SQLite

Timeline
Login

Many hyperlinks are disabled.
Use anonymous login to enable hyperlinks.

1109 check-ins using file ext/fts5/fts5parse.y version e51b3754

2019-09-03
19:40
Fix a buffer overread that could occur when running fts5 prefix queries inside a transaction. (Leaf check-in: b54aa18b user: dan tags: branch-3.18)
17:46
Disable the undocumented rtreenode() SQL function that is only used for testing, except when doing a build that is specifically intended for testing. (check-in: 0a1cce49 user: drh tags: branch-3.18)
2018-12-19
01:38
Add extra defenses against strategically corrupt databases to fts3/4. (check-in: 4bb21d82 user: drh tags: branch-3.18)
2017-07-21
07:56
Add new interfaces sqlite3_result_pointer(), and sqlite3_value_pointer() and use them to transfer the eponymous FTS3 column pointer to the snippet() and offsets() routines. This changes is a cherry-pick of [f0f49224] via [69906880cee] and is ultimately inspired by check-in [72de49f2]. (check-in: 8201f4e1 user: drh tags: branch-3.18)
2017-07-15
10:44
Add new interfaces sqlite3_bind_pointer(), sqlite3_result_pointer(), and sqlite3_value_pointer() used to safely move pointer values through SQL without exposing underlying memory address information. (Leaf check-in: d6a44b35 user: drh tags: branch-3.17)
2017-06-17
09:59
Version 3.18.2 (check-in: 036ebf72 user: drh tags: release, branch-3.18, version-3.18.2)
00:39
When building an ephemeral table of integers, do not assume that the table does not already contain N if the insert cursor points to N-1. Fix for ticket [61fe9745]. Also increase the version number to 3.18.2. This is the release candidate. (check-in: e1b71029 user: drh tags: branch-3.18)
2017-06-16
13:41
Version 3.18.1 (check-in: 77bb4623 user: drh tags: release, branch-3.18, version-3.18.1)
12:15
An unintentional fork was created due to a typo when creating branch-3.18. This check-in resolves the fork and is the 3.18.1 release candidate. (check-in: 20e5ffb1 user: drh tags: branch-3.18)
2017-06-15
07:47
Increase the version number to 3.18.1. (check-in: 4be2eba8 user: drh tags: branch-3.18)
2017-06-12
23:44
Ensure pointer map entries are always added when a row that does use overflow pages replaces one that does not in an auto-vacuum database. Fix for [fda22108]. (check-in: 30c50f0e user: drh tags: branch-3.18)
2017-04-12
17:38
Improved \n and \r escapes in the ext/misc/dbdump.c utility function. The implementation of dbdump.c now matches the implementation in the CLI. (check-in: f2643315 user: drh tags: trunk)
2017-04-11
20:48
Avoid updating unaffected indexes on a table as part of an UPDATE that requires foreign key processing in some cases. (check-in: 7aae5c0f user: dan tags: trunk)
19:58
Avoid updating unaffected indexes on a table as part of an UPDATE that requires foreign key processing in some cases. (Closed-Leaf check-in: 477bea9e user: dan tags: fkey-optimization)
19:00
Update this branch with latest trunk changes. (check-in: 0f66a093 user: dan tags: schemalint)
18:55
Limit the depth of recursion for valid JSON in the JSON1 extension in order to avoid using excess stack space in the recursive descent parser. Fix for ticket [981329adeef51011052667a9]. (check-in: 1f68c184 user: drh tags: trunk)
18:29
Fix a formatting issue in the output of the sqlite3_expert program. (check-in: cc8c3581 user: dan tags: schemalint)
18:06
Smaller and faster implementation of exprMightBeIndexed(). (check-in: 76cd611d user: drh tags: trunk)
17:43
Add header comments to the API functions in sqlite3expert.h. Include a list of all candidate indexes in the report output by the sqlite3_expert program. (check-in: 0c45c5eb user: dan tags: schemalint)
16:44
Very slight smaller and faster sqlite3SelectNew() (check-in: 4143650c user: drh tags: trunk)
12:20
Add an ALWAYS() around an unreachable condition in sqlite3VdbeMemGrow(). (check-in: 0f3eb61f user: drh tags: trunk)
11:52
Fix a segfault that could occur if an indexed expression was used in a comparison operation within the result-set of a SELECT statement. (check-in: d6bb7c42 user: dan tags: trunk)
01:30
When evaluating a query that uses an index on an expression, try to use the expression values read directly from the index rather than reevaluating the expression. (check-in: a47efb7c user: drh tags: trunk)
01:01
Improved comments. Fix a problem when an indexed expression is used in an ORDER BY clause. (Closed-Leaf check-in: c59eaf2b user: drh tags: covering-index-on-expr)
2017-04-10
23:42
Merge changes from trunk. (check-in: 8978465f user: drh tags: covering-index-on-expr)
20:51
Smaller and faster vdbeFreeOpArray() (check-in: e052436d user: drh tags: trunk)
20:27
Slightly smaller and faster sqlite3VdbeMemGrow(). (check-in: efd1702a user: drh tags: trunk)
20:00
Add ext/expert/README.md. (check-in: 9318f1b9 user: dan tags: schemalint)
16:13
Changes to allow the code in sqlite3expert.c to be tested directly (via the API in sqlite3expert.h) instead of by invoking the sqlite3_expert application. Fix memory leaks and other problems. (check-in: 5dd98317 user: dan tags: schemalint)
12:31
Add a new JSON test case to verify that all control characters are escaped in the json_quote() function. (check-in: 6ee12221 user: drh tags: trunk)
12:25
Change the JSON extension so that it disallows control characters inside of strings. Fix for ticket [6c9b5514077fed34551f98e64c09a10dc2fc8e16]. (check-in: 475d8f82 user: drh tags: trunk)
2017-04-09
19:23
Do not expose the name of the internal Mem object in the public interface defined by sqlite3.h. (check-in: 19dd753f user: drh tags: trunk)
08:38
Fix the -file option on the sqlite3_expert program. (check-in: 0857c48e user: dan tags: schemalint)
2017-04-08
19:00
Merge latest trunk with this branch. (check-in: e051e8f2 user: dan tags: schemalint)
18:56
Rename shell6.test to expert1.test. Have it invoke the sqlite3_expert binary if it is present. (check-in: be0deff9 user: dan tags: schemalint)
18:24
Merge all recent changes from trunk. (check-in: c23ef5e2 user: drh tags: apple-osx)
18:18
Disallow leading zeros on numeric constants in JSON. Fix for ticket [b93be8729a895a528e2] (check-in: 204e72f0 user: drh tags: trunk)
17:41
Use hash tables instead of in-memory database tables for a few purposes in sqlite3expert.c. (check-in: bf10e68d user: dan tags: schemalint)
14:11
Expand on the comment above OP_Destroy to explain why it throws an error if there are any active reader VMs. (check-in: b9a8c2b9 user: dan tags: trunk)
13:52
Have the rtree module close any open blob-handle within the xSavepoint method. This prevents such an open blob handle from interfering with DROP TABLE operations. (check-in: fa4416ad user: dan tags: trunk)
13:42
Fix the quoting mechanism for ".dump" so that it is not applied for the ".mode quote" output. (check-in: 78c1e903 user: drh tags: trunk)
09:12
Have fts5 close any open blob-handle when a new savepoint is opened. This ensures that fts5 does not prevent DROP TABLE statements (which always open a savepoint) from succeeding. (check-in: a921ada8 user: dan tags: trunk)
01:09
Make sure the RTree sqlite3_blob handle is reset prior to renaming the table. (check-in: 1cdae2db user: drh tags: trunk)
00:55
Fix the ".dump" command so that it works with ".headers on". Also fix the display of ".mode insert" with ".headers on". (check-in: a6ce57ee user: drh tags: trunk)
00:25
Merge the latest trunk changes. (check-in: dd16439e user: drh tags: apple-osx)
2017-04-07
20:20
Use replace() instead of char() to quote newline and return characters in strings in the output of .dump, to avoid excess expression complexity. (check-in: 73073529 user: drh tags: trunk)
20:14
Refactor code to suggest indexes from the shell tool into an extension in ext/expert. Unfinished. (check-in: 305e19f9 user: dan tags: schemalint)
19:52
Remove an unused token type. (check-in: 13a42223 user: drh tags: covering-index-on-expr)
19:41
Proof of concept for the ability to use the expression columns in an index on expressions in place of equivalent expressions in the result set or in the WHERE clause. This check-in compiles but is mostly untested. (check-in: a52ef2ad user: drh tags: covering-index-on-expr)
11:45
Use replace() instead of char() to quote newline and return characters in strings in the output of .dump, to avoid excess expression complexity. (Closed-Leaf check-in: 4c2b5729 user: drh tags: shell-fix)
2017-04-06
18:44
Changes to allow indexes to be recommended for queries on SQL views. (check-in: 0884ff1d user: dan tags: schemalint)
14:56
Fix the ".lint fkey" shell command for cases where the child key is also an INTEGER PRIMARY KEY. (check-in: 48826b22 user: dan tags: trunk)
12:06
Fix the ".lint fkey" shell tool command so that it works for foreign keys that refer implicitly to primary key columns with non-BINARY default collation sequences. (check-in: 327eff25 user: dan tags: trunk)
2017-04-05
13:44
Remove a confusing and pointless sentence from the documentation for the sqlite3_interrupt() interface. (check-in: c5f1a2b6 user: drh tags: trunk)
12:39
Split off sqlite3DbFreeNN() from sqlite3DbFree() and use it in cases where we know that the argument to be freed is never NULL. (check-in: ad90e8bb user: drh tags: trunk)
11:57
Save a few bytes and a few CPU cycles in sqlite3ExprListDelete() routine. (check-in: 9e6c9391 user: drh tags: trunk)
11:49
Remove a conditional made unreachable by the previous ExprList enhancement. (check-in: a1cf4476 user: drh tags: trunk)
11:32
Combine the ExprList_item objects and the ExprList wrapper into a single memory allocation, for improved performance and reduced footprint. (check-in: 2b6560ad user: drh tags: trunk)
10:54
Attempt to remove bash-isms from configure.ac. Use -O0 with --enable-debug. (check-in: 71ed35cc user: drh tags: trunk)
2017-04-04
19:58
Fix a problem causing "PRAGMA integrity_check" to disable the xfer optimization for subsequent VACUUM operations on tables with one or more CHECK constraints. This could result in VACUUM producing slightly larger database files. (check-in: e5bb7db5 user: dan tags: trunk)
17:50
Add some support for OR terms to sqlite3_whereinfo_hook(). (check-in: 5cd07000 user: dan tags: schemalint)
04:23
Add the sqlite3_whereinfo_hook() API - an experimental API replacing the DBCONFIG_WHEREINFO hack on this branch. (check-in: a54aef35 user: dan tags: schemalint)
2017-04-03
14:07
Avoid an unnecessary call to sqlite3WhereGetMask() inside of whereShortCut(). (check-in: 5c11f430 user: drh tags: trunk)
13:59
Do not attempt to run sync2.test with SQLITE_DISABLE_DIRSYNC builds. (check-in: 658f08ce user: dan tags: trunk)
13:33
Force a schema load prior to "PRAGMA optimize". (check-in: 86897c24 user: drh tags: trunk)
13:17
Fix typos in the documentation for OP_Column. (check-in: 777b43e6 user: drh tags: trunk)
12:04
Slightly smaller and faster implementation for vdbeSorterCompareInt(). (check-in: 84fa069c user: drh tags: trunk)
2017-04-01
20:44
Remove an unnecessary setting of the Mem.enc field for the output of the OP_Record opcode, for a performance improvement and size reduction. (check-in: e6e36b28 user: drh tags: trunk)
20:14
Minor performance enhancements to the OP_Affinity opcode. (check-in: c45cd3b9 user: drh tags: trunk)
19:45
Remove an unnecessary clearing of the Vdbe.iCurrentTime value. (check-in: fcd2acdd user: drh tags: trunk)
11:59
Faster implementation for sqlite3VdbeIntValue() and sqlite3VdbeRealValue(). (check-in: 8698df60 user: drh tags: trunk)
11:40
Remove unused fields from the BtCursor object. (check-in: 1c0d82e0 user: drh tags: trunk)
00:20
Save a 78 bytes of code space and a million CPU cycles in speedtest1 by storing the cell index for the leaf page in the BtCursor object in its own field (BtCursor.ix), rather than as an entry in the BtCursor.aiIdx array. (check-in: 2452f061 user: drh tags: trunk)
2017-03-31
11:20
Merge latest trunk into this branch. (check-in: cb721d0b user: dan tags: schemalint)
2017-03-30
20:35
Merge changes from the 3.18.0 release. (check-in: 785c37d9 user: drh tags: begin-concurrent)
20:26
Merge all changes from the 3.18.0 release. (check-in: ed28f15e user: drh tags: apple-osx)
17:13
Declare the Lemon-generated parser object as itself. (Duh) (check-in: c8000e94 user: drh tags: trunk)
16:51
Increase the version number for the next development cycle. (check-in: 9e550ccc user: drh tags: trunk)
16:37
Ensure that the stack space used to hold the Lemon-generated parser object is always 8-byte aligned. (check-in: 1279de0b user: drh tags: trunk)
2017-03-29
23:22
When compiling with MSVC, disable extra SELECT/WHERE tracing for dynamically built shells. (check-in: a6891942 user: mistachkin tags: trunk)
18:03
Avoid unnecessary blob handle invalidation when changes are made to an unrelated table. (check-in: 4a01880b user: drh tags: trunk)
17:06
Slightly smaller and faster implementation of pcache1InitBulk(). (check-in: 06c2b685 user: drh tags: trunk)
16:55
Fix some tests in malloc5.test to account for the sqlite3_release_memory() change in the previous commit. (check-in: d336858d user: dan tags: trunk)
15:18
Fix the sqlite3_release_memory() interface so that it works even if SQLITE_DEFAULT_PCACHE_INITSZ is non-zero. (check-in: 5d902b7f user: drh tags: trunk)
2017-03-28
18:48
Version 3.18.0 (check-in: 424a0d38 user: drh tags: release, branch-3.18, version-3.18.0)
2017-03-25
19:16
Add the new sqlite3_set_last_insert_rowid() interface to the extension loader thunk. (check-in: 8469fc0d user: drh tags: trunk)
18:31
Do not run tests for ".dump --preserve-rowids" when testing SQLITE_OMIT_VIRTUALTABLE builds. (check-in: 55df410b user: dan tags: trunk)
18:15
Disable the new --preserve-rowids option on the CLI ".dump" command if compiled with SQLITE_OMIT_VIRTUALTABLE. (check-in: bd5bbe4e user: drh tags: trunk)
18:03
Make the default value for SQLITE_DEFAULT_SYNCHRONOUS a simple integer literal, so that it does not show up goofy in the output of "PRAGMA compile_options;". (check-in: 833ab321 user: drh tags: trunk)
12:08
Fix two harmless compiler warnings. (check-in: 32be9c3f user: drh tags: trunk)
2017-03-24
19:45
Fix a harmless compiler warning in the JSON1 extension. (check-in: c2c3dd84 user: drh tags: trunk)
18:38
Previous check-in was not correct. This is a better fix for the OP_Once problem of ticket [06796225f59c057cd120f1]. (check-in: 8194dd28 user: drh tags: trunk)
17:59
Fix the OP_Once opcode so that it works correctly for recursive triggers. Ticket [06796225f59c057cd120f1]. (check-in: 25560145 user: drh tags: trunk)
15:09
Merge all recent enhancements from trunk. (check-in: fd5676fe user: drh tags: begin-concurrent)
13:31
Add the RFC-7396 Appendix A test cases for json_patch(). (check-in: c5441d2d user: drh tags: trunk)
12:35
Fix an error in the new json_patch() routine discovered by Ralf Junker. (check-in: 9d535041 user: drh tags: trunk)
2017-03-23
23:44
Add the json_patch() SQL function to the JSON1 extension. (check-in: 47608848 user: drh tags: trunk)
20:33
Change the name of the json_merge_patch() function to just json_patch(). (Closed-Leaf check-in: 04d41004 user: drh tags: json_mergepatch)
19:51
Improvement to the amalgamation configure.ac file contributed by Bob Friesenhahn. (check-in: bf28a55d user: drh tags: trunk)
17:22
Do not run test file autoanalyze1.test if OMIT_VIRTUALTABLE is defined. (check-in: e1d06a57 user: dan tags: trunk)
17:03
Remove an invalid assert() statement failing when VACUUMing a database that contains an index on a column explicitly declared "COLLATE BINARY". (check-in: 9f2e04d3 user: dan tags: trunk)
12:56
Avoid redundant edits in the json_merge_patch() function. (check-in: 4a8e6437 user: drh tags: json_mergepatch)
00:46
Fix harmless compiler warnings in the new json_merge_patch() logic. (check-in: 5d2cf5a2 user: drh tags: json_mergepatch)
00:13
Various fixes to the json_merge_patch() function. (check-in: f49fd255 user: drh tags: json_mergepatch)
2017-03-22
21:45
Change the name of the new function to "json_merge_patch()". (check-in: 53bf70f3 user: drh tags: json_mergepatch)
21:24
Initial implementation of the json_mergepatch(A,B) SQL function. (check-in: a2674440 user: drh tags: json_mergepatch)
12:51
Fix harmless compiler warnings in the shell. (check-in: a7868297 user: drh tags: trunk)
2017-03-21
20:17
New simplified memory initialization for MacOS. (check-in: 055b36f1 user: drh tags: trunk)
18:56
