From: shess <shess@noemail.net>
Date: Tue, 22 Jul 2008 23:41:26 +0000 (+0000)
Subject: Delete all fts2 index data the table becomes empty.
X-Git-Tag: version-3.6.10~720
X-Git-Url: http://git.ipfire.org/?a=commitdiff_plain;h=08904673c8fc8b6b5be50c50a411ebe458cd2e50;p=thirdparty%2Fsqlite.git

Delete all fts2 index data the table becomes empty.
Backports check-in (5413) from fts3. (CVS 5457)

FossilOrigin-Name: 4c98179be258319f441ae4e123cf59af77e96409
---

diff --git a/ext/fts2/fts2.c b/ext/fts2/fts2.c
index a7a6fc86b0..524e1d83d4 100644
--- a/ext/fts2/fts2.c
+++ b/ext/fts2/fts2.c
@@ -1774,10 +1774,12 @@ typedef enum fulltext_statement {
   CONTENT_SELECT_STMT,
   CONTENT_UPDATE_STMT,
   CONTENT_DELETE_STMT,
+  CONTENT_EXISTS_STMT,
 
   BLOCK_INSERT_STMT,
   BLOCK_SELECT_STMT,
   BLOCK_DELETE_STMT,
+  BLOCK_DELETE_ALL_STMT,
 
   SEGDIR_MAX_INDEX_STMT,
   SEGDIR_SET_STMT,
@@ -1786,6 +1788,7 @@ typedef enum fulltext_statement {
   SEGDIR_DELETE_STMT,
   SEGDIR_SELECT_SEGMENT_STMT,
   SEGDIR_SELECT_ALL_STMT,
+  SEGDIR_DELETE_ALL_STMT,
 
   MAX_STMT                     /* Always at end! */
 } fulltext_statement;
@@ -1800,10 +1803,12 @@ static const char *const fulltext_zStatement[MAX_STMT] = {
   /* CONTENT_SELECT */ "select * from %_content where rowid = ?",
   /* CONTENT_UPDATE */ NULL,  /* generated in contentUpdateStatement() */
   /* CONTENT_DELETE */ "delete from %_content where rowid = ?",
+  /* CONTENT_EXISTS */ "select rowid from %_content limit 1",
 
   /* BLOCK_INSERT */ "insert into %_segments values (?)",
   /* BLOCK_SELECT */ "select block from %_segments where rowid = ?",
   /* BLOCK_DELETE */ "delete from %_segments where rowid between ? and ?",
+  /* BLOCK_DELETE_ALL */ "delete from %_segments",
 
   /* SEGDIR_MAX_INDEX */ "select max(idx) from %_segdir where level = ?",
   /* SEGDIR_SET */ "insert into %_segdir values (?, ?, ?, ?, ?, ?)",
@@ -1824,6 +1829,7 @@ static const char *const fulltext_zStatement[MAX_STMT] = {
   /* SEGDIR_SELECT_ALL */
   "select start_block, leaves_end_block, root from %_segdir "
   " order by level desc, idx asc",
+  /* SEGDIR_DELETE_ALL */ "delete from %_segdir",
 };
 
 /*
@@ -2100,6 +2106,25 @@ static int content_delete(fulltext_vtab *v, sqlite_int64 iRow){
   return sql_single_step(s);
 }
 
+/* Returns SQLITE_ROW if any rows exist in %_content, SQLITE_DONE if
+** no rows exist, and any error in case of failure.
+*/
+static int content_exists(fulltext_vtab *v){
+  sqlite3_stmt *s;
+  int rc = sql_get_statement(v, CONTENT_EXISTS_STMT, &s);
+  if( rc!=SQLITE_OK ) return rc;
+
+  rc = sqlite3_step(s);
+  if( rc!=SQLITE_ROW ) return rc;
+
+  /* We expect only one row.  We must execute another sqlite3_step()
+   * to complete the iteration; otherwise the table will remain locked. */
+  rc = sqlite3_step(s);
+  if( rc==SQLITE_DONE ) return SQLITE_ROW;
+  if( rc==SQLITE_ROW ) return SQLITE_ERROR;
+  return rc;
+}
+
 /* insert into %_segments values ([pData])
 **   returns assigned rowid in *piBlockid
 */
@@ -2273,6 +2298,23 @@ static int segdir_delete(fulltext_vtab *v, int iLevel){
   return sql_single_step(s);
 }
 
