# 2015 April 21
+#
+# The author disclaims copyright to this source code.  In place of
+# a legal notice, here is a blessing:
+#
+#    May you do good and not evil.
+#    May you find forgiveness for yourself and forgive others.
+#    May you share freely, never taking more than you give.
+#
+#***********************************************************************
+#
+# This test is focused on really large position lists. Those that require
+# 4 or 5 byte position-list size varints. Because of the amount of memory
+# required, these tests only run on 64-bit platforms.
+#
+
+source [file join [file dirname [info script]] fts5_common.tcl]
+set testprefix fts5bigpl
+
+if { $tcl_platform(wordSize)<8 } {
+  finish_test
+  return
+}
+
+do_execsql_test 1.0 { CREATE VIRTUAL TABLE t1 USING fts5(x) }
+
+do_test 1.1 {
+  foreach t {a b c d e f g h i j} {
+    set doc [string repeat "$t " 1200000]
+    execsql { INSERT INTO t1 VALUES($doc) }
+  }
+  execsql { INSERT INTO t1(t1) VALUES('integrity-check') }
+} {}
+
+do_test 1.2 {
+  execsql { DELETE FROM t1 }
+  foreach t {"a b" "b a" "c d" "d c"} {
+    set doc [string repeat "$t " 600000]
+    execsql { INSERT INTO t1 VALUES($doc) }
+  }
+  execsql { INSERT INTO t1(t1) VALUES('integrity-check') }
+} {}
+
+
+# 5-byte varint. This test takes 30 seconds or so on a 2014 workstation.
+# The generated database is roughly 635MiB.
+#
+do_test 2.1...slow {
+  execsql { DELETE FROM t1 }
+  foreach t {a} {
+    set doc [string repeat "$t " 150000000]
+    execsql { INSERT INTO t1 VALUES($doc) }
+  }
+  execsql { INSERT INTO t1(t1) VALUES('integrity-check') }
+} {}
+
+finish_test
+
diff --git a/manifest b/manifest
index 994b10599..cb1f4c510 100644
--- a/manifest
+++ b/manifest
@@ -1,5 +1,5 @@
-C Logically\sstore\supdates\sas\s(insert+delete)\swithin\sthe\sFTS\stree.\sThis\sallows\skeys\sto\sbe\sannihilated\smore\squickly\sunder\ssome\scircumstances.
-D 2015-04-15T18:49:20.008
+C Fix\ssome\sfts5\sproblems\swith\svery\slarge\sposition\slists.
+D 2015-04-20T18:48:57.780
 F Makefile.arm-wince-mingw32ce-gcc d6df77f1f48d690bd73162294bbba7f59507c72f
 F Makefile.in 88a3e6261286db378fdffa1124cad11b3c05f5bb
 F Makefile.linux-gcc 91d710bdc4998cb015f39edf3cb314ec4f4d7e23
@@ -112,7 +112,7 @@ F ext/fts5/fts5_buffer.c 3ba56cc6824c9f7b1e0695159e0a9c636f6b4a23
 F ext/fts5/fts5_config.c 0847facc8914f57ea4452c43ce109200dc65e894
 F ext/fts5/fts5_expr.c 5215137efab527577d36bdf9e44bfc2ec3e1be98
 F ext/fts5/fts5_hash.c 3cb5a3d04dd2030eb0ac8d544711dfd37c0e6529
-F ext/fts5/fts5_index.c 28f1bfadf3eb4f860c8b978f4d8d6ea0cf7c724d
+F ext/fts5/fts5_index.c f840e35cceafcd0597688467010a4d12feea9c76
 F ext/fts5/fts5_storage.c ac0f0937059c8d4f38a1f13aa5f2c2cd7edf3e0d
 F ext/fts5/fts5_tcl.c 617b6bb96545be8d9045de6967c688cd9cd15541
 F ext/fts5/fts5_tokenize.c c07f2c2f749282c1dbbf46bde1f6d7095c740b8b
@@ -133,6 +133,7 @@ F ext/fts5/test/fts5aj.test 05b569f5c16ea3098fb1984eec5cf50dbdaae5d8
 F ext/fts5/test/fts5ak.test 7b8c5df96df599293f920b7e5521ebc79f647592
 F ext/fts5/test/fts5al.test 6a5717faaf7f1e0e866360022d284903f3a4eede
 F ext/fts5/test/fts5auxdata.test c69b86092bf1a157172de5f9169731af3403179b
+F ext/fts5/test/fts5bigpl.test b1cfd00561350ab04994ba7dd9d48468e5e0ec3b
 F ext/fts5/test/fts5content.test 8dc302fccdff834d946497e9d862750ea87d4517
 F ext/fts5/test/fts5corrupt.test dbdcfe75749ed2f2eb3915cf68fd55d3dc3b058d
 F ext/fts5/test/fts5ea.test 04695560a444fcc00c3c4f27783bdcfbf71f030c
@@ -1292,7 +1293,7 @@ F tool/vdbe_profile.tcl 67746953071a9f8f2f668b73fe899074e2c6d8c1
 F tool/warnings-clang.sh f6aa929dc20ef1f856af04a730772f59283631d4
 F tool/warnings.sh 0abfd78ceb09b7f7c27c688c8e3fe93268a13b32
 F tool/win/sqlite.vsix deb315d026cc8400325c5863eef847784a219a2f
-P b29109a083e5cd442cdd19f29d7be45b09c4f661
-R 3f3d372bbefa9c63dd6c3e57000a1fa8
+P 50fae1f0006c0e946b5214e73eedf2687a0016f9
+R e3e266d7af429931ea61b2ad868bf28e
 U dan
-Z ff9a464604d7627b634f349ee851852e
+Z 88f214a9049d68201f885f825375d535
diff --git a/manifest.uuid b/manifest.uuid
index ab0ef8daa..6b9e7e2cd 100644
--- a/manifest.uuid
+++ b/manifest.uuid
@@ -1 +1 @@
-50fae1f0006c0e946b5214e73eedf2687a0016f9
\ No newline at end of file
+2ea8f9cbe67dac60c1a0a661c95a03ecfa9a0b9a
\ No newline at end of file