Fix an incorrect assert in the ANALYZE logic for STAT4 on WITHOUT ROWID tables. (check-in: ad741976 user: drh tags: trunk)
17:19
Add short script ext/fts3/tool/fts3cov.sh. To help measure test-coverage of fts3 source code. (check-in: ee9588e8 user: dan tags: trunk)
10:45
Do not run sync2.test as part of the "inmemory_journal" permutation. (check-in: 9f680bc7 user: dan tags: trunk)
2017-03-20
22:58
Add the --native-malloc option to fuzzcheck. Fix ossfuzz.c and fuzzcheck.c so that they both deallocate the temp_store_directory before closing. (check-in: 0dd18ec8 user: drh tags: trunk)
20:42
Fix the fuzzcheck program so that it can create new databases again. (check-in: 021e8874 user: drh tags: trunk)
19:35
Avoid passing NULL as the second argument to memcpy() in fts3. (check-in: 49b93d97 user: dan tags: trunk)
19:26
Avoid a technically undefined right-shift of a signed value in rtree.c. (check-in: a144875f user: dan tags: trunk)
18:53
Fix some problems in fts3 found by address-sanitizer. (check-in: 16a8e84f user: dan tags: trunk)
16:34
Avoid the possibility of signed integer overflow with oversized precisions in %d conversions in the printf() implementation. (check-in: ef3a7c87 user: drh tags: trunk)
16:06
Do not run sync2.test as part of the "journaltest" permutation, as it uses "PRAGMA synchronous = off". (check-in: 285005a9 user: dan tags: trunk)
15:29
Ensure that a "--" prefix is added to sqlite3_trace_v2() output for nested SQL statements. (check-in: 673a7b67 user: drh tags: trunk)
15:11
Only run sync2.test on unix, as it depends on instrumentation in os_unix.c. (check-in: ecb9321e user: dan tags: trunk)
14:44
Fix the check.test script so that it works on Windows. (check-in: 8822eb5d user: drh tags: trunk)
13:03
Documentation fix: SQLITE_SOURCE_ID is a now a SHA3-256 hash. (check-in: 2aa22509 user: drh tags: trunk)
2017-03-18
13:59
Only do the specialized MacOS single-core zone_malloc initialization if compiled with the SQLITE_MIGHT_BE_SINGLE_CORE flag. This avoids a (harmless) warning about OSAtomicCompareAndSwapPtrBarrier() being deprecated. (check-in: 4e6a03d9 user: drh tags: trunk)
2017-03-17
23:08
Fix an error in the newly revised documentation for SQLITE_LIMIT_VDBE_OP. No changes to code. (check-in: f4cf8635 user: drh tags: trunk)
22:51
Set a reasonable limit on the number of opcodes in a prepared statement for ossfuzz.c. This should prevent timeouts in OSS-Fuzz when it generates totally unreasonable queries. (check-in: f74899ed user: drh tags: trunk)
22:50
Begin enforcing the SQLITE_LIMIT_VDBE_OP. The documentation warned that this day might come. (check-in: ef591461 user: drh tags: trunk)
14:59
Add the --show-errors and --show-max-delay command-line options to the ossshell test program. (check-in: 626bdca9 user: drh tags: trunk)
14:15
Fix the Makefile.in so that it builds the ossshell test program correctly. (check-in: 36f5602e user: drh tags: trunk)
13:59
Fix a buffer overread in debugging routine sqlite3VdbeMemPrettyPrint(). Problem discovered by OSS-Fuzz. (check-in: f336fba7 user: dan tags: trunk)
03:21
Fix a problem in the enhanced PRAGMA integrity_check where it verifies CHECK constraints: Do not be confused by the reuse of the Table.pCheck field by VIEWs with named columns. Problem discovered by OSS-Fuzz. (check-in: 019dd3d5 user: drh tags: trunk)
2017-03-16
18:55
If the user has not set it explicitly, set the "PRAGMA synchronous" setting to SQLITE_DEFAULT_SYNCHRONOUS when a database connection changes from wal to rollback journal mode. (check-in: 5c604479 user: dan tags: trunk)
18:54
Fix a typo causing the build to fail if SQLITE_DEFAULT_SYNCHRONOUS==SQLITE_DEFAULT_WAL_SYNCHRONOUS. (Closed-Leaf check-in: df39adea user: dan tags: default-synchronous-fix)
18:14
If the user has not set it explicitly, set the "PRAGMA synchronous" setting to SQLITE_DEFAULT_SYNCHRONOUS when a database connection changes from wal to rollback journal mode. (check-in: 78030c0f user: dan tags: default-synchronous-fix)
14:28
Simplified OOM detection in the instr() SQL function. (check-in: 6d85eb57 user: drh tags: trunk)
13:30
Fix an uninitialized variable reference in the text generator for "PRAGMA vdbe_trace=ON" output. Problem discovered by OSS-Fuzz. (check-in: e3d48716 user: drh tags: trunk)
13:14
Fix a problem in the "showdb" utility that prevents it from correctly decoding cells with no content beyond the record header. (check-in: eb7680a1 user: drh tags: trunk)
12:11
Fix a crash that could follow an OOM condition in the instr() SQL function. Problem discovered by OSS-Fuzz. (check-in: 6e59e903 user: dan tags: trunk)
2017-03-15
20:27
Updates to README files under the ext/ hierarchy. No changes to code. (check-in: 029bc5d2 user: drh tags: trunk)
19:11
Add the "Obtaining The Code" subsection in the top-level README.md file. No changes to code. (check-in: b1b1aa8b user: drh tags: trunk)
14:20
Mention the ".selftest" command in the ".help" output of the CLI. (check-in: 37f766db user: drh tags: trunk)
14:04
Make "PRAGMA temp.synchronous=N" a no-op to ensure that the TEMP schema always has synchronous=OFF. This fixes an issue discovered by OSS-Fuzz. (check-in: bcf6bb08 user: drh tags: trunk)
13:47
Improved header comment on the shathree.c extension. No changes to code. (check-in: 84f2e3d5 user: drh tags: trunk)
2017-03-14
15:27
Updates to the MSVC makefiles. (check-in: 1e4b9e7c user: mistachkin tags: trunk)
2017-03-13
22:10
Add the dbdump.c extension that implements functionality similar to the ".dump" command of the CLI, though in a separate library. (check-in: 2b9980a2 user: drh tags: trunk)
22:02
Add dbdump.exe to the MSVC makefile. (Closed-Leaf check-in: 59241a50 user: drh tags: dbdump)
21:49
Fixes to the dump logic. All appears to be working in preliminary tests. (check-in: 007b11e3 user: drh tags: dbdump)
21:26
First draft of the complete dbdump.c library. (check-in: 84ea4fcc user: drh tags: dbdump)
19:26
Infrastructure for an extension C-library to implement sqlite3_db_dump() and a corresponding "dbdump" command-line utility - both of which do the same work as the ".dump" command of the CLI. (check-in: 74c5ace4 user: drh tags: dbdump)
18:31
Merge all recent changes from trunk. (check-in: 3d04b2cd user: drh tags: apple-osx)
18:24
In the output of the ".dump" command in the CLI, quote newline and carriage-return characters using the char() function, so that they do not get eaten by end-of-line processing logic in the OS or in other command-line utilities and/or libraries. (check-in: 68f6dc7a user: drh tags: trunk)
17:37
Fix the sqlite3TreeViewSelect() routine so that it works with a null pointer to the Select object. (check-in: 9034cf7e user: drh tags: trunk)
14:30
Fix a segfault that could occur following an OOM error in the flattenSubquery() routine. (check-in: c6dda3f7 user: dan tags: trunk)
13:45
In the OSSFuzz test module, invoke the progress handler much more frequently so that timeouts are detected punctually even if the test script is running opcodes that individually take a long time (for example, an OP_Function opcode that invokes "randomblob(1.5e6)"). (check-in: f3b6959c user: drh tags: trunk)
2017-03-12
20:28
Fix a possible NULL pointer dereference in following an OOM error in sqlite3ExprIsInteger(). Problem found by OSS-Fuzz. (check-in: 5ec655e8 user: drh tags: trunk)
19:39
Remove an obsolete assert() in the IN operator code generation. (check-in: 18bf6aca user: drh tags: trunk)
2017-03-11
13:02
Make sure the translateColumnToCopy() routine in the query planner does not try to access an array that failed to be fully allocated due to a prior OOM. This fixes an issue discovered by OSSFuzz. (check-in: 3299a261 user: drh tags: trunk)
01:56
The output of the ".dump" command in the CLI quotes newline and carriage-return characters using "char(10)" and "char(13)". (Closed-Leaf check-in: 8b2954dd user: drh tags: string-quoting-dump)
00:46
Increase the number of significant digits in floating point literals on ".dump" output from the shell. (check-in: 7359fcac user: drh tags: trunk)
2017-03-10
18:36
Remove the rbu_round_trip.tcl script. It is now part of project "test-dbs". (check-in: b5bf2957 user: dan tags: trunk)
17:03
Fix an error in the SQLITE_MAX_MEMORY implementation resulting from a bad merge. Update the OSSFuzz interface so that it times out after running the byte-code engine for 10 seconds. (check-in: f8560c60 user: drh tags: trunk)
16:22
Add the -DSQLITE_MAX_MEMORY=N compile-time option. The default is no limit. (check-in: eabd4ef4 user: drh tags: trunk)
15:55
Add the SQLITE_MAX_MEMORY compile-time option that provides a hard upper bound on the amount of memory that SQLite will use, per process. (Closed-Leaf check-in: 77dfe2ab user: drh tags: max-memory-option)
14:36
Enhance the ".stats" dot-command in the CLI to use sqlite3_status64() instead of sqlite3_status(). (check-in: 118f5c05 user: drh tags: trunk)
01:05
Improvements to ".selftest --init". Tests are number in increments of 10 starting with 100. The tests are generated inside a SAVEPOINT. Errors are reported during test generation. Tests can be appended to existing tests. Add a test case to verify the schema. (check-in: b044b152 user: drh tags: trunk)
2017-03-09
22:00
Remove the test/dbselftest.c program. In its place, add the ".selftest" command to the CLI. The new CLI version is .selftest is slightly different in that it uses SHA3 hashing instead of SHA1, so the new is subtly incompatible with the old. (check-in: f4fcd46f user: drh tags: trunk)
18:13
Fix the ".dump" command to correctly extract tail data from corrupt WITHOUT ROWID tables. (check-in: 6c627e50 user: drh tags: trunk)
16:32
Add the --recovery-mode option to the CLI ".dump" command. This involves enhancing the "PRAGMA reverse_unordered_selects" command to accept a "TOGGLE" option. Recovery mode used to be always on. Now it is opt-in. Not sure this is the correct approach. Also not sure if TOGGLE is a good feature to have on boolean PRAGMA statements. (Closed-Leaf check-in: 5dc82a15 user: drh tags: dump-recovery)
13:50
Begin moving separate boolean variables in the ShellState object of the CLI into the shellFlgs bitmask. (check-in: 50eec5d9 user: drh tags: trunk)
2017-03-08
23:29
Refactor the output logic in the CLI. Not sure we want to go this way. Just saving the work for historical reference, or in case I decide to come back to it. (Leaf check-in: aacac906 user: drh tags: shell-print-refactor)
18:37
Merge recent enhancements from trunk, including the --preserve-rowids option to .dump and the .sha3sum commands of the CLI. (check-in: 9b439173 user: drh tags: apple-osx)
18:06
Make sure the database connection is open prior to running the ".sha3sum" command. (check-in: 2ea300fb user: drh tags: trunk)
17:56
In the ".sha3sum" command, if there is a LIKE pattern, show the hashes for each table separately. Without a LIKE pattern, show a single hash over the entire database. (check-in: 30f87883 user: drh tags: trunk)
16:10
Add the ".sha3sum" command to the CLI - used to compute a cryptographic hash of the content of a database file or of individual tables with in that file. (check-in: fc663799 user: drh tags: trunk)
15:03
Improved comments explaining the sha3_query() format. Fix a performance #ifdef error in the sha3 implementation. (check-in: 54ef7abd user: drh tags: trunk)
13:50
Add the shathree.c extension for implementing SHA3() and SHA3_QUERY() SQL functions. (check-in: f7ca9193 user: drh tags: trunk)
12:25
In the CLI, avoid unnecessary identifier quoting in the ".dump" output. Also add new ".dump" test cases. (check-in: de65f907 user: drh tags: trunk)
11:44
Add the --preserve-rowids option to the ".dump" command in the CLI. (check-in: c60aee24 user: drh tags: trunk)
2017-03-07
20:03
Add test script ext/rbu/rbu_round_trip.tcl. Uses "dbselftest" to test that "rbu" and "sqldiff" work together. (check-in: 961e79da user: dan tags: trunk)
14:47
Before beginning an incremental checkpoint in RBU, sync the directory containing the target database file. This ensures that the new directory entry created by renaming the *-oal file to *-wal is synced to disk. Cherrypick of [915a9a28]. (Leaf check-in: 694fe0b2 user: dan tags: version-3.17.0-rbu-fixes)
14:46
Fix another RBU case similar to the previous. This one for systems where the sector-size is larger than the page-size. Cherrypick of [4012bb3a]. (check-in: 59a11b7f user: dan tags: version-3.17.0-rbu-fixes)
14:45
When saving the state of an RBU update in the incremental-checkpoint phase, sync the database file. Otherwise, if a power failure occurs and the RBU update resumed following system recovery, the database may become corrupt. Cherrypick of [edee6a80]. (check-in: 811a5599 user: dan tags: version-3.17.0-rbu-fixes)
14:38
Performance optimization in the tokenizer/parser loop. (check-in: 2cb71583 user: drh tags: trunk)
12:18
Minor comment typo fixes. No changes to code. (check-in: 01cd9fb1 user: drh tags: trunk)
03:40
More size and speed improvements in the expression name resolver. (check-in: e0a3d39f user: drh tags: trunk)
03:25
Small size reduction and performance increase in the name resolver routine for expressions. (check-in: 1a3554e1 user: drh tags: trunk)
2017-03-06
23:51
Make the default MASK argument for "PRAGMA optimize" be 0xfffe, to allow for future expansion of up to 14 new default-on optimizations. (check-in: 73019a8b user: drh tags: trunk)
20:44
Fix a typo in a comment. No changes to code. (check-in: ec529bf1 user: drh tags: trunk)
20:00
Merge all pending 3.18 enhancements from trunk. (check-in: 1650c3f4 user: drh tags: apple-osx)
17:33
Add an initial implementation of the "PRAGMA optimize" command. (check-in: 137aeb2b user: drh tags: trunk)
11:39
Merge updates from trunk. (Closed-Leaf check-in: 5f7fc79a user: drh tags: auto-analyze)
2017-03-03
21:51
If a reprepare is needed after binding to a variable with a number larger than 32, set only the high-order bit of the Vdbe.expmask rather than setting all bits. This could potentially result in fewer false-positive reprepares. (check-in: 45797fee user: drh tags: trunk)
21:36
Remove an redundant function call from the date/time function implementation. (check-in: 4a04c48a user: drh tags: trunk)
20:43
Fix another corner-case for the 'start of ...' modifier in the date/time functions. Related to ticket [6097cb92745327a1]. (check-in: 8831f439 user: drh tags: trunk)
20:02
Fix a case introduced by [4cd2a967] where a corrupt database could cause a buffer overwrite. (check-in: 5d0455fe user: dan tags: trunk)
16:51
Before beginning an incremental checkpoint in RBU, sync the directory containing the target database file. This ensures that the new directory entry created by renaming the *-oal file to *-wal is synced to disk. (check-in: 915a9a28 user: dan tags: trunk)
2017-03-02
23:40
Fix a bug in the 'start of ...' date/time modifiers when they follow a julian day number. Fix for ticket [6097cb92745327a1]. (check-in: 081dbcfb user: drh tags: trunk)
16:56
Fix another RBU case similar to the previous. This one for systems where the sector-size is larger than the page-size. (check-in: 4012bb3a user: dan tags: trunk)
14:51
When saving the state of an RBU update in the incremental-checkpoint phase, sync the database file. Otherwise, if a power failure occurs and the RBU update resumed following system recovery, the database may become corrupt. (check-in: edee6a80 user: dan tags: trunk)
14:17
Add an optional bitmask of allowed optimizations on the "PRAGMA optimize" command. The 0x01 bit is Debug Mode. (check-in: a35388ee user: drh tags: auto-analyze)
13:22
Merge recent trunk enhancements. (check-in: c60cdb47 user: drh tags: auto-analyze)
13:13
Make sure the Vdbe.expmask value is set correctly in sqlite3VdbeSwap(). This fixes a problem introduced by [a8fd7052]. (check-in: 29f54b89 user: drh tags: trunk)
2017-03-01
11:30
Fix a use-after-free problem in the shell tool code that could occur if an SQL statement were executed after an ".open" command with invalid options. (check-in: ac760db0 user: dan tags: trunk)
2017-02-27
17:16
Make PATH handling more robust in the batch tools for MSVC. (check-in: 23a8917e user: mistachkin tags: trunk)
17:06
Improve performance of fts5 writes by writing the "averages" record once at the end of each transaction instead of every time the table is updated. (check-in: 2b210d69 user: dan tags: trunk)
16:15