+/* Delete entire fts index, SQLITE_OK on success, relevant error on
+** failure.
+*/
+static int segdir_delete_all(fulltext_vtab *v){
+  sqlite3_stmt *s;
+  int rc = sql_get_statement(v, SEGDIR_DELETE_ALL_STMT, &s);
+  if( rc!=SQLITE_OK ) return rc;
+
+  rc = sql_single_step(s);
+  if( rc!=SQLITE_OK ) return rc;
+
+  rc = sql_get_statement(v, BLOCK_DELETE_ALL_STMT, &s);
+  if( rc!=SQLITE_OK ) return rc;
+
+  return sql_single_step(s);
+}
+
 /* TODO(shess) clearPendingTerms() is far down the file because
 ** writeZeroSegment() is far down the file because LeafWriter is far
 ** down the file.  Consider refactoring the code to move the non-vtab
@@ -5775,6 +5817,23 @@ static int fulltextUpdate(sqlite3_vtab *pVtab, int nArg, sqlite3_value **ppArg,
 
   if( nArg<2 ){
     rc = index_delete(v, sqlite3_value_int64(ppArg[0]));
+    if( rc==SQLITE_OK ){
+      /* If we just deleted the last row in the table, clear out the
+      ** index data.
+      */
+      rc = content_exists(v);
+      if( rc==SQLITE_ROW ){
+        rc = SQLITE_OK;
+      }else if( rc==SQLITE_DONE ){
+        /* Clear the pending terms so we don't flush a useless level-0
+        ** segment when the transaction closes.
+        */
+        rc = clearPendingTerms(v);
+        if( rc==SQLITE_OK ){
+          rc = segdir_delete_all(v);
+        }
+      }
+    }
   } else if( sqlite3_value_type(ppArg[0]) != SQLITE_NULL ){
     /* An update:
      * ppArg[0] = old rowid
diff --git a/manifest b/manifest
index 7ea48d4c7e..f373014583 100644
--- a/manifest
+++ b/manifest
@@ -1,5 +1,5 @@
-C fts2\sfunctions\sfor\stesting\sscripts.\r\nBackports\s(5340)\sfrom\sfts3.\s(CVS\s5456)
-D 2008-07-22T23:32:28
+C Delete\sall\sfts2\sindex\sdata\sthe\stable\sbecomes\sempty.\r\nBackports\scheck-in\s(5413)\sfrom\sfts3.\s(CVS\s5457)
+D 2008-07-22T23:41:26
 F Makefile.arm-wince-mingw32ce-gcc fcd5e9cd67fe88836360bb4f9ef4cb7f8e2fb5a0
 F Makefile.in 77ff156061bb870aa0a8b3d545c670d08070f7e6
 F Makefile.linux-gcc d53183f4aa6a9192d249731c90dbdffbd2c68654
@@ -39,7 +39,7 @@ F ext/fts1/simple_tokenizer.c 1844d72f7194c3fd3d7e4173053911bf0661b70d
 F ext/fts1/tokenizer.h 0c53421b832366d20d720d21ea3e1f6e66a36ef9
 F ext/fts2/README.tokenizers 21e3684ea5a095b55d70f6878b4ce6af5932dfb7
 F ext/fts2/README.txt 8c18f41574404623b76917b9da66fcb0ab38328d
-F ext/fts2/fts2.c af6d11365c4ae66be9779dde0948887e92d8b867
+F ext/fts2/fts2.c c0d287669e35e3f0998b42fb06ebd2f89e5b2593
 F ext/fts2/fts2.h da5f76c65163301d1068a971fd32f4119e3c95fa
 F ext/fts2/fts2_hash.c 2689e42e1107ea67207f725cf69cf8972d00cf93
 F ext/fts2/fts2_hash.h 9a5b1be94664139f93217a0770d7144425cffb3a
@@ -316,6 +316,7 @@ F test/fts2m.test 4b30142ead6f3ed076e880a2a464064c5ad58c51
 F test/fts2n.test 12b9c5352128cebd1c6b8395e43788d4b09087c2
 F test/fts2o.test c6a79567d85403dc4d15b89f3f9799a0a0aef065
 F test/fts2p.test 4b48c35c91e6a7dbf5ac8d1e5691823cc999aafb
+F test/fts2q.test ee2cb138a599a07de3a03cd7a71a4e9e40120f93
 F test/fts2token.test d8070b241a15ff13592a9ae4a8b7c171af6f445a
 F test/fts3.test 6ee4c38b0864583c80e82a2d4372f63aae8b10c7
 F test/fts3aa.test 432d1d5c41939bb5405d4d6c80a9ec759b363393
@@ -610,7 +611,7 @@ F tool/speedtest16.c c8a9c793df96db7e4933f0852abb7a03d48f2e81
 F tool/speedtest2.tcl ee2149167303ba8e95af97873c575c3e0fab58ff
 F tool/speedtest8.c 1dbced29de5f59ba2ebf877edcadf171540374d1
 F tool/speedtest8inst1.c 293327bc76823f473684d589a8160bde1f52c14e
-P 3f614453d2d7c753a5963b027fe8618b50b4f6b9
-R 76720fcdb90c8738e414944367c3b0e7
+P 4e47394be9dfbf0f9309e55eb6c6a3a517ea2006
+R 47eed6cf50908f452be50506a6925d0d
 U shess
-Z 807f812f3e691d0e714ba8b89e45124c
+Z f9e0e9386065f5f55b31594485ae0dd8
diff --git a/manifest.uuid b/manifest.uuid
index c80705daff..7800c6c661 100644
--- a/manifest.uuid
+++ b/manifest.uuid
@@ -1 +1 @@
-4e47394be9dfbf0f9309e55eb6c6a3a517ea2006
\ No newline at end of file
+4c98179be258319f441ae4e123cf59af77e96409
\ No newline at end of file
diff --git a/test/fts2q.test b/test/fts2q.test
new file mode 100644