Add an sqlite3_set_last_insert_rowid() method. Use it to work around fts4 and fts5 modifying the last-insert-rowid unintuitively from within commit processing. (check-in: 952a3906 user: dan tags: trunk)
14:52
Add an sqlite3_set_last_insert_rowid() method. Use it to work around fts4 and fts5 modifying the last-insert-rowid unintuitively from within commit processing. (Closed-Leaf check-in: fe41bb56 user: dan tags: set-last-insert-rowid)
12:23
Remove references to special handling in virtual table methods from the documentation for sqlite3_last_insert_rowid(). (check-in: 660f9569 user: dan tags: trunk)
2017-02-25
20:57
Add an 'extern "C"' block to header file sqlite3userauth.h. (check-in: ffd61fb4 user: dan tags: trunk)
17:47
Optimize defragmentPage() in the case where the page contains either one or two free-blocks and a small number of fragmented bytes. (check-in: 4cd2a967 user: dan tags: trunk)
16:24
Tweak the code on this branch to detect b-tree page corruption in the same cases as the trunk. (Closed-Leaf check-in: f9863b39 user: dan tags: defragmentpage-opt)
2017-02-24
19:58
Optimize defragmentPage() in the case where the page contains either one or two free-blocks and a small number of fragmented bytes. (check-in: 202b1c02 user: dan tags: defragmentpage-opt)
2017-02-23
16:30
Save a few bytes and a few cycles by setting Vdbe.expmask to zero for statements prepared using legacy interface sqlite3_prepare(). (check-in: a8fd7052 user: dan tags: trunk)
02:15
Add two NEVER() operators in the sqlite3BtreeRowCountEst() routine. (check-in: 7a959f6d user: drh tags: auto-analyze)
00:58
Do a single OP_Expire at the very end of "PRAGMA optimize", and omit the OP_Expire on ANALYZE commands invoked by the pragma. (check-in: 188300a3 user: drh tags: auto-analyze)
2017-02-22
20:24
Change the name of the analyze_as_needed pragma to "optimize". Enhance the comment (which will become documentation, assuming these changes land on trunk) to explain that the optimize pragma is likely to be enhanced in various ways in future releases and that applications should not depend upon the current behavior. (check-in: 9fced545 user: drh tags: auto-analyze)
19:49
Merge integrity_check and other improvements from trunk. (check-in: fe073905 user: drh tags: auto-analyze)
19:41
Move a branch condition in analyze.c inside an #ifdef SQLITE_ENABLE_STAT4 block. (check-in: d6afd98d user: dan tags: trunk)
19:27
Improve the performance of ANALYZE when SQLITE_ENABLE_STAT4 is defined. (check-in: 737a8244 user: dan tags: trunk)
18:53
Improvements to PRAGMA integrity_check. Verify CHECK constraints. Verify NOT NULL constraints even on table that lack indexes. Verify CHECK and NOT NULL constraints with PRAGMA quick_check. (check-in: aa02bd3c user: drh tags: trunk)
18:04
Cleanup and simplification of the output row count limit control of PRAGMA integrity_check. (Closed-Leaf check-in: 5af7d72e user: drh tags: integrity-check-improvements)
15:11
Fix integrity_check so that it verifies NOT NULL constraints even for tables that have no indexes. Enhance quick_check so that it verifies NOT NULL and CHECK constraints. (check-in: 5264844b user: drh tags: integrity-check-improvements)
14:15
Enhance "PRAGMA integrity_check" so that it verifies CHECK constraints. (check-in: 549bae08 user: drh tags: integrity-check-improvements)
2017-02-21
21:24
In sqlite3VdbeHalt(), return as soon as possible if Vdbe.magic!=VDBE_MAGIC_RUN. This makes sqlite3_reset() slightly faster in some cases. (check-in: 80adc0cb user: dan tags: trunk)
17:52
Fix an FTS5 problem that could cause a crash when certain queries were interrupted using sqlite3_interrupt(). (check-in: e400909f user: dan tags: trunk)
15:27
Very small enhancement to dispatch speed for SQL functions. (check-in: 3c3228ed user: drh tags: trunk)
14:04
The VDBE cycle counts for the sqlite3_progress_handler() callback are now cumulative. Leftovers from the previous statement are applied to the next statement. (Leaf check-in: 7a62fc6a user: drh tags: cumulative-progress-count)
13:29
Proof of concept for a "PRAGMA vdbe_cycle_limit=N" command. When N>0, invoke sqlite3_interrupt() whenever any byte code program uses more than N virtual machine cycles. (Leaf check-in: 9626b41e user: drh tags: vdbe_cycle_limit)
2017-02-20
23:32
Correct a harmless typo in the previous check-in. (check-in: 1589db01 user: mistachkin tags: trunk)
19:13
Avoid unsigned integer overflows for SQLITE_WIN32_HEAP_INIT_SIZE when the Win32 heap subsystem is used with very large values of SQLITE_DEFAULT_CACHE_SIZE and/or SQLITE_DEFAULT_PAGE_SIZE. (check-in: 96b6a98e user: mistachkin tags: trunk)
14:30
Small grammar simplification. (check-in: 0d8a868a user: drh tags: trunk)
13:11
Merge fixes from trunk. (check-in: ff213f2e user: drh tags: auto-analyze)
2017-02-18
22:52
Updated comments. No code changes. (check-in: e842ad39 user: drh tags: auto-analyze)
20:05
Add an optimization to OP_Column to speed up sequential OP_Column instructions that read earlier fields from the same cursor. Attempt to reorder OP_Column opcodes so as to take advantage of this. (Leaf check-in: c1adf959 user: dan tags: sort-column-opcodes)
15:58
Add the OP_SqlExec opcode and use it to implement "PRAGMA analyze_as_needed", invoking ANALYZE subcommands as necessary. This simplifies the implementation. (check-in: d386015f user: drh tags: auto-analyze)
13:47
Add the SQLITE_BUG_COMPATIBLE_20160819 compile-time option to omit the error message when an unrecognized argument is provided to the VACUUM command. (check-in: 49181427 user: drh tags: trunk)
02:42
Fix errors in the table resize detection. (check-in: 4229caec user: drh tags: auto-analyze)
02:19
In the analyze_as_needed pragma, avoid running unnecessary OP_LoadAnalysis and OP_Expire opcodes. Make the analyze_as_needed pragma responsive to the schema name. (check-in: 882599a4 user: drh tags: auto-analyze)
2017-02-17
23:52
Fix the #endif location for an #ifndef SQLITE_UNTESTABLE macro in the command-line shell. (check-in: 8cc9d74c user: drh tags: trunk)
21:23
Increase the estimated cost of sorting when sorting wide results sets, to account for the extra storage space and I/O required for the external sort. (Leaf check-in: 38298ef9 user: drh tags: apple-increased-sorting-cost)
19:24
The analyze_as_needed pragma now responds to table size growth and will automatically rerun the analysis after each 10x size increase. (check-in: bfbdd074 user: drh tags: auto-analyze)
16:26
Add the "PRAGMA analyze_as_needed" command. (check-in: e93db237 user: drh tags: auto-analyze)
15:26
Set the TF_StatsUsed flag on tables when the query planner outcome is affected by the sqlite_stat1 data. Also, change the column names of the "PRAGMA stats" command so that they are not keywords. (check-in: fb2b8ae8 user: drh tags: auto-analyze)
13:38
Enhance the Index and Table objects so that they remember if their stats come from the sqlite_stat1 table. Make the "PRAGMA stats" an SQLITE_DEBUG only pragma. Add the flags column to "PRAGMA stats". These are all preliminary steps toward a "PRAGMA analyze_ifneeded;" feature. (check-in: 85026c8e user: drh tags: auto-analyze)
02:07
Merge fixes from trunk. (Leaf check-in: 662e8ccf user: drh tags: increased-sorting-cost)
02:04
Fix a test case that was made to fail by the LIKE optimization enhancement in check-in [158290c0ab] but which went unnoticed because test builds were running with ICU enabled and ICU disables the LIKE optimization. (check-in: 218b2bbb user: drh tags: trunk)
01:43
Enable the ".wheretrace" and ".selecttrace" extensions in the command-line shell when compiled on Windows using DEBUG=3 or higher. Fix a harmless warning in the shell that comes up when compiled this way. (check-in: 8a03be1d user: drh tags: trunk)
2017-02-16
21:29
Increase the estimated cost of sorting when sorting wide results sets, to account for the extra storage space and I/O required for the external sort. (check-in: aa0703e5 user: drh tags: increased-sorting-cost)
20:52
Change the name of WhereInfo.pDistinctSet to pResultSet, since it is now used for more than just DISTINCT processing. (check-in: 9fc5cd50 user: drh tags: trunk)
17:18
Change two MallocZero() calls into MallocRaw() to avoid unnecessary memset(). (check-in: ff5e733c user: drh tags: trunk)
16:26
Fix a comment on a field of the ExprList object. No changes to code. (check-in: bb8e2642 user: drh tags: trunk)
15:57
Increase Table.tabFlags from 8 to 32 bits. (check-in: 7e14044c user: drh tags: trunk)
15:06
Remove two redundant initializations from the virtual table logic. (check-in: 6bd82b95 user: drh tags: trunk)
14:48
Always use the IsVirtual() macro to determine if a Table object is a virtual table. Slightly smaller and faster code. (check-in: 6affb1c8 user: drh tags: trunk)
14:02
Merge recent enhancements from trunk. (check-in: 325ccfa9 user: drh tags: est_count_pragma)
2017-02-15
22:36
Query planner optimization to detect empty tables in a join early and bail out without doing excess work. (check-in: 58797e9b user: drh tags: trunk)
18:30
Minor enhancement to mutex tracing on Win32. (check-in: 830b9235 user: mistachkin tags: trunk)
17:47
simple test cases for carray_asc. (Leaf check-in: 1d4759c1 user: drh tags: carray_asc)
16:11
Add the companion "carray_asc" table-valued function to the carray extension. (check-in: a2b4f60b user: drh tags: carray_asc)
16:04
Bummer. Checked this in as a branch off of the wrong branch.... Was: Add the companion "carray_asc" table-valued function to the carray extension. (Closed-Leaf check-in: 396b9d99 user: drh tags: mistake)
15:11
Remove the CLANG_VERSION macro, since we have learned that version numbers in clang are "marketing" and are inconsistent and unreliable. Builds using clang will still use the GCC_VERSION macro since clang works hard to be gcc compatible. (check-in: 8d3f485d user: drh tags: branch-3.17)
15:09
Remove the CLANG_VERSION macro, since we have learned that version numbers in clang are "marketing" and are inconsistent and unreliable. Builds using clang will still use the GCC_VERSION macro since clang works hard to be gcc compatible. (check-in: 810d2932 user: drh tags: trunk)
04:16
Further reforms to Tcl_*Alloc() usage. (check-in: ee1e6896 user: mistachkin tags: trunk)
01:39
In the blob test code, avoid crashing on low-memory systems by using Tcl_AttemptAlloc(). (check-in: 1d267757 user: mistachkin tags: trunk)
01:02
Make SQLITE_GET_MUTEX_TIME a proper noop on non-Win32/non-POSIX. (Leaf check-in: 4ca2b2f8 user: mistachkin tags: mutexDbg2)
00:52
Another compilation fix for POSIX. (check-in: 82ef47ea user: mistachkin tags: mutexDbg2)
00:45
Make fine time granularity work on POSIX. (check-in: 79057716 user: mistachkin tags: mutexDbg2)
00:02
Compilation fix for POSIX. (check-in: 9964ad24 user: mistachkin tags: mutexDbg2)
2017-02-14
23:58
Experimental enhancements to mutex debugging. (check-in: 80481754 user: mistachkin tags: mutexDbg2)
21:47
Clarification of the help text for the command-line shell. (check-in: ca4f1e49 user: drh tags: trunk)
20:00
Enable the SQLITE_ENABLE_NULL_TRIM option for WITHOUT ROWID tables. (check-in: 54836270 user: drh tags: trunk)
16:30
More realistic lengths of string values in speedtest1 with --testset orm. (check-in: e4731fd6 user: drh tags: trunk)
15:58
Fix a testcase number on the ORM testset of speedtest1. (check-in: 58b2f911 user: drh tags: trunk)
15:57
Add the new "--testset orm" to the speedtest1 utility. (check-in: 1836adc1 user: drh tags: trunk)
2017-02-13
16:17
Merge all changes up through the 3.17.0 release. (check-in: 1913a758 user: drh tags: apple-osx)
16:02
Version 3.17.0 (check-in: ada05cfa user: drh tags: release, version-3.17.0, branch-3.17)
13:35
Change all legacy instances of "#if SQLITE_DEBUG" to "#ifdef SQLITE_DEBUG" for consistency. (check-in: 670f10b2 user: drh tags: trunk)
13:26
Fix typos in using the MSVC_VERSION macro. (check-in: f3b65926 user: drh tags: trunk)
13:20
Avoid a duplication #define in FTS5 (check-in: c447441c user: drh tags: trunk)
11:35
Fix typos in using the MSVC_VERSION macro. (check-in: 25ebadd0 user: drh tags: branch-3.17)
2017-02-11
15:02
Ensure that indexed expressions with collating sequences are handled correctly. Fix for ticket [eb703ba7b50c1a5] backported from trunk. (check-in: b2e49ae3 user: drh tags: branch-3.17)
14:59
Fix indexes on expressions so that they can be actually used with a COLLATE clause. (check-in: e464b919 user: drh tags: trunk)
13:51
Ensure that indexed expressions with collating sequences are handled correctly. Proposed fix for ticket [eb703ba7b50c1a5]. (check-in: 9689d04b user: drh tags: trunk)
2017-02-10
21:40
Bump the version number up to 3.18.0. (check-in: 7520c238 user: drh tags: trunk)
21:37
Enhance the LIKE optimization so that it works for arbitrary expressions on the LHS as long as the pattern on the RHS does not begin with a digit or a minus sign. (check-in: 158290c0 user: drh tags: trunk)
19:38
Add the "," flag to printf(). (check-in: 064445b1 user: drh tags: trunk)
17:38
Version 3.17.0 release candidate (check-in: ad867e87 user: drh tags: branch-3.17)
2017-02-09
17:12
Cleanup the usage of the SQLITE_DISABLE_INTRINSIC compile-time option. Remove the SQLITE_RUNTIME_BYTEORDER compile-time option. Use -DSQLITE_BYTEORDER=0 instead. Fix a bug in R-Tree that occurs when compiling on a known little-endian machine without the use of intrinsic byteswapping functions. (check-in: 798fb9d7 user: drh tags: trunk)
2017-02-08
19:12
Avoid preparing a SELECT statement each time an UPDATE or DELETE by docid is executed against an fts3 table. (check-in: 1afec575 user: dan tags: trunk)
19:10
Avoid preparing a SELECT statement each time an UPDATE or DELETE by docid is executed against an fts3 table. (Leaf check-in: 9962c10a user: dan tags: fts3-seekstmt-cache)
18:13
Typo fixes in comment. No changes to code. (check-in: c09dd5c0 user: mistachkin tags: trunk)
16:55
Merge the 3.17.0beta changes from trunk. (check-in: 92dbd875 user: drh tags: apple-osx)
16:01
Always invoke the xRoundup() method of the memory allocator before calling xMalloc(). (check-in: 77b470b0 user: drh tags: trunk)
12:18
Round up the size of all memory allocations to a multiple of 8 bytes when using the system memory allocator. (check-in: c46e06fa user: drh tags: trunk)
2017-02-07
21:44
When generating the hash on the sqlite_master table in dbselftest, use an ORDER BY clause, since the sqlite_master table is reordered by VACUUM. (check-in: c8bfd99b user: drh tags: trunk)
21:09
MSVC makefile enhancments for several command line tools and the session extensions. (check-in: fc1dfe87 user: mistachkin tags: trunk)
21:00
Fix harmless compiler warnings in dbfuzz. (check-in: 61242267 user: drh tags: trunk)
20:57
Fix harmless compiler warnings in kvtest.c (check-in: db6b3993 user: drh tags: trunk)
20:51
The dbselftest utility now generates hashes in the selftest table with --init. It also accepts multiple database files on the command-line. (check-in: e68829c9 user: drh tags: trunk)
19:36
Omit fts5fault1.test from the inmemory_journal permutation. (check-in: cb1e83f9 user: dan tags: trunk)
19:23
Update a test parameter in malloc5.test to account for the increase in default lookaside buffer size. (check-in: be82d5ae user: dan tags: trunk)
17:36
Add the initial version of the "dbselftest" utility program. (check-in: 1fcac836 user: drh tags: trunk)
16:15
Update an error message in corruptC.test. (check-in: 722e57fe user: dan tags: trunk)
14:45
Fix a build problem affecting non-amalgamation rtree builds. (check-in: bb7f445b user: dan tags: trunk)
14:22
Fix a test case to account for the fact that sqlite3_blob_reopen() now returns SQLITE_IOERR_NOMEM instead of SQLITE_NOMEM if an OOM occurs in the VFS layer. (check-in: e11cc523 user: dan tags: trunk)
13:51
Fix SQLITE_BYTEORDER #defines in R-Tree when compiled separately from the amalgamation. (check-in: a136609c user: drh tags: trunk)
12:58
Make the cellMargin() routine of R-Tree slightly smaller and faster while also fixing a harmless compiler warning. (check-in: 07fe6228 user: drh tags: trunk)
03:44
Back out [ec8ff892ac] which is causing missed OOM errors on some tests. (check-in: 0611770d user: drh tags: trunk)
00:55
Fix a minor typo in a comment in R-Tree. No changes to code. (check-in: f77ee9e9 user: drh tags: trunk)
2017-02-06
16:52
Remove a small amount of unnecessary code from R-Tree. (check-in: e5aea894 user: drh tags: trunk)
15:27
Add the "max-limits" utility program to the tools/ subdirectory. (check-in: c54173b6 user: drh tags: trunk)
07:37
Fix a bug in kvtest causing "init --variance 0" runs to generate single byte blob values only. (check-in: 6b0276f9 user: dan tags: trunk)
01:19
Remove incorrect testcase() macros from R-Tree. (check-in: 853a58a7 user: drh tags: trunk)
2017-02-04
20:15
Simplification to the error handling to extension loading in sqlite3_open(). (check-in: ec8ff892 user: drh tags: trunk)
17:33
Allow sqlite3session_apply() to apply changesets to tables that have been extended using ALTER TABLE ADD COLUMN. (check-in: b20ff81f user: dan tags: trunk)
15:29
Merge recent trunk enhancements. (check-in: 6c3f0902 user: drh tags: apple-osx)
14:30
Merge the latest trunk changes, especially the RTREE enhancement to use sqlite3_blob objects. (Leaf check-in: 61bd2a88 user: drh tags: sqlite3_blob_reset)
14:24
In RTREE, use an sqlite3_blob object rather than an sqlite3_stmt object for reading content out of the %_node shadow table. (check-in: 97ccf3e4 user: drh tags: trunk)
13:12
Close sqlite3_blob objects on xSync rather than waiting until xCommit. (Closed-Leaf check-in: 95ee745f user: drh tags: rtree-blob-agressive-release)
2017-02-03
20:54
Improved performance and stack usage when processing VALUES clauses with a very large number of rows. (check-in: 5706d470 user: drh tags: trunk)
19:16
Avoid a performance problem when very large "VALUES(..), (..), (..)" terms are used in queries. (Closed-Leaf check-in: f5306ad6 user: dan tags: recursive-selectdup)
15:16
Improved tests for the carray() table-valued function. (check-in: 83a099f1 user: drh tags: trunk)
14:44
Modify the sqlite3SelectDup() routine to avoid recursing on Select.pPrior. (check-in: a7674ead user: dan tags: recursive-selectdup)
00:07
Merge in the sqlite3_blob_reset() interface. (Closed-Leaf check-in: e9c94a1f user: drh tags: rtree-sqlite3_blob)
2017-02-02
23:57
Add the sqlite3_blob_reset() interface. Enhance the behavior of sqlite3_blob objects so that they can go active again after encountering an error by rerunning sqlite3_blob_reopen(). More work needed on the documentation. (check-in: 53b77838 user: drh tags: sqlite3_blob_reset)
20:32
Ensure that all cursors have their positions saved prior to rolling back a savepoint. (check-in: 8e03a8e9 user: drh tags: trunk)
19:24
Fix issues in the sha1 extension seen with MSVC. (check-in: 5a0da77c user: mistachkin tags: trunk)
16:30
Remove the unused pReadNode prepared statement from each RTREE object. (check-in: e51dc0ec user: drh tags: rtree-sqlite3_blob)
16:08
Fix a potential uninitialized (though harmless) variable in RTREE. (check-in: a1c74e09 user: drh tags: rtree-sqlite3_blob)
15:35
The sqlite3_blob_close() interface can cause recursive invocations of nodeBlobReset() in RTREE. Make sure that does not cause problems. (check-in: 88333441 user: drh tags: rtree-sqlite3_blob)
14:40
Change RTREE so that the sqlite3_blob object is closed whenever the cursor count drops to zero and there is not a pending write transaction. (check-in: 9bb4eafe user: drh tags: rtree-sqlite3_blob)
02:28
Use the sqlite3_blob interface for reading values from the %_node shadow table in RTREE. This is a work in progress. There are still some minor problems. (check-in: fc4917d7 user: drh tags: rtree-sqlite3_blob)
00:46
This is an experimental patch that ensures that all cursors have their position saved prior to starting a ROLLBACK TO. (Closed-Leaf check-in: 01d97e5b user: drh tags: savepoint-rollback)
2017-02-01
23:06
Fix harmless compiler warnings seen with MSVC. (check-in: 0c66cf0f user: mistachkin tags: trunk)
23:03
Backout the change in [02f6293f27] as it causes MSVC to complain. (check-in: aaae74d0 user: mistachkin tags: trunk)
22:59
Fix C99-style variable declaration issue seen with older versions of MSVC. (check-in: 54d28546 user: mistachkin tags: trunk)
22:43
Fix harmless compiler warnings seen with MSVC. (Closed-Leaf check-in: 997f765b user: mistachkin tags: msvcWarn)
22:32
Add an option to the MSVC makefile to enable treating warnings as errors. (check-in: 6a378c29 user: mistachkin tags: trunk)
17:08
Unwind the RTREE dimension loop inside of rtreeCallbackConstraint(). (check-in: 4854ea9c user: drh tags: trunk)
16:41
Completely unroll the dimension loop inside of cellArea() in RTREE. (check-in: 3c4c0126 user: drh tags: trunk)
15:49
Precompute the nDim2 value in the Rtree object and use that to make loops over coordinates faster. (check-in: f1f3c8cc user: drh tags: trunk)
15:24
Use compiler intrinsic functions (when available) for byteswapping in RTREE. (check-in: 82fcd54a user: drh tags: trunk)
15:19
Fix the build by making the OPFLAG_ISNOOP macro available unconditionally. (check-in: 510933cb user: drh tags: trunk)
14:19
Update the documentation comment in sqlite.h.in for sqlite3_preupdate_hook(). (check-in: 7f857020 user: dan tags: trunk)
14:10
Changes so that the pre-update hook and the sessions module work with WITHOUT ROWID tables. (check-in: 964bdc27 user: dan tags: trunk)
02:25
More RTREE performance optimizations related to decoding values. (check-in: c5395e74 user: drh tags: trunk)
01:34
Improved comments on the statGet() implementation in ANALYZE. No changes to code. (check-in: 9663eea2 user: drh tags: trunk)
2017-01-31
21:22
R-TREE optimization: unwrap the coordinate decode loop in rtreeCallbackConstraint(). (check-in: 0bf7b518 user: drh tags: trunk)
19:10
Very small performance improvements and size reductions in sqlite3VdbeExec() and blobSeekToRow(). (check-in: 85dddf2b user: drh tags: trunk)
19:02
Simplifications to blobSeekToRow(). (Closed-Leaf check-in: 495ea824 user: drh tags: micro-optimizations)
17:31
Add a speed-test program for the sessions module. (Closed-Leaf check-in: 25f1275f user: dan tags: preupdate-without-rowid)
16:49
Remove a C99-style comment. Fixes to the kvtest-speed.sh script. (check-in: 91eb6b62 user: drh tags: micro-optimizations)
16:43
Remove another unnecessary local variable initialization from sqlite3VdbeExec() (check-in: 2361b03b user: drh tags: micro-optimizations)
16:34
Remove an unnecessary initialization of the pOp variable in sqlite3VdbeExec(). (check-in: 02f6293f user: drh tags: micro-optimizations)
15:29
Add the "stat" command to kvtest.c. Also add the --variance option to the "init" command. Add the tool/kvtest-speed.sh script used for doing performance testing on key/value access patterns. (check-in: b63deed6 user: drh tags: trunk)
15:27
Fix a typo in a comment. (check-in: bd22bf9c user: drh tags: trunk)
14:08
Merge latest trunk with this branch. (check-in: 4a592abb user: dan tags: preupdate-without-rowid)
12:41
Further minor enhancement and size reduction in sqlite3ExprAssignVarNumber(). (check-in: eacfdcf2 user: drh tags: trunk)
03:52
Performance optimization in sqlite3ExprAssignVarNumber(). (check-in: 5987ca1f user: drh tags: trunk)
2017-01-30
19:44
Experimental change to invoke the preupdate hook when WITHOUT ROWID tables are written. (check-in: 856f8604 user: dan tags: preupdate-without-rowid)
11:38
Fix building with SQLITE_OMIT_FOREIGN_KEY defined. (check-in: e93d2c49 user: dan tags: trunk)
2017-01-28
20:46
In the amalgamation, allocate the parser engine object from stack rather than from heap, for improved performance. This only happens in the amalgamation, since otherwise the sqlite3RunParser() routine does not know the object size. (check-in: 4fe879d4 user: drh tags: trunk)
19:53
Fix a couple comment typos. No changes to code. (check-in: 2a2e7d86 user: mistachkin tags: trunk)
19:45
Avoid redundant table b-tree cursor seeks in UPDATE statements that use the two-pass strategy. (check-in: dc555b10 user: dan tags: trunk)
15:26
Updates to the sqlite3_blob documentation. No changes to code. (check-in: 426b440a user: drh tags: trunk)
13:40
In the speed-check.sh script, automatically invoke "fossil test-diff --tk" on the cachegrind output against trunk, if not generating a trunk run. (check-in: aa1ab371 user: drh tags: trunk)
06:50
Fix a typo in the docs for sqlite3_update_hook(). (check-in: 7c029655 user: dan tags: trunk)
2017-01-27
19:27
Add the sha1.c loadable extension that implements the sha1() and sha1_query() SQL functions. (check-in: 24e77c1c user: drh tags: trunk)
17:02
Fix a problem causing the pre-update hook to be invoked by DROP TABLE statements. (check-in: fbb6bf1b user: dan tags: trunk)
16:39
Merge all recent trunk enhancements into the apple-osx branch. (check-in: 0e14fe1b user: drh tags: apple-osx)
16:24
Merge all recent enhancements from trunk. (check-in: fe86ff54 user: drh tags: est_count_pragma)
13:14
Alternative ICU fix (compare to check-in [50e60cb4]) that avoids casting integers to pointers. (check-in: d9752c8f user: drh tags: trunk)
01:52
In the command-line shell, enhance the ".mode" command so that it restores the default column and row separators for modes "line", "list", "column", and "tcl". (check-in: 58f02e6e user: drh tags: trunk)
01:25
Another micro-optimization in accessPayload(). Slightly smaller and faster. (check-in: c012619b user: drh tags: trunk)
01:13
Performance optimization in accessPayload(). (check-in: ebb1fd98 user: drh tags: trunk)
00:31
Simplify the accessPayload() routine so that it always populates the overflow page cache. In the one case where populating the page cache can lead to problems, simply invalidate the cache as soon as accessPayload() returns. This simplification reduces code size and helps accessPayload() to run a little faster. This backs out the eOp==2 mode of accessPayload() added by check-in [da59198505]. (check-in: 68e7a8c6 user: drh tags: trunk)
2017-01-26
21:30
Remove an unreachable branch in the error handling logic for sqlite3BtreePayloadChecked(). (check-in: 293bf3ed user: drh tags: trunk)
18:00
Improvement to the OP_Permutation opcode to prevent it from using CPU cycles for initialization in prepared statements that do not use that opcode. (check-in: b4a98f65 user: drh tags: trunk)
17:26
Remove the obsolete lastRowid cache from the sqlite3VdbeExec() for a size reduction and performance improvement. (check-in: b4803184 user: drh tags: trunk)
17:11
Performance optimization in sqlite3VdbeCloseStatement(). (check-in: 1e96e5ec user: drh tags: trunk)
16:54
Performance optimization to sqlite3_blob_read(). (check-in: 7459f4b7 user: drh tags: trunk)
16:27
Minor simplification and performance optimization for Direct Overflow Read. (check-in: 3e96d6ef user: drh tags: trunk)
02:26
Make SQLITE_DIRECT_OVERFLOW_READ work for in WAL mode as long as the page being read is not in the the -wal file. (check-in: 9879be18 user: drh tags: trunk)
01:54
Enhancements to the kvtest utility program. Add the --jmode option. Improved output formatting. (check-in: 62a4851c user: drh tags: trunk)
00:58
Modify the ICU extension to use a static initializer, as VC++ complains about a dynamic initialization. Maybe the dynamic structure initialization is a GCC extension. This fixes a problem identified on the SQLite mailing list by Ziemowit Laski. (check-in: 50e60cb4 user: drh tags: trunk)
2017-01-25
20:55
Trim NULL values off the end of records when the SQLITE_ENABLE_NULL_TRIM compile-time option is used. Increase the size of the P5 operand to 16 bits. Fix a problem with short records in the sessions extension. (check-in: 4801bd59 user: drh tags: trunk)
18:53
Fix another pre-update hook issue, this time in sqlite3preupdate_old(). (Closed-Leaf check-in: c7651d21 user: dan tags: trim-nulls)
18:12
Fix a test script problem in exclusive2.test causing it to fail on this branch. (check-in: f66614dc user: dan tags: trim-nulls)
17:44
Fix a problem with the pre-update hook on this branch. (check-in: 6fc4fbfa user: dan tags: trim-nulls)
14:58
Experimental enhancement to automatically trim NULL values from the end of records, for a reduced disk footprint. This change also involves increasing the P5 operand from 8 to 16 bits. (check-in: 118ded40 user: drh tags: trim-nulls)
14:38
Ensure that sqlite3_blob_reopen() correctly handles short rows. Fix for ticket [e6e962d6b0f06f46e]. (check-in: 8cd1a445 user: drh tags: trunk)
13:54
Fix SQLITEINT_H macro usage in two extensions. (check-in: 0803390c user: drh tags: trunk)
04:41
Ensure that sqlite3_blob_reopen() correctly handles short rows. Proposed fix for ticket [e6e962d6b0f06f46e]. Further testing needed. (Closed-Leaf check-in: 57d8dad3 user: drh tags: blob_reopen-fix)
2017-01-23
21:12
Experimental changes to permit a VFS to directly override the initial page size. (Leaf check-in: 64131724 user: mistachkin tags: pagePerSector)
19:11
Document the --mmap option in the --help screen for kvtest. Enhance kvtest so that numeric arguments can have suffixes like "K" or "M". Add kvtest to the unix makefiles. (check-in: 175bda87 user: drh tags: trunk)
18:40
Add the --mmap option to the kvtest utility program. (check-in: 4948f7e6 user: drh tags: trunk)
16:56
Optimization: Try to avoid unnecessary btree searching when repositioning a cursor to the next row. (check-in: ee793d30 user: drh tags: trunk)
15:58
Add the missing SQLITE_API symbol to test_delete.c (it is not added automatically as this file is not part of the amalgamation). (check-in: 7a4f512d user: dan tags: trunk)
07:06
Fix a problem preventing FTS5 from correctly passing the FTS5_TOKENIZE_PREFIX flag to custom tokenizer implementations. (check-in: 03c60134 user: dan tags: trunk)
2017-01-22
16:34
Fix an uninitialized variable in the command-line shell. (check-in: 06b8001a user: drh tags: trunk)
02:04
Fixes to documentation comments in the public header file. (check-in: 772dcb08 user: mistachkin tags: trunk)
00:11
Fix an initialized variable in kvtest. (check-in: ed62c5a6 user: drh tags: trunk)
2017-01-21
21:47
A better implementation of the moveto-neighbor optimization that checks for nearby rows on adjacent pages. (Closed-Leaf check-in: 2c4ecb85 user: drh tags: btree-moveto-neighbor)
16:54
B-tree optimization: When seeking on a rowid table that has already been positioned, check to see if the new row happens to be the next row on the same leaf page. That is a reasonably common case, and if it is true it avoids a full binary search. (check-in: 8e5cfb20 user: drh tags: btree-moveto-neighbor)
16:27
Change sqlite3_blob_reopen() to call sqlite3VdbeExec() directly rather than going through sqlite3_step(). Performance enhancement. (check-in: 347df3c1 user: drh tags: trunk)
16:21
Fix problems in the previous commit. (Leaf check-in: 74ad80eb user: dan tags: trigger-trace)
15:58
Add extra (somewhat inefficient) trace callbacks for triggers if SQLITE_TRACE_TRIGGER is defined. (check-in: ffda1d1e user: dan tags: trigger-trace)
15:55
In the kvtest.c test utility, reuse the buffer into which blobs are read, rather than reallocating it for each row. This is a closer match to how other test programs work, and thus provides a better comparison. (check-in: 0d1ad13a user: drh tags: trunk)
15:30
A proof-of-concept for running sqlite3_blob_open() without using OP_Column when operating on a pure key/value table. This demo does not include any corrupt database checking. Uses about 3% fewer CPU cycles on a key/value performance test. (Leaf check-in: 4cda3b30 user: drh tags: kv-access-opt-demo)
14:11
Remove an unnecessary sqlite3_bind_int64() call from sqlite3_blob_open(). Also other minor refactoring of the sqlite3_blob implementation. (check-in: 9d197a53 user: drh tags: trunk)
2017-01-20
20:43
Minor performance optimization and size reduction to the accessPayload() routine in btree.c. (check-in: 264e5c10 user: drh tags: trunk)
16:47
Fix a typo in the help message for kvtest. (check-in: 8971d98f user: dan tags: trunk)
16:46
Add option "--stats" to test program kvtest. Specifying --stats causes kvtest to output information similar to the shell tool option of the same name. (check-in: 90291327 user: dan tags: trunk)
16:09
Get the "--testset rtree" option working on speedtest1. Add the --rtree, --lookaside, and --clang options to the speed-check.sh script. (check-in: 87b640c8 user: drh tags: trunk)
00:40
Minor performance optimizations to sqlite3_blob_open() and sqlite3_blob_reopen(). (check-in: 52a61967 user: drh tags: trunk)
2017-01-19
21:20
If compiled with SQLITE_INLINE_MEMCPY, all memcpy() calls are replaced with in-line code. With that change, cachegrind shows which memcpy() calls are taking the most time. This is a performance-measurement hack only and is not for production use. (check-in: 9ed38521 user: drh tags: trunk)
18:20