index 0000000000..fdce634a87
--- /dev/null
+++ b/test/fts2q.test
@@ -0,0 +1,128 @@
+# 2008 June 26
+#
+# 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 file implements regression tests for SQLite library.  The focus
+# of this script is testing the FTS2 module's optimize() function.
+#
+# $Id: fts2q.test,v 1.1 2008/07/22 23:41:26 shess Exp $
+#
+
+set testdir [file dirname $argv0]
+source $testdir/tester.tcl
+
+# If SQLITE_ENABLE_FTS2 is not defined, omit this file.
+ifcapable !fts2 {
+  finish_test
+  return
+}
+
+#*************************************************************************
+# Probe to see if support for the FTS2 dump_* functions is compiled in.
+# TODO(shess): Change main.mk to do the right thing and remove this test.
+db eval {
+  DROP TABLE IF EXISTS t1;
+  CREATE VIRTUAL TABLE t1 USING fts2(c);
+  INSERT INTO t1 (rowid, c) VALUES (1, 'x');
+}
+
+set s {SELECT dump_terms(t1, 1) FROM t1 LIMIT 1}
+set r {1 {unable to use function dump_terms in the requested context}}
+if {[catchsql $s]==$r} {
+  finish_test
+  return
+}
+
+#*************************************************************************
+# Utility function to check for the expected terms in the segment
+# level/index.  _all version does same but for entire index.
+proc check_terms {test level index terms} {
+  # TODO(shess): Figure out why uplevel in do_test can't catch
+  # $level and $index directly.
+  set ::level $level
+  set ::index $index
+  do_test $test.terms {
+    execsql {
+      SELECT dump_terms(t1, $::level, $::index) FROM t1 LIMIT 1;
+    }
+  } [list $terms]
+}
+proc check_terms_all {test terms} {
+  do_test $test.terms {
+    execsql {
+      SELECT dump_terms(t1) FROM t1 LIMIT 1;
+    }
+  } [list $terms]
+}
+
+# Utility function to check for the expected doclist for the term in
+# segment level/index.  _all version does same for entire index.
+proc check_doclist {test level index term doclist} {
+  # TODO(shess): Again, why can't the non-:: versions work?
+  set ::term $term
+  set ::level $level
+  set ::index $index
+  do_test $test {
+    execsql {
+      SELECT dump_doclist(t1, $::term, $::level, $::index) FROM t1 LIMIT 1;
+    }
+  } [list $doclist]
+}
+proc check_doclist_all {test term doclist} {
+  set ::term $term
+  do_test $test {
+    execsql {
+      SELECT dump_doclist(t1, $::term) FROM t1 LIMIT 1;
+    }
+  } [list $doclist]
+}
+
+#*************************************************************************
+# Test results when all rows are deleted and one is added back.
+# Previously older segments would continue to exist, but now the index
+# should be dropped when the table is empty.  The results should look
+# exactly like we never added the earlier rows in the first place.
+db eval {
+  DROP TABLE IF EXISTS t1;
+  CREATE VIRTUAL TABLE t1 USING fts2(c);
+  INSERT INTO t1 (rowid, c) VALUES (1, 'This is a test');
+  INSERT INTO t1 (rowid, c) VALUES (2, 'That was a test');
+  INSERT INTO t1 (rowid, c) VALUES (3, 'This is a test');
+  DELETE FROM t1 WHERE 1=1; -- Delete each row rather than dropping table.
+  INSERT INTO t1 (rowid, c) VALUES (1, 'This is a test');
+}
+
+# Should be a single initial segment.
+do_test fts2q-1.segments {
+  execsql {
+    SELECT level, idx FROM t1_segdir ORDER BY level, idx;
+  }
+} {0 0}
+do_test fts2q-1.matches {
+  execsql {
+    SELECT OFFSETS(t1) FROM t1
+     WHERE t1 MATCH 'this OR that OR was OR a OR is OR test' ORDER BY rowid;
+  }
+} {{0 0 0 4 0 4 5 2 0 3 8 1 0 5 10 4}}
+
+check_terms_all fts2q-1.1 {a is test this}
+check_doclist_all fts2q-1.1.1 a {[1 0[2]]}
+check_doclist_all fts2q-1.1.2 is {[1 0[1]]}
+check_doclist_all fts2q-1.1.3 test {[1 0[3]]}
+check_doclist_all fts2q-1.1.4 this {[1 0[0]]}
+
+check_terms   fts2q-1.2   0 0 {a is test this}
+check_doclist fts2q-1.2.1 0 0 a {[1 0[2]]}
+check_doclist fts2q-1.2.2 0 0 is {[1 0[1]]}
+check_doclist fts2q-1.2.3 0 0 test {[1 0[3]]}
+check_doclist fts2q-1.2.4 0 0 this {[1 0[0]]}
+
+# TODO(shess): optimize() tests here.
+
+finish_test