In the 'vtshim' extension, avoid accessing freed memory when handling errors from xCreate/xConnect. (check-in: ffd559af user: mistachkin tags: trunk)
12:07
Add a prototype of the sqlite3_kv_reset() interface. (Leaf check-in: 088c5903 user: drh tags: sqlite3_kv)
11:52
Add test cases for the instrumentation on this branch. Fix some OOM handling issues in the same. (Leaf check-in: 50ca94b9 user: dan tags: transaction-pages)
2017-01-18
23:12
For Windows 8.x and higher, require the SQLITE_WIN32_WIN8_SECTOR_SIZE compile-time option in order to enable sector size determination. (Leaf check-in: 47beb55c user: mistachkin tags: winSectorSize)
22:47
In the 'windirent' test module, use a macro for the hidden/system attribute checking. (check-in: a84a08d0 user: mistachkin tags: trunk)
22:19
Remove superfluous option to Tcl 'lsort' in the vtabH test file. (check-in: b92cc6e5 user: mistachkin tags: trunk)
22:16
Make the vtabH-3.1 test more portable and robust. (check-in: d3c91c1f user: mistachkin tags: trunk)
22:16
Fix handling of initial hidden and/or system files in the opendir() implementation for Windows. No changes to non-test code. (check-in: 26dd42b4 user: mistachkin tags: trunk)
20:14
Add temporary code to record and report on the set of b-tree pages read and written by the current transaction. This is likely still buggy. (check-in: 2a8f6c89 user: dan tags: transaction-pages)
19:54
Begin fleshing out the key/value accessor implementation. This is an incremental check-in. (check-in: 02d60e19 user: drh tags: sqlite3_kv)
19:36
Improve and update comments. (check-in: 2dc16d34 user: mistachkin tags: winSectorSize)
19:06
Minimize the requested permissions when opening the volume on Windows 7 and Vista. (check-in: 8d429a59 user: mistachkin tags: winSectorSize)
17:20
Baseline interface definition for the experimental sqlite3_kv accessor object. (check-in: a435841e user: drh tags: sqlite3_kv)
01:11
Add runtime version checking for winSectorSize. (check-in: cb9d1ab3 user: mistachkin tags: winSectorSize)
00:27
When determining sector sizes on Windows 7 and Vista, make sure the target file is on the same volume as corresponding root directory. (check-in: de699ead user: mistachkin tags: winSectorSize)
2017-01-17
10:41
Fix a problem that could cause a spurious SQLITE_NOMEM error when attempting to resume an RBU operation if the previous client failed right after completing the incremental checkpoint. Also a "cannot vacuum wal db" error that could occur when resuming an RBU vacuum if an error (OOM or IO error) occurs during the incremental checkpoint. (check-in: 681d96eb user: dan tags: trunk)
00:10
Disable intrinsic functions for Windows using Clang, due to reports of linkage errors. This causes a 0.6% performance reduction. We will want to revisit this change in the future. (check-in: 7fd560c6 user: drh tags: trunk)
2017-01-16
18:10
Back out check-in [0b3174e0b1364c] and replace it with a better fix for \ticket [91e2e8ba6ff2e2] - a fix that does not cause the problem identified by ticket [7ffd1ca1d2ad4ec]. Add new test cases for both tickets. (check-in: 9b64af7b user: drh tags: trunk)
16:43
Back out check-in [0b3174e0b1364c] and replace it with a better fix for ticket [91e2e8ba6ff2e2] - a fix that does not cause the problem identified by ticket [7ffd1ca1d2ad4ec]. (Closed-Leaf check-in: 06136652 user: drh tags: automatic-index-affinity)
16:01
Add test cases for tickets [91e2e8ba6ff2e2] and [7ffd1ca1d2ad4ec]. (check-in: 9d0dfe0b user: drh tags: automatic-index-affinity)
11:54
An example showing how to improve performance of sqlite3VdbeSerialPut() using the GCC intrinsic function __builtin_bswap64(). (Leaf check-in: e42ed9b4 user: drh tags: builtin-bswap64)
2017-01-13
22:21
Merge updates from trunk. (check-in: 8b42b8e3 user: mistachkin tags: winSectorSize)
18:24
Fix a problem preventing resumption of RBU operations after recovering from a process or system failure that occurs during the incremental-checkpoint phase. (check-in: 97914266 user: dan tags: trunk)
12:53
Fix the build for SQLITE_ENABLE_MEMORY_MANAGEMENT. (check-in: 8c85b8fd user: drh tags: trunk)
2017-01-12
23:37
Attempt to detect physical sector sizes on Windows Vista and higher. (check-in: 6e388423 user: mistachkin tags: winSectorSize)
19:10
Remove a branch that is probably unreachable, and which adds no value. (check-in: 9acc7238 user: drh tags: trunk)
16:21
Fix harmless compiler warnings in the UPDATE code generator. (check-in: 385db266 user: drh tags: trunk)
16:14
Remove an unnecessary corruption test from the btree balancer. If corruption is present, it will be found harmlessly by later tests. (check-in: bddf3956 user: drh tags: trunk)
15:19
Test case macros to show that the previous check-in is well tested. (Leaf check-in: bd8b977a user: drh tags: extra-testcase-macros)
15:11
Improved detection of cells that extend into the reserved space at the end of the page while adjusting overflow page pointers during autovacuum. (check-in: 8097712c user: drh tags: trunk)
11:52
Remove invalid test case from cursorhints2.test (check-in: 163cc1b2 user: drh tags: trunk)
11:50
Make sure Tcl_AppendResult() always has a NULL-pointer argument at the end. (check-in: c07aef6f user: drh tags: trunk)
2017-01-11
21:03
Changes to allow some multi-row UPDATE statements to avoid the two-pass approach. (check-in: 7ae6104a user: dan tags: trunk)
20:10
Fix a problem causing the pre-update hook to be passed an incorrect rowid value in some single-pass multi-row updates. (Closed-Leaf check-in: 62257eb5 user: dan tags: onepass-update)
19:03
Fix a problem with single-pass multi-row UPDATE statements that invoke REPLACE conflict handling. (check-in: 0a2b8e1b user: dan tags: onepass-update)
16:52
Attempt to detect physical sector sizes on Windows 8 and higher. (check-in: 381fd34b user: mistachkin tags: winSectorSize)
15:42
Fix a problem preventing UPDATE statements that use a range-scan on the PK index of a WITHOUT ROWID table from using a one-pass strategy. (check-in: cab86c90 user: dan tags: onepass-update)
14:15
In the STAT4 computations, ensure that the aAvgEq values do not go negative. (check-in: f58f75b5 user: drh tags: trunk)
2017-01-10
20:04
Changes to allow some multi-row UPDATE statements to avoid the two-pass approach. (check-in: 46db23cc user: dan tags: onepass-update)
17:37
Add a test case for ticket [25e335f802dd]. (check-in: e500c15a user: dan tags: trunk)
17:33
Throw an error if the ON clause of a LEFT JOIN references tables to the right of the ON clause. Fix for ticket [25e335f802dd]. (check-in: c92ecff2 user: drh tags: trunk)
16:09
Avoid unnecessary calls to the xRoundup() method of the memory allocator when the soft heap limit is not set. (check-in: 4209b89e user: drh tags: trunk)
15:08
Fix a potential assertion fault discovered by OSS-Fuzz. (check-in: 71c03b59 user: drh tags: trunk)
2017-01-09
20:57
Merge latest changes from trunk, and also move the perf-counter into the inner loop of sqlite3BtreeMovetoUnpacked(). (Leaf check-in: ff9eab95 user: drh tags: vdbe-aux-perf)
19:55
Remove a redundant assignment statement. (check-in: a5fa0965 user: drh tags: trunk)
19:27
Performance optimization and size reduction in the OP_Variable opcode. (check-in: 237aa974 user: drh tags: trunk)
18:22
Fix typo in a comment. No changes to code. (check-in: d38fd229 user: mistachkin tags: trunk)
18:20
Try to move OP_Concat operations outside the inner loop. This turns out to make things very slightly slower, at least in speedtest1.c. (Closed-Leaf check-in: 8a90f691 user: drh tags: failed-optimization)
15:44
Modify the OP_RowData opcode so that when P3!=0 it is allowed to hold an ephemeral copy of the content. This avoids unnecessary memcpy() operations in the xfer-optimization and VACUUM. (check-in: 6e106acd user: drh tags: trunk)
13:43
Add the SQLITE_UINT64_TYPE compile-time option. (check-in: a5fe03bc user: drh tags: trunk)
07:00
Merge the "changebatch" functionality into this branch. (check-in: 50fb1eb3 user: dan tags: begin-concurrent)
06:53
Disable the optimization from [8cb8516d] (omit TableLock instructions for non-sharable databases) on this branch. This branch uses the TableLock instructions to ensure that the db schema is not written from within a BEGIN CONCURRENT transaction. (check-in: c8ca3e0a user: dan tags: begin-concurrent)
06:33
Upgrade this branch to 3.16 plus the various fixes that appeared after its release. (check-in: d0e212d0 user: dan tags: begin-concurrent)
2017-01-07
14:47
Improvements to the iIdxNoSeek optimization of sqlite3GenerateRowDelete() so that it is automatically disabled for BEFORE triggers but works in all other cases. (check-in: 3178ec4c user: drh tags: trunk)
14:26
Critical fix to the previous check-in so that it works when there are BEFORE triggers that move the cursor before the OP_Delete has a chance to be applied. (check-in: db2c0960 user: drh tags: trunk)
03:26
Avoid an unnecessary btree seek while deleting an index entry due to a conflict on a REPLACE operation. (check-in: f0495c51 user: drh tags: trunk)
00:56
Avoid duplicate b-tree searches in the duplicate row detector used to implement DISTINCT. (check-in: d577dda0 user: drh tags: trunk)
00:42
This hack illustrates how to use the VDBE_PROFILE mechanism to show which bytecode operators are using resources other than time. In this case, the number of loops through the binary search code in sqlite3BtreeMovetoUnpacked() is measured, for the purpose of helping to identify unnecessary btree searches. (check-in: 746b1836 user: drh tags: vdbe-aux-perf)
2017-01-06
17:01
Merge the changes for version 3.16.2. (check-in: c59c8880 user: drh tags: apple-osx)
16:32
Version 3.16.2 (Leaf check-in: a65a6289 user: drh tags: release, branch-3.16, version-3.16.2)
13:53
Clarify the documentation on sqlite3_column_count(). (check-in: 01b4cb26 user: dan tags: branch-3.16)
13:52
Ensure that "PRAGMA case_sensitive_like" and "PRAGMA shrink_memory" set the number of output columns to 0 (as they are statements that return no data). (check-in: 4a97ba4e user: dan tags: branch-3.16)
13:49
Ensure that "PRAGMA case_sensitive_like" and "PRAGMA shrink_memory" set the number of output columns to 0 (as they are statements that return no data). (check-in: 6696cd18 user: dan tags: trunk)
11:56
Improve handling of corrupt databases in fts5. (check-in: b0726941 user: dan tags: branch-3.16)
11:55
Improve handling of corrupt data in fts5. (check-in: 609ac1c7 user: dan tags: trunk)
01:09
Clarify the documentation on sqlite3_column_count(). (check-in: 9ea0badd user: drh tags: trunk)
00:02
Remove several non-portable tests for the deprecated temp_store_directory PRAGMA. (check-in: 28d87789 user: mistachkin tags: trunk)
2017-01-05
20:01
Ensure that sqlite3_column_count() returns 0 for the "set" mode of "get/set" PRAGMA statements that do not return a value in that case (e.g. page_size, cache_size, auto_vacuum). (check-in: e745da6b user: dan tags: branch-3.16)
20:00
Ensure that sqlite3_column_count() returns 0 for the "set" mode of "get/set" PRAGMA statements that do not return a value in that case (e.g. page_size, cache_size, auto_vacuum). (check-in: 2f57939a user: dan tags: trunk)
19:32
Ensure that sqlite3_column_count() returns 0 for the "set" mode of "get/set" PRAGMA statements that do not return a value in that case (e.g. page_size, cache_size, auto_vacuum). (Closed-Leaf check-in: 5c05d8ec user: dan tags: pragma-columncount-fix)
17:27
Version 3.16.2 second release candidate (check-in: 285c5a54 user: drh tags: branch-3.16)
17:25
Fix handling the case where a sub-query in a FROM clause is itself a UNION ALL, and one side of that UNION ALL is a query on a view that includes an ORDER BY. Fix for ticket [190c2507]. (check-in: ec1e85a6 user: dan tags: branch-3.16)
17:23
Fix handling the case where a sub-query in a FROM clause is itself a UNION ALL, and one side of that UNION ALL is a query on a view that includes an ORDER BY. Fix for ticket [190c2507]. (check-in: 590ca83b user: dan tags: trunk)
15:50
Version 3.16.2 release candidate (check-in: 9d8ffe13 user: drh tags: branch-3.16)
14:05
Increase the version number to 3.16.2 (check-in: 9592cbcf user: drh tags: branch-3.16)
13:56
Fix problems in trigger and foreign key handling when doing REPLACE on a WITHOUT ROWID table that has no secondary indexes. Fix for ticket [30027b613b4]. (check-in: 8a788594 user: drh tags: branch-3.16)
13:52
Ensure that the sqlite3_value_text() interface returns a buffer that is long enough to hold the complete string plus the zero terminator even when the input is a zeroblob. Fix for a problem detected by OSS-Fuzz. (check-in: ca185808 user: drh tags: branch-3.16)
13:50
Fix problems in trigger and foreign key handling when doing REPLACE on a WITHOUT ROWID table that has no secondary indexes. Fix for ticket [30027b613b4]. (check-in: 571f166e user: drh tags: trunk)
07:58
Ensure that the sqlite3_value_text() interface returns a buffer that is long enough to hold the complete string plus the zero terminator even when the input is a zeroblob. Fix for a problem detected by OSS-Fuzz. (check-in: 2dc7eeb5 user: drh tags: trunk)
06:57
Fix some problems with foreign key processing within REPLACE ops on WITHOUT ROWID tables with no triggers or auxiliary indexes. (Closed-Leaf check-in: c1220b1a user: dan tags: replace-fix)
2017-01-04
22:02
Improved the comment on the block of code the provides the performance optimization originally added by check-in [925840cfdb]. The original check-in omitted condition 4, which was the cause of bug [30027b613b]. (check-in: c6506b82 user: drh tags: replace-fix)
20:13
Possible fix for [30027b61]. There may still be problems surrounding foreign key processing. (check-in: 71ccb1f4 user: dan tags: replace-fix)
14:53
Adjust filename creation in test_sqllog.c so that it uses leading zeros on counters, so that files appear in order when sorted lexicographically. (check-in: 8d670973 user: drh tags: trunk)
13:09
Performance enhancements in the printf implementation. (check-in: 82e4001b user: drh tags: trunk)
04:18
Factor constant functions out of the inner loop since they are more expensive than the extra OP_Copy needed to move the result into place. (check-in: 80ad317f user: drh tags: trunk)
04:10
Clean up the implementation of constant function factorization. (Closed-Leaf check-in: 2ab997e4 user: drh tags: factor-constant-funcs)
01:07
Attempt to factor out constant functions from the interior of table scans, since functions can often be expensive to compute. (check-in: 62e9270a user: drh tags: factor-constant-funcs)
00:26
Changes to the printf implementation for better performance. (Closed-Leaf check-in: acdb8f6f user: drh tags: printf-optimization)
2017-01-03
21:57
Use compiler intrinsic functions for signed integer math when overflow detection is needed. (check-in: d3ac32a6 user: drh tags: trunk)
21:50
Back out the use of __sync_fetch_and_sub() as it does not appear to work. (Closed-Leaf check-in: 4c2efd42 user: drh tags: gnu-safe-math)
21:03
Fix a typo in a comment. (check-in: ae087123 user: drh tags: gnu-safe-math)
20:20
Merge changes from the 3.16.1 release (check-in: dcbec226 user: drh tags: apple-osx)
20:01
Use the CLANG_VERSION macro to control clang-specific features. (check-in: f8ebeec2 user: drh tags: gnu-safe-math)
18:27
Version 3.16.1 (check-in: 979f0439 user: drh tags: release, version-3.16.1, branch-3.16)
18:05
Use the GCC built-in __sync_fetch_and_sub() to make the sqlite3StatusDown() routine atomic, and thereby avoid some mutexing. (check-in: f69ce75b user: drh tags: gnu-safe-math)
17:33
Make use of the __buildin_OP_overflow() functions from GCC when doing 64-bit signed integer arithmetic. (check-in: 82cbebb8 user: drh tags: gnu-safe-math)
16:03
Bump the version number up to 3.16.1. (check-in: a60b60d6 user: drh tags: branch-3.16)
15:59
Defer size checking on row-value assignments for when the RHS is a SELECT until after the "*" wildcards have been expanded. (check-in: 5c892938 user: drh tags: branch-3.16)
15:57
Fix the row-values in UPDATE statements within TRIGGER problem identified by ticket [8c9458e7]. (check-in: bf984e98 user: drh tags: branch-3.16)
14:45
Add the experimental affinity() SQL function when SQLITE_DEBUG is defined. (check-in: bed0eaa5 user: drh tags: trunk)
14:39
Put the affinity() function implementation inside of #ifdef SQLITE_DEBUG. (Closed-Leaf check-in: f778f58a user: drh tags: affinity-sql-func)
14:30
Merge all the latest changes from trunk. (check-in: c27cd8a8 user: drh tags: affinity-sql-func)
13:45
Defer size checking on row-value assignments for when the RHS is a SELECT until after the "*" wildcards have been expanded. (check-in: 36944be6 user: drh tags: trunk)
11:59
Fix a typo on a comment. No code changes. (check-in: abc27b60 user: drh tags: trunk)
08:11
Add test cases for the fix in [f12ed3ce]. No problems discovered. (check-in: 548532fd user: dan tags: trunk)
02:58
Improvements to the way vector assignment size checking is done. Size checks when the RHS is a SELECT are deferred until after "*" wildcards are expanded. (Closed-Leaf check-in: 696219b1 user: drh tags: vector-size-check)
01:24
Fix the row-values in UPDATE statements within TRIGGER problem identified by ticket [8c9458e7]. (check-in: f12ed3ce user: drh tags: trunk)
00:27
Improved assert()s on the sqlite3ExprListDup() logic for TK_SELECT_COLUMN. (Closed-Leaf check-in: 14da99d4 user: drh tags: rowvalue-update-trigger)
2017-01-02
23:43
Test cases for the row-value UPDATE TRIGGER fix. (check-in: b9b96437 user: drh tags: rowvalue-update-trigger)
23:18
Put an ALWAYS() on an unreachable branch in the sqlite3ExprListDup() routine. (check-in: 2caaaab5 user: drh tags: rowvalue-update-trigger)
22:36
Proposed fix for the row-value TRIGGER UPDATE problem described in ticket [8c9458e7]. (check-in: 61a442ea user: drh tags: rowvalue-update-trigger)
19:10
Increase the version number to 3.17.0 for the next release cycle. (check-in: 16415b5a user: drh tags: trunk)
19:02
Provide the SQLITE_DEFAULT_LOOKASIDE compile-time option and make it's default value be 1200,100 (raised from 512,125 in the latest release). (check-in: 584da48f user: drh tags: trunk)
18:40
Change SQLITE_DEFAULT_PCACHE_INITSZ from 100 to 20, which is experimentally determined to be slightly faster. (check-in: 12d9493c user: drh tags: trunk)
18:19
Avoid unnecessary calls to sqlite3BtreeEnterAll() and sqlite3BtreeLeaveAll() when no btree is using shared-cache. (check-in: cfb31582 user: drh tags: trunk)
12:20
Add the --all option to the wordcount test program. Fix the speedtest1 test program so that it builds on MSVC and so that the --lookaside 0 0 option works. (check-in: cb338f36 user: drh tags: trunk)
11:57
Version 3.16.0 (check-in: 04ac0b75 user: drh tags: trunk, release, version-3.16.0)
2017-01-01
12:44
Detect row-value comparison size mismatches even when the size of one operand is obscured by an unexpanded subquery. (check-in: 2c4d167c user: drh tags: trunk)
2016-12-31
21:55
Add the --help and --all options to the wordcount test utility. (Closed-Leaf check-in: 18baeadf user: drh tags: wordcount-enhancement)
18:37
In speedtest1.c, make the "--lookaside 0 0" option completely disable lookaside. (Closed-Leaf check-in: 52b99bcb user: drh tags: speedtest1)
14:33
Minor #include change to speedtest1.c so that it will compile under MSVC. (check-in: 8c28fde0 user: drh tags: speedtest1)
2016-12-30
17:40
Fix a crash that could occur following an OOM in the group_concat() function if the second argument is an SQLITE_BLOB value. (check-in: 14d855d2 user: dan tags: trunk)
15:16
Strengthen the defense against OOM in the instr() SQL function. (check-in: a0971e71 user: drh tags: trunk)
14:25
Add a test to ensure that the app-defined pcache xFetch() method is never passed a key of zero. (check-in: 0bdbe49c user: drh tags: trunk)
14:15
Avoid passing NULL pointers to memcmp() or memcpy(), even when the "number-of-bytes" argument is passed 0. (check-in: 56ff72ab user: dan tags: trunk)
13:55
Clarify the expectations for the behavior of the xFetch method on the sqlite3_pcache_methods object for the case when the key is zero. (Closed-Leaf check-in: b0810ac1 user: drh tags: pcache1-zero-page)
13:40
Improved detection of zero page numbers in the page cache. (check-in: 5550e815 user: drh tags: pcache1-zero-page)
12:10
Fix a harmless compiler warning in fuzzcheck.c (check-in: 2842bc60 user: drh tags: trunk)
12:06
Disable the pagerAcquireMapPage() routine if memory-mapped I/O is disabled. This fixes a harmless compiler warning on OpenBSD. (check-in: 3e25ba6e user: drh tags: trunk)
00:09
Encode a 64-bit integer literal in date.c as a constant expression so that it works on older compilers. Also fix a harmless compiler warning in vdbe.c. (check-in: f57952ba user: drh tags: trunk)
2016-12-29
19:48
Fix harmless compiler warnings in the command-line shell and in Lemon. (check-in: afcdc4a6 user: drh tags: trunk)
17:25
In kvtest.c, use stat() instead of fseek()/ftell() to determine the size of a BLOB to be read directly from disk. This makes the pile-of-files database more competative against SQLite. (check-in: a7dca29f user: drh tags: trunk)
16:58
Add the kvtest.c test program used to show that it is many times faster to read thumbnail and similar BLOBs out of an SQLite database than it is to read them as separate files from the filesystem. (check-in: 8074d59c user: drh tags: trunk)
16:49
Add more detail to the header command to further explain how to compile and use the kvtest.c utility. (Closed-Leaf check-in: 55d29839 user: drh tags: kvtest)
16:18
Fix kvtest.c so that it compiles using MSVC. (check-in: e2bbeae7 user: drh tags: kvtest)
15:26
In kvtest.c: do not show the --cache-size setting on fopen() runs. Show progress for the (potentially slow) "export" command. (check-in: b0f58d24 user: drh tags: kvtest)
14:44
Improvements to the kvtest.c utility. Added the --cache-size option. Additional reporting of version and settings at the end of "./kvtest run". (check-in: f6fcac6a user: drh tags: kvtest)
03:57
Add the kvtest.c test program for measuring key/value read performance under various scenarios. (check-in: 489e0787 user: drh tags: kvtest)
2016-12-27
15:59
Do not allow the nockpt.test module to run under the inmemory_journal permutation. (check-in: a6af06f1 user: drh tags: trunk)
13:33
Allow sqlite3_interrupt() to be invoked on a database connection that is in the process of being closed even if SQLITE_ENABLE_API_ARMOR is defined. (check-in: 7854bee2 user: drh tags: trunk)
12:45
Adjust a corruption test case to accommodate the sqlite3BtreeInsert() optimization of check-in [0b86fbca66]. (check-in: 4cb0945f user: drh tags: trunk)
12:35
Disable new test modules nockpt.test and interrupt2.test for incompatible permutations. Add a CORRUPT_DB term to an assert() in vdbe.c. (check-in: a54e619e user: drh tags: trunk)
12:08
Avoid a potential (harmless) signed integer overflow in memory usage accounting when calling sqlite3_realloc() to reduce the size of an allocation. (check-in: 2d71cbdf user: drh tags: trunk)
02:43
Minor changes to make some newer test cases work under more compile-time options. (check-in: 382eea36 user: drh tags: trunk)
2016-12-26
12:25
During fuzz testing with the fuzzcheck utility program, impose a record length limit of 100MB and a limit on the length of LIKE strings of 50 bytes. (check-in: edc9db41 user: drh tags: trunk)
12:14
Enable ossfuzz.c to build even if SQLITE_OMIT_PROGRESS_CALLBACK is defined. (check-in: a37c00dc user: drh tags: trunk)
01:41
Remove an incorrect ALWAYS() macro from balance_nonroot(), which could result in corrupt databases if it were optimized out. This ALWAYS was added only two weeks ago (check-in [f9f2e23bbd68a]) and has never appeared in an official release. (check-in: e447b23c user: drh tags: trunk)
00:18
Add the built-in affinity() SQL function. (check-in: 57e40e1c user: drh tags: affinity-sql-func)
00:15
Enhance the fuzztest utility with the --prng-seed option. Always reseed the PRNG prior to each test. (check-in: 8c5187f6 user: drh tags: trunk)
2016-12-24
21:32
Combine the implementations of the ".tables" and ".indexes" commands in the command-line shell. The ".indexes" command now puts the indexes in multiple columns, just like ".tables" and shows all indexes in all attached databases. (check-in: def29333 user: drh tags: trunk)
19:37
The schema name "main" is always an acceptable alias for the primary database even if the primary database is renamed using SQLITE_DBCONFIG_MAINDBNAME. (check-in: 2f481b85 user: drh tags: trunk)
18:18
Change the output format of the ".databases" command in the command-line shell so that it shows the schema name, a colon, and the corresponding filename. (check-in: 8558512e user: drh tags: trunk)
18:04
In the command-line shell, improve the formatting to the ".databases" command and on the ".tables" command, only show the schema prefix for databases past the first ("main") database. (check-in: c0d5626e user: drh tags: trunk)
2016-12-23
16:05
Fill in missing opcode documentation entries for OP_SorterNext and OP_SorterSort. (check-in: 2940661b user: drh tags: trunk)
13:54
Use the VList object to capture the mapping between SQL parameter names and variable numbers in a single memory allocation. (check-in: 99fddf2e user: drh tags: trunk)
13:52
Add check to prevent a VList from growing after pointers to labels have been taken. (Closed-Leaf check-in: aa23d7ea user: drh tags: VList)
13:30
Fix the VList object so that it actually works. (check-in: 9dcd8569 user: drh tags: VList)
03:59
Use the VList object to replace Parse.azVar for tracking the mapping between SQL parameter names and parameter numbers. There is a performance improvement, though there are still a few hiccups in the current code. (check-in: 68ecafa1 user: drh tags: VList)
2016-12-22
14:53
Fix harmless compiler warnings. (check-in: fa86db2f user: drh tags: trunk)
2016-12-21
21:26
Change the default lookaside configuration from 500 slots of 128 bytes each to 125 slots of 512 bytes each. This uses the same amount of memory (64,000 bytes) but seems to perform much better in test applications. (check-in: 52a12e47 user: drh tags: trunk)
19:45
Simplifications to the way UnpackedRecord objects are allocated. Smaller and faster code that also fixes a subtle (currently unreachable) bug. (check-in: f7ab01f2 user: drh tags: trunk)
2016-12-18
17:42
Change the OP_IfNotZero opcode so that it decrements register P1 by 1 rather than the value in P3, and so that it only decrements if originally positive. This avoids decrementing the smallest 64-bit signed integer. (check-in: 165c0446 user: drh tags: trunk)
2016-12-17
21:07
Refinements to the new dbfuzz test program. (check-in: 50325abe user: drh tags: trunk)
20:27
Enhance fuzzershell.c to read and execute SQL commands in the autoexec table of the database under test. Add the dbfuzz.c test program combining selected features of fuzzershell.c and fuzzcheck.c. (check-in: ef6e071a user: drh tags: trunk)
08:18
Fix a problem in the shell tools readfile() command causing blobs to be truncated at the first embedded 0x00 byte in release builds, or an assert() to fail in a debug build. (check-in: 8dedd6ad user: dan tags: trunk)
2016-12-16
21:29
Merge recent enhancements from trunk, and especially the pragma-as-vtab change. (check-in: 4b1e7804 user: drh tags: apple-osx)
21:15
Merge the pragma-as-vtab enhancement from trunk. (check-in: 89250777 user: drh tags: est_count_pragma)
21:02
Fix a potential memory leak following an OOM error in the pragma processing. (check-in: 396ab8f8 user: drh tags: est_count_pragma)
18:43
Add the ".lint fkey-indexes" command to the command-line shell. (check-in: 94689e3b user: drh tags: trunk)
18:41
Fix minor issues with the ".lint" command implemention in the shell and the shell help text related to ".lint". (Closed-Leaf check-in: 1268dc77 user: drh tags: fkey-missing-indexes)
18:14
Built-in PRAGMA statements without side-effects can be invoked as table-valued functions by prefixing their name with "pragma_". (check-in: d66ec5cf user: drh tags: trunk)
17:28
Change ".fkey_missing_indexes" to ".lint fkey-indexes". (check-in: ee621ade user: dan tags: fkey-missing-indexes)
16:49
Experimental merge of the est_count_pragma and the pragma-as-vtab branches. (Closed-Leaf check-in: 4b73ee33 user: drh tags: est-count-pragma-vtab)
16:44
Update .fkey_missing_indexes to use the built-in pragma vtabs. (check-in: 3ab05987 user: dan tags: fkey-missing-indexes)
16:13
Merge the pragma-as-vtab change into this branch. (check-in: 4ba45e72 user: dan tags: fkey-missing-indexes)
15:57
Merge recent trunk enhancements. (check-in: 74a0ca1f user: drh tags: est_count_pragma)
15:05
Fix a problem causing the planner to generate sub-optimal plans for some queries that use recursive WITH sub-queries with LIMIT clauses. (check-in: 053a149c user: dan tags: trunk)
04:20
Fix an error in the way the "schema" argument to some pragma virtual tables is handled. (Closed-Leaf check-in: 546821e2 user: drh tags: pragma-as-vtab)
02:31
Simplifications to facilitate full test coverage. (check-in: 01afc515 user: drh tags: pragma-as-vtab)
02:14
Minor bug fixes and enhancements to the pragma eponymous virtual tables. (check-in: 3f8f4615 user: drh tags: pragma-as-vtab)
01:09
Merge enhancements from trunk, and especially the ability to use temp.sqlite_master as an alias for sqlite_temp_master. (check-in: 994a4401 user: drh tags: pragma-as-vtab)
01:00
All temp.sqlite_master to be used as an alias for sqlite_temp_master. (check-in: 8d646905 user: drh tags: trunk)
2016-12-15
21:33
Fix the cost estimation in the BestIndex method of the eponymous pragma virtual table implementation. (check-in: 7126807a user: drh tags: pragma-as-vtab)
21:11
The pragma as eponymous virtual table mechanism now appears to work. (check-in: 2c274a1a user: drh tags: pragma-as-vtab)
20:59
Code to automatically create eponymous virtual tables for read-only pragmas. Compiles, but does not yet work. (check-in: 988a61e8 user: drh tags: pragma-as-vtab)
18:59
Do exponential rather than linear expansion of the SrcList.a array when appending new elements, to reduce the number of malloc() calls. (check-in: 4e374912 user: drh tags: trunk)
16:01
Do more pragma processing from tables rather than in-line code. (Closed-Leaf check-in: a88ca352 user: drh tags: table-driven-pragma)
06:01
Add the -groupbyparent option to the ".fkey_missing_indexes" command. (check-in: 976c51b4 user: dan tags: fkey-missing-indexes)
2016-12-14
19:28
Add the experimental ".fkey_missing_indexes" command to the shell tool. To identify indexes that should be created on child keys if FK processing is to be enabled. (check-in: 7df23aca user: dan tags: fkey-missing-indexes)
14:07
Refactor the Table.nRef field as Table.nTabRef for easier grepping. (check-in: 9cae4c2e user: drh tags: trunk)
13:54
Increase the size of the reference count on Table objects to 32 bits. (check-in: d08b72c3 user: drh tags: trunk)
11:39
Fix the configure makefile to align it with the manual makefile. Unsaved changes from the previous check-in. (check-in: 00a184b6 user: drh tags: trunk)
11:23
Adjust the unix makefiles so that "make test" runs fuzzcheck prior to even attempting to build the other test programs. (check-in: bb48f7a5 user: drh tags: trunk)
11:14
Three times faster sqlite3SrcListAppend() in the common case by avoiding the call to sqlite3SrcListEnlarge() for the first allocation. (check-in: 0ea2762f user: drh tags: trunk)
10:30
Fix the optimization that prevents writing freelist pages to the journal. (check-in: 6aa9b265 user: drh tags: trunk)
2016-12-13
23:22
In the command-line shell, in the output of the ".dump", ".schema", and ".fullschema" commands, convert CREATE TABLE statements that appear to come from shadow tables into CREATE TABLE IF NOT EXISTS statements. (check-in: c7021960 user: drh tags: trunk)
20:30
Fix harmless compiler warnings. (check-in: ed2c9f37 user: drh tags: trunk)
18:47
Convert sqlite3PagerGet() into a pointer-dispatched virtual method. This makes it about 25% faster. (check-in: 7f88bb44 user: drh tags: trunk)
18:34
In the pager, avoid checking for the illegal page number 0 except when creating a new page. (Closed-Leaf check-in: dee20ba9 user: drh tags: pager-get-method)
16:57
Fix a problem causing SQLite to return false "foreign key violation" errors when there is a partial (i.e. WHERE constrained) UNIQUE index on the parent key columns. This bug did not cause SQLite to allow illegal data to be inserted into the database, only to reject legal operations. (check-in: 850877d1 user: dan tags: trunk)
15:53
Further refinements to the virtual method implementation of sqlite3PagerGet(). (check-in: 67df4446 user: drh tags: pager-get-method)
14:32
Make the sqlite3PagerGet() interface into a virtual method, with different implementations based on the current state of the pager. This gives a small performance increase by avoiding unnecessary branches inside the various methods. (check-in: df5bb90d user: drh tags: pager-get-method)
2016-12-12
23:24
Add the --mmap option to the speedtest1 program and to the speed-check.sh script that is frequently used to run speedtest1. (check-in: 1a636d5e user: drh tags: trunk)
18:12
Remove unnecessary MemPage initialization resets in the autovacuum ptrmap processing. (check-in: 1b36fbb9 user: drh tags: trunk)
16:15
Merge all the latest performance enhancements from trunk. (check-in: a24f805b user: drh tags: apple-osx)
16:08
Faster operation for large in-memory databases. (check-in: 9675518b user: drh tags: trunk)
12:58
Merge recent enhancements from trunk. (check-in: dc006e08 user: drh tags: est_count_pragma)
11:05
Performance optimization in sqlite3PagerWrite(). (check-in: d4dff10a user: drh tags: trunk)
01:53
Updates to the tokenizer for EBCDIC. No changes for standard builds. (check-in: 4fc6580f user: drh tags: trunk)
01:30
Size and performance optimization in btreeInitPage() in the freeblock validation step. (check-in: 57deb1b4 user: drh tags: trunk)
00:58
Change the order of comparison for divider cell overflow in balance non-root to avoid a (harmless) reference to an uninitialized variable. (check-in: f9f2e23b user: drh tags: trunk)
2016-12-10
13:12
Avoid an unnecessary memset() in the sqlite3BtreeInsert() routine. (check-in: 35ecd4eb user: drh tags: trunk)
12:58
Reorder the fields in the VdbeCursor object so that those that need to be bulk zeroed on allocation are grouped at the beginning, and the memset() only runs over those fields that really need it. (check-in: 8165f88b user: drh tags: trunk)
04:06
Avoid signed integer overflow when dealing with a LIMIT and OFFSET whose sum exceeds the maximum integer value. (check-in: c9bdf7ad user: drh tags: trunk)
00:14
In balance_nonroot, try to combine dropCell/insertCell combinations for the dividers into a cell overwrites. This results in a very small (0.05%) performance gain which is probably not worth the added complexity. (Closed-Leaf check-in: 478627c9 user: drh tags: failed-dropCell-opt)
2016-12-09
19:42
When doing the sqlite3BtreeInsert() overwrite optimization, make sure the memcpy() does not extend off the end of the page. (check-in: 684ef458 user: drh tags: trunk)
18:09
Additional comments and an assert on the sqlite3BtreeInsert() overwrite optimization. (check-in: c1f0ae9d user: drh tags: trunk)
17:32
In sqlite3BtreeInsert() when replacing a re-existing row, try to overwrite the cell directly rather than deallocate and reallocate the cell. (check-in: 0b86fbca user: drh tags: trunk)
16:12
Fix an fts5 bug that could cause a crash following an OOM error or sqlite3_interrupt() interrupt. (check-in: 0ea3ece9 user: dan tags: trunk)
16:02
Avoid unnecessary zeroing of fields in the MemPage object that are going to be reinitialized before use anyhow. A smaller and faster binary results. (check-in: 01ada3d1 user: drh tags: trunk)
00:15
Fix an obscure problem in range estimation with STAT4. (check-in: 1f16c9a7 user: drh tags: trunk)
2016-12-08
23:52
Fix the showstat4 utility program so that is displays strings using standard SQL notation (single quotes) rather than C-style notation. (check-in: 92998e4a user: drh tags: trunk)
19:04
Fix a minor error in the cg_anno.tcl script causing the cycles-per-file counts to be miscomputed. (check-in: b26df26e user: drh tags: trunk)
18:36
Update the tool/cg_anno.tcl script to give a summary of cycle counts by canonical source file name. (check-in: 254a83bf user: drh tags: trunk)
01:38
More changes to take advantage of the sqlite3VdbeAppendP4() method. (check-in: 83bc5e40 user: drh tags: trunk)
2016-12-07
21:35
Add the sqlite3VdbeAppendP4() method for adding P4 content to the most recently coded instruction. (check-in: 28883e8f user: drh tags: trunk)
20:22
The use P4_DYNAMIC in place of the P4_MPRINTF operand type. (check-in: 3954f837 user: drh tags: trunk)
20:09
Performance optimization for sqlite3VdbeAddOp4(). (check-in: 04716c90 user: drh tags: trunk)
17:06
Put the SQLITE_UNTESTABLE compile-time option in the correct order for PRAGMA compile_options. (check-in: 24574a15 user: drh tags: trunk)
15:49
Rename the SQLITE_OMIT_BUILTIN_TEST compile-time option to SQLITE_UNTESTABLE. (check-in: f3608187 user: drh tags: trunk)
15:38
Prevent the flattening or where-term push-down optimizations from obscuring misuses of SQL row values that can lead to crashes or assert() failures. (check-in: 433d16ff user: dan tags: trunk)
13:49
Always honor the sqlite3.dbOptFlags bitmask, regardless of compile-time options. Continuing fix for ticket [da78413751863]. (check-in: afab1663 user: drh tags: trunk)
07:46
Merge latest trunk changes into this branch. (Leaf check-in: 391344d8 user: dan tags: changebatch)
2016-12-06
22:47
Performance improvement and size reduction in the Expr node allocator function sqlite3PExpr(). (check-in: 2a81763e user: drh tags: trunk)
19:33
Add missing nul-terminator to a Tcl_AppendResult() call in tclsqlite.c. (check-in: 0820f8b3 user: dan tags: trunk)
17:59
In the LEMON parser-generator, fix the stack overflow processing so that it correct invokes the destructor on the top-level of the parse stack. (check-in: e8247065 user: drh tags: trunk)
2016-12-05
20:16
Fix a harmless redundant local variable declaration in the date/time function logic. (check-in: b4bc40d1 user: drh tags: trunk)
20:06
Work around a bug in the definition of "ino_t" on some versions of Android. (check-in: 8df492c1 user: drh tags: trunk)
2016-12-02
19:07
Simplify the date/time function logic for improved rebustness and also to decrease the size of the binary. (check-in: 6e144735 user: drh tags: trunk)
17:08
Extend the valid range of 'unixepoch' dates from JulianDay 0 through 9999-12-31. (check-in: 768c9859 user: drh tags: trunk)
14:15
Fix a potential integer overflow during out-of-bound date computations. (Closed-Leaf check-in: 3c58b173 user: drh tags: date-overflow-fix)
02:19
Test cases for the improved LIKE/GLOB matcher. (check-in: 0bdb7ae1 user: drh tags: trunk)
2016-12-01
19:58
Avoid clearing the EP_FromJoin flag from terms in ON clauses when flattening sub-selects. Possible fix for [2df0107b]. (check-in: a427c405 user: dan tags: trunk)
19:38
Avoid clearing the EP_FromJoin flag from terms in ON clauses when flattening sub-selects. Possible fix for [2df0107b]. (Closed-Leaf check-in: 8bed4cd5 user: dan tags: left-join-fix)
18:57
Performance improvement for GLOB and LIKE matching for patterns with two or more multi-character wildcards ("*" or "%"). (check-in: 2df0ebf9 user: drh tags: trunk)
18:49
Faster version of patternCompare() that uses new return values rather than an extra parameter to communicate wildcard information back up to parent searches. (Closed-Leaf check-in: a1e2b6ce user: drh tags: pattern-compare-optimization)
17:34
Modify the patternCompare() function (used for GLOB, LIKE) to better handle patterns containing multiple wildcard characters ("*", "%"). (check-in: c5e5614d user: dan tags: pattern-compare-optimization)
2016-11-30
16:54
Add the remember(V,PTR) extension function which copies an SQL value into an application variable. (check-in: d2d30914 user: drh tags: trunk)
16:39
Merge all the latest changes from trunk. (check-in: 7ca58a07 user: drh tags: est_count_pragma)
14:47
Fix then handling of the (oversized) integer literal -0x8000000000000000. (check-in: 3816bb41 user: drh tags: trunk)
05:08
Change the OP_DecrJumpZero opcode back to its old behavior of always decrementing, except top the decrementing when it reaches the largest negative number. (check-in: 9d0d8c2e user: drh tags: trunk)
04:07
More improvements to boundary cases in the date/time functions, flowing out of branch coverage testing. (check-in: 1218005a user: drh tags: trunk)
01:05
Prevent a warning about integer overflow when using a very large negative LIMIT. (check-in: 96106d56 user: drh tags: trunk)
00:48
Further changes to the date/time functions to suppress harmless signed integer overflow warnings that could have occurred when doing out-of-range date calculations which, according to the docs, give undefined results. (check-in: dc453b34 user: drh tags: trunk)
2016-11-29
20:39
The documentation says that the built-in date-time functions give undefined results for dates before 0000-01-01 and after 9999-12-31. Change the actually implementation so that the answer given is really NULL. This also avoids unnecessary hand-wringing over an signed integer overflow that might otherwise occur when processing out-of-bound dates. (check-in: d410a839 user: drh tags: trunk)
2016-11-28
21:22
Fix broken hyperlinks to the sqlite3_snapshot objection definition in the documentation. No changes to code. (check-in: 9021f687 user: drh tags: trunk)
19:13
Version 3.15.2 (Leaf check-in: bbd85d23 user: drh tags: release, version-3.15.2, branch-3.15)
14:52
Simplify some code in fts5_expr.c. Add some extra test cases for the same. (check-in: 7cc2d60d user: dan tags: trunk)
2016-11-26
20:44
Merge all recent trunk changes, and especially the new sqlite3_snapshot_recover() interface. (check-in: 41a3af54 user: drh tags: apple-osx)
20:12
Update test case for opening SHM files read/write on a read-only connection so that they are only attempted on Darwin with SQLITE_ENABLE_PRESIST_WAL. (check-in: d6a7bf80 user: drh tags: apple-osx)
19:44
Add the sqlite3_snapshot_recover() interface and related functionality. (check-in: b70c85ce user: drh tags: trunk)
2016-11-25
20:20
Clarification of the behavior of sqlite3_stmt_readonly() on BEGIN IMMEDIATE statements. The is comment changes only - no changes to code. (check-in: a4205a83 user: drh tags: trunk)
19:32
Remove the OP_RowKey opcode. Use OP_RowData in its place. (check-in: 6ac7b07a user: drh tags: trunk)
19:18
Refactor the sqlite3BtreeKey() and sqlite3BtreeData() internal interfaces into sqlite3BtreePayload() and sqlite3BtreePayloadChecked(), respectively. This is a continuation of the optimization started by check-in [2d831074cf]. The result is a slightly smaller and faster binary. (check-in: 49ebc219 user: drh tags: trunk)
17:03
Small performance increase and size reduction in the OP_Column opcode. (check-in: a9498407 user: drh tags: trunk)
15:47
Performance enhancement to sqlite3_snprintf(). (check-in: c53dca7f user: drh tags: trunk)
15:11
Performance improvement in sqlite3StrAccumFinish() for the common case where no memory allocation is required. (check-in: b6acf5d4 user: drh tags: trunk)
14:30
Add a comment to help clarify the EP_FromJoin hack in exprCodeBetween(). (check-in: 6df7c574 user: drh tags: trunk)
2016-11-24
13:42
Fix a test case in vacuum5.test so that it works with the in-memory journal permutation. Cherrypick of [81c86401]. (check-in: e31d646a user: dan tags: branch-3.15)
2016-11-23
23:18
Fix the error message for invalid ".mode" arguments in the command-line shell, so that it mentions the new "quote" mode. (check-in: 9a603a18 user: drh tags: trunk)
21:33
Updates makefiles and build scripts on windows to used Tcl 8.6. (check-in: f693850a user: drh tags: branch-3.15)
21:14
Update the version number to 3.15.2. (check-in: 86dc74cb user: drh tags: branch-3.15)
21:13
Code constant vector components outside of any loops in CASE and BETWEEN expressions. Fix for [1a684656]. (check-in: 4c6efe60 user: drh tags: branch-3.15)
21:08
Prevent a possible use-after-free bug in the query optimizer. (check-in: 661b4506 user: drh tags: branch-3.15)
21:01
Disable the query flattener optimization for SELECT statements that are on the RHS of vector IN operators. This is a hack that fixes the bug described in ticket [da7841375186386c]. A better solution that does not disable the query flattener is needed, but this will server for the time being. (check-in: 27438fb4 user: drh tags: branch-3.15)
20:44
Fix a potential use-after-free error during parsing of malformed CREATE TABLE statement. (check-in: 0f956597 user: drh tags: branch-3.15)
20:37
Fix an fts5 problem causing a crash in phrase queries where the first token of the phrase matches one or more rows but some other token within the phrase matches zero. (check-in: 4efd331e user: drh tags: branch-3.15)
20:31
Fix OOM error reporting in the json1_group_object() SQL function. Remove an unreachable branch from the JSON1 logic. (check-in: 6492e57e user: drh tags: branch-3.15)
20:24
Fix the JSON1 extension so that the JSON validator correctly rejects malformed backslash escapes within string literals. (check-in: 7c466283 user: drh tags: branch-3.15)
20:19
Avoid a crash that can occur after an obscure OOM in the built-in INSTR() function. (check-in: 8a55b8e1 user: drh tags: branch-3.15)
20:12
Mark the ICU extension functions as deterministic. (check-in: 8fd2fcce user: drh tags: branch-3.15)
19:43
Handle some obscure "row value misused" cases that could cause segfaults or assertion failures. (check-in: 794763fd user: drh tags: branch-3.15)
19:40
Take care not to try to generate code for the ATTACH and DETACH commands if there were syntax errors during parsing. Fix for ticket [2f1b168ab4d4844] (check-in: f8cf7ff1 user: drh tags: branch-3.15)
19:31
Code constant vector components outside of any loops in CASE and BETWEEN expressions. Fix for [1a684656]. (check-in: c08aff13 user: dan tags: trunk)
14:19
Correctly display the names of SQL functions in the VDBE_PROFILE output. (check-in: 161192f1 user: drh tags: trunk)
2016-11-22
22:23
Merge bug fixes and enhancements from trunk. (Closed-Leaf check-in: e2c4bdf5 user: drh tags: serializable-snapshot)
21:11
Enclose the sqlite3WalSnapshotRecover() routine within (check-in: e7be3183 user: drh tags: serializable-snapshot)
20:29
Add a new requirement mark for CAST expressions. (check-in: bee2859b user: drh tags: trunk)
19:15
Update requirement marks to reflect changes in wording in the documentation. No changes to code. (check-in: f5ac98ef user: drh tags: trunk)
01:26
Remove unnecessary OP_Close opcodes for a size reduction and performance increase. (check-in: 32be7aae user: drh tags: trunk)
2016-11-21
21:33
Remove unnecessary OP_Close opcodes at the end of INSERT operations, resulting in smaller and faster code. (check-in: abeddb6d user: drh tags: trunk)
18:15
Enhance speedtest1.c so that is works on older versions of SQLite, before the introduction of the group_concat() aggregate function. (check-in: 9979ba80 user: drh tags: trunk)
2016-11-20
17:59
Fix harmless compiler warnings. (check-in: b3b7b42d user: drh tags: trunk)
12:00
Prevent a possible use-after-free bug in the query optimizer. (check-in: 0a98c8d7 user: drh tags: trunk)
2016-11-19
18:31
Remove the requirement to open the wal file before sqlite3_snapshot_recover() is called. Also add some comments to new functions. (check-in: 28393c41 user: dan tags: serializable-snapshot)
17:30
Test some extra error conditions in sqlite3_recover_snapshot(). (check-in: db314213 user: dan tags: serializable-snapshot)
17:20
Add another fault-injection test for sqlite3_snapshot_recover(). (check-in: 7e040406 user: dan tags: serializable-snapshot)
16:35
Fix a problem causing sqlite3_snapshot_recover() to return SQLITE_IOERR_SHORT_READ. (check-in: 525f75fa user: dan tags: serializable-snapshot)
14:53
Fix a bug in sqlite3_snapshot_recover() that could cause subsequent read transactions to use out-of-data cache entries. (check-in: 9abeb798 user: dan tags: serializable-snapshot)
2016-11-18
20:49
Add experimental sqlite3_snapshot_recover() API. (check-in: 174a6076 user: dan tags: serializable-snapshot)
18:43
Require that the database handle be in autocommit mode for sqlite3_snapshot_get() to succeed. This is because it may open a read transaction on the database file. (check-in: 83b658da user: dan tags: serializable-snapshot)
18:22
Add tests for snapshot interfaces. (check-in: 1f7ee7af user: dan tags: serializable-snapshot)
14:38
Enhance existing snapshot tests to serialize/deserialize snapshots. No new tests. (check-in: 16b9bf92 user: dan tags: serializable-snapshot)
2016-11-17
21:52
Experimental changes for faster in-memory DB operation for large databases. (Closed-Leaf check-in: 4fe94b0f user: drh tags: memdb-opt)
20:05
Fix a problem in rtreeD.test causing it to fail in OMIT_BUILTIN_TEST builds. (check-in: d6b3779e user: dan tags: trunk)
14:02
When opening the *-shm file for a readonly database, try to open it in read-write mode before falling back to readonly. This is in case some other read/write connection within the same process uses the same file descriptor. (check-in: a07c581e user: dan tags: apple-osx)
13:54
Change the --enable-debug option on configure so that it enables the ".selecttrace" and ".wheretrace" commands in the command-line shell. (check-in: 2331192b user: drh tags: trunk)
13:13
Disable the query flattener optimization for SELECT statements that are on the RHS of vector IN operators. This is a hack that fixes the bug described in ticket [da7841375186386c]. A better solution that does not disable the query flattener is needed, but this will server for the time being. (check-in: 005d5b87 user: drh tags: trunk)
2016-11-15
21:17
Add a needed #include to speedtest1.c. (check-in: 77aeb705 user: drh tags: trunk)
17:37
Experimental changes toward making snapshots serializable. (check-in: b6a81fa1 user: drh tags: serializable-snapshot)
04:00
Change the OP_Last opcode so that it is a no-op if the cursor is already pointing at the end of the b-tree. (check-in: 66347385 user: drh tags: trunk)
02:46
Make the VACUUM command about 9% faster by avoiding unnecessary calls to sqlite3BtreeMovetoUnpacked() while copying rowid tables. (check-in: 5ed0bd38 user: drh tags: trunk)
2016-11-14
20:47
Add 8794 new test cases received from the OSS-FUZZ project. (check-in: 03f75a67 user: drh tags: trunk)
20:08
Fix a potential use-after-free error during parsing of malformed CREATE TABLE statement. (check-in: c5dbc599 user: drh tags: trunk)
18:27
Enhance fuzzcheck.c to accept database-specific configuration parameters in the option CONFIG table. Add the fuzzdata5.db fuzz database containing test cases from OSS-FUZZ. (check-in: a71fad45 user: drh tags: trunk)
17:25
Fix the ossfuzz.c test module so that it does not segfault after a "PRAGMA empty_result_callbacks=1;". Add the ossshell.c program for simple command-line testing of ossfuzz.c. (check-in: 6f2d43ec user: drh tags: trunk)
15:28
Avoid unnecessary calls to sqlite3ExprCacheAffinityChange() when generating OP_MakeRecord opcodes that do not change any register affinities. (check-in: e211c579 user: drh tags: trunk)
08:19
Fix an fts5 problem causing a crash in phrase queries where the first token of the phrase matches one or more rows but some other token within the phrase matches zero. (check-in: e78f6f3b user: dan tags: trunk)
2016-11-11
20:37
Fix harmless compiler warnings in test code for MSVC. (check-in: 7b76be41 user: drh tags: trunk)
19:08
Avoid storing redundant fields in sorter records when the sort-key and data have fields in common. (check-in: b835cf3e user: drh tags: trunk)
19:01
Enhance the OP_IdxInsert opcode so that it can used unpacked key values if they are available. Update the code generator to take advantage of this new capability. The speedtest1.c test is about 2.6% faster as a result. Later: This also introduced bug [30027b613b]. Bummer. (check-in: 925840cf user: drh tags: trunk)
18:45
Merge trunk with this branch. (Closed-Leaf check-in: dd62d2de user: dan tags: sorter-opt)
18:19
Remove obselete and unused logic in the update code generator. Fix the sqlite3BtreeMovetoUnpacked() routine so that it remembers the rowid of the row that it landed on. (Closed-Leaf check-in: 1a587d72 user: drh tags: unpacked-IdxInsert)
18:08
Reenable the SQLITE_EXPR_REF optimization for "SELECT DISTINCT ... ORDER BY" queries. (check-in: 6e2e9d38 user: dan tags: sorter-opt)
17:52
Merge enhancements and bug-fixes from trunk. (check-in: 5515b827 user: drh tags: unpacked-IdxInsert)
17:08
Fix a problem with switching from wal to rollback mode when SQLITE_DBCONFIG_NO_CKPT_ON_CLOSE is configured. (check-in: 46e00162 user: dan tags: trunk)
16:33
Add the test/ossfuzz.c interface adaptor for OSS-FUZZ. Make previsions for testing the adaptor using fuzzcheck.c. (check-in: 119d6ef8 user: drh tags: trunk)
15:49
Handle some obscure "row value misused" cases that could cause segfaults or assertion failures. (check-in: fba5fddb user: dan tags: trunk)
14:54
Back out the "--raw" option on ".read" in the command-line shell. Instead, fix the command-line shell so that if EOF is reached without seeing a final semicolon, it goes ahead and passes the accumulated SQL text to SQLite. (check-in: f98c8ac8 user: drh tags: trunk)
09:51
Add extra tests for the xBestIndex() virtual table method. (check-in: 642a8fba user: dan tags: trunk)
05:19
In the command line shell, avoid using utf8_printf() in a couple places where it is superfluous. (check-in: 6311a8bd user: mistachkin tags: trunk)
04:37
Add the --raw option to the ".read" dot-command of the command-line shell, to cause the named file to be read and sent directly into sqlite3_exec() without any interpretation. (check-in: 09233770 user: drh tags: trunk)
03:37
Take care not to try to generate code for the ATTACH and DETACH commands if there were syntax errors during parsing. Fix for ticket [2f1b168ab4d4844] (check-in: b0ff183b user: drh tags: trunk)
2016-11-10
21:19
Break up the sqlite3BtreeMovetoUnpacked() routine into an eponymous routine and sqlite3BtreeMovetoIntkey(). Each routine specializes in a single kind of btree. (Leaf check-in: 9a64a4f2 user: drh tags: split-moveto)
20:42
Avoid a few unnecessary calls to sqlite3BtreeMovetoUnpacked(). (check-in: eac0e827 user: drh tags: unpacked-IdxInsert)
20:14
Avoid storing redundant fields in sorter records when the sort-key and data have fields in common (as in "SELECT a FROM t1 ORDER BY 1"). (check-in: 0af62fdb user: dan tags: sorter-opt)
17:01
When doing a REPLACE on a WITHOUT ROWID table with no secondary indexes, bypass the OP_NoConflict/OP_Delete sequence and directly overwrite any preexisting row, for a performance increase. (check-in: f7041cbb user: drh tags: unpacked-IdxInsert)
16:07
Remove C++ style comments. No code changes. (check-in: 42f8aa62 user: drh tags: unpacked-IdxInsert)
14:24
Add the unpacked key fields to OP_IdxInsert for INSERT and UPDATE statements. (check-in: 67602e40 user: drh tags: unpacked-IdxInsert)
2016-11-09
20:14
Comment changes that attempt to better explain the behavior of the "seekResult" field on VdbeCursor objects and the seekResult parameter to sqlite3BtreeInsert() and the OPFLAG_USESEEKRESULT flag on insert opcodes. No changes to code. (check-in: 345b46be user: drh tags: unpacked-IdxInsert)
16:03
Fix a simple comment typo. No changes to code. (check-in: 51d0aed8 user: drh tags: unpacked-IdxInsert)
01:46
Fix typo in the CSV extension. (check-in: b4889588 user: mistachkin tags: trunk)
01:38
Make use of the unpacked key on the OP_IdxInsert on sorters with a LIMIT. (check-in: 42db7cd2 user: drh tags: unpacked-IdxInsert)
01:19
Fix a typo on the OP_IdxInsert documentation. No code changes. (check-in: e4acd982 user: drh tags: unpacked-IdxInsert)
01:07
Add a DISTINCT query to speedtest1.c. Also change INSERT OR REPLACE into just REPLACE. (check-in: 754ee844 user: drh tags: trunk)
00:57
Remove the "experimental" marking from the sqlite3_preupdate interfaces. But be sure all the interface definitions are within documentation. (check-in: d6dd2ad3 user: drh tags: trunk)
00:10
Enhance the OP_IdxInsert opcode to optionally accept unpacked key material. (check-in: 89d958ab user: drh tags: unpacked-IdxInsert)
2016-11-08
19:22
Avoid superfluous cursor seeks in "INSERT OR REPLACE" statements. (check-in: bec5b6d4 user: dan tags: trunk)
17:19
Avoid generating OP_TableLock unnecessary instructions on btrees that are not sharable. (check-in: 8cb8516d user: drh tags: trunk)
16:27
Fix the speedtest1.c test program so that it gives closing "%" wildcards on LIKE searches. (check-in: c2f84f63 user: drh tags: trunk)
00:30
Tweaks to speedtest1.c to better reflect mobile usage based on findings in the "Pocket Data" paper by Kennedy, Ajay, Challen, and Ziarek. (check-in: b5733cae user: drh tags: trunk)
2016-11-07
18:18
Fix OOM error reporting in the json1_group_object() SQL function. Remove an unreachable branch from the JSON1 logic. (check-in: b5409849 user: drh tags: trunk)
15:15
Replace some unreachable branches from the JSON1 extension with assert(). (check-in: 145cd13e user: drh tags: trunk)
13:37
Fix the JSON1 extension so that the JSON validator correctly rejects malformed backslash escapes within string literals. (check-in: 7a635391 user: drh tags: trunk)
2016-11-04
19:09
Merge recent trunk changes, including all the fixes that appeared in version 3.15.1. (check-in: 0e5ffd91 user: drh tags: apple-osx)
12:08
Version 3.15.1 (check-in: 1136863c user: drh tags: release, version-3.15.1, branch-3.15)
12:05
Avoid a crash that can occur after an obscure OOM in the built-in INSTR() function. (check-in: b86b79c4 user: dan tags: trunk)
12:03
Mark the ICU extension functions as deterministic. (check-in: afbbb6c6 user: drh tags: trunk)
11:23
Add an assert() to silence a coverity false-positive. (check-in: 97354093 user: drh tags: trunk)
07:25
Fix a test case in vacuum5.test so that it works with the in-memory journal permutation. (check-in: 81c86401 user: dan tags: trunk)
2016-11-03
21:35
Replace a couple more references to Tcl 8.5 in the Windows build tools. (check-in: 1d0e43c5 user: mistachkin tags: trunk)
19:27
Fix a test case in vacuum5.test so that it works with -DSQLITE_TEMP_STORE=2 and -DSQLITE_ENABLE_ATOMIC_WRITE.. (check-in: 752e14d9 user: dan tags: branch-3.15)
19:24
Fix a test case in vacuum5.test sot aht it works with -DSQLITE_ENABLE_ATOMIC_WRITE. (check-in: 3ca80010 user: dan tags: trunk)
19:01
Fix a test case in vacuum5.test so that it works with -DSQLITE_TEMP_STORE=2. (check-in: f5dd95e2 user: dan tags: trunk)
18:49
Add the SQLITE_FCNTL_WIN32_GET_HANDLE file control. (check-in: 5221e3cc user: drh tags: branch-3.15)
18:44
Increase the version number to 3.15.1. (check-in: ccdcce46 user: drh tags: branch-3.15)
18:38
Escape non-ASCII character from an ICU extension comment. (check-in: ee16fedd user: drh tags: branch-3.15)
18:36
Fix an issue that was causing the new database image to be assembled entirely in heap memory when VACUUMing a database, even if it should use a temp file. This could cause SQLITE_NOMEM errors when vacuuming very large databases on 32-bit systems. (check-in: 3fae483f user: drh tags: branch-3.15)
18:35
Make sure left-join markings are transferred to the virtual scalar subexpressions when decomposing a vector comparison in the ON clause of a LEFT JOIN. Fix for ticket [fef4bb4bd9185ec8f]. (check-in: aba1e22b user: drh tags: branch-3.15)
18:33
If all branches of an OR optimize scan that is the rhs of a LEFT JOIN use the same index, set the index cursor to return NULL values if there are no matches for a row on the lhs. Fix for ticket [34a579141b2c5ac] (check-in: 54eeddec user: drh tags: branch-3.15)
18:31
In the sessions module, avoid recording a change if an UPDATE statement overwrites a column with REAL affinity containing an integer value with the same value. (check-in: 0fc4f0f4 user: drh tags: branch-3.15)
18:30
Avoid using the "direct overflow read" optimization to read large blobs if the pager layer has a wal file open - even if the database header indicates that the db is not a wal database. (check-in: 63cfe197 user: drh tags: branch-3.15)
13:12
In the command-line shell, fix the .imposter command so that it works with indexes on expressions. Fix the ".mode quote" mode so that it works with ".headers ON". (check-in: ba9873da user: drh tags: trunk)
13:01
Add the ".imposter" command to the command-line shell. (check-in: be3ec8fd user: drh tags: trunk)
02:25
Fix an possible NULL pointer deference in the command-line shell that can occur when using imposter mode. (check-in: ad08753a user: drh tags: trunk)
2016-11-02
19:49
Adjust the MSVC makefile to use Tcl8.6 by default. (check-in: def8f598 user: drh tags: trunk)
14:50
Fix an issue that was causing the new database image to be assembled entirely in heap memory when VACUUMing a database, even if it should use a temp file. This could cause SQLITE_NOMEM errors when vacuuming very large databases on 32-bit systems. (check-in: 30288453 user: dan tags: trunk)
13:18
Fix a requirement mark. No changes to code. (check-in: d18f61b7 user: drh tags: trunk)
02:17
Use NOT NULL partial indexes when the query is constrained by a single LIKE or GLOB. (check-in: 9b3a3b41 user: drh tags: trunk)
2016-10-31
16:49
Add the SQLITE_DBCONFIG_NO_CKPT_ON_CLOSE sqlite3_dbconfig() option - for disabling SQLite's default checkpoint-on-close behaviour. (check-in: 6d142025 user: dan tags: trunk)
16:16
Add the SQLITE_DBCONFIG_NO_CKPT_ON_CLOSE sqlite3_dbconfig() option - for disabling SQLite's default checkpoint-on-close behaviour. (Closed-Leaf check-in: 093d2fc2 user: dan tags: no-ckpt-on-close)
2016-10-27
14:51
Remove the mutex from test_multiplex.c. (check-in: 6374978e user: dan tags: trunk)
01:02
Make sure left-join markings are transferred to the virtual scalar subexpressions when decomposing a vector comparison in the ON clause of a LEFT JOIN. Fix for ticket [fef4bb4bd9185ec8f]. (check-in: 619f5cc7 user: drh tags: trunk)
2016-10-26
18:41
Small size reduction and performance enhancement in whereScanInit(). (check-in: c23f5ef3 user: drh tags: trunk)
17:57
Small size reduction and performance improvement in whereScanNext(). (check-in: d861ee17 user: drh tags: trunk)
16:05
If all branches of an OR optimize scan that is the rhs of a LEFT JOIN use the same index, set the index cursor to return NULL values if there are no matches for a row on the lhs. Fix for ticket [34a579141b2c5ac] (check-in: ec9dab80 user: dan tags: trunk)
15:46
If all branches of an OR optimize scan that is the rhs of a LEFT JOIN use the same index, set the index cursor to return NULL values if there are no matches for a row on the lhs. (Closed-Leaf check-in: e7b9bc67 user: dan tags: leftjoin-or-fix)
13:58
Merge the SQLITE_ENABLE_URI_00_ERROR compile-time option. (check-in: 86675ae0 user: drh tags: trunk)
13:44
Add extra tests to check the result of including a %00 escape in a URI when ENABLE_URI_00_ERROR is defined. (Closed-Leaf check-in: 1aaa06e3 user: dan tags: uri-00-error)
12:58
Add an elapsed-time output for faststat1.c. (check-in: 71167951 user: drh tags: est_count_pragma)
12:15
Add compile time option SQLITE_ENABLE_URI_00_ERROR. If defined, any "%00" escape found in a URI is treated as an error. (check-in: e8a9bfec user: dan tags: uri-00-error)
2016-10-25
19:39
Simplifications to faststat1.c. Fix a bug in sqlite3MovetoProportional() for very large b-trees. (check-in: f7f78147 user: drh tags: est_count_pragma)
19:21
Further refinements to the faststat1.c utility. (check-in: b051fd19 user: drh tags: est_count_pragma)
18:28
Enhance faststat1.c to deal better with WITHOUT ROWID tables. (check-in: 65444f2e user: drh tags: est_count_pragma)
17:28
Merge recent trunk changes, and especially the PRAGMA index_info enhancement which is needed on this branch. (check-in: c3570e46 user: drh tags: est_count_pragma)
15:39
Enhance the "PRAGMA index_info" and "PRAGMA index_xinfo" statements so that they work on WITHOUT ROWID tables and provide information about the underlying index btree that implements the WITHOUT ROWID table. (Leaf check-in: fe49fb03 user: drh tags: index-info-on-table)
15:06
Add test case to demonstrate a "BEGIN EXCLUSIVE" command returning SQLITE_BUSY_SNAPSHOT. (check-in: b1158564 user: dan tags: trunk)
13:57
First attempt at a utility program to compute sqlite_stat1 without doing a full table scan. (check-in: 7b83581a user: drh tags: est_count_pragma)
2016-10-24
01:01
Performance optimization in moveToRoot(). Avoid repeated validity checking of the root page on each call. Once is enough. (check-in: 98795c2d user: drh tags: trunk)
2016-10-22
01:22
In the sessions module, avoid recording a change if an UPDATE statement overwrites a column with REAL affinity containing an integer value with the same value. (check-in: 5f3e6028 user: drh tags: trunk)
2016-10-21
21:21
In the sessions module, avoid recording a change if an UPDATE statement overwrites a column with REAL affinity containing an integer value with the same value. (Closed-Leaf check-in: b861328a user: dan tags: sessions-fix)
18:01
When reading from an index, the shared-cache lock must be on the corresponding table. (check-in: 04fe12b5 user: drh tags: est_count_pragma)
17:45
Merge updates from trunk, and especially the ".mode quote" enhancement to the shell. (check-in: 0c8a5b88 user: drh tags: est_count_pragma)
17:39
Add ".mode quote" to the command-line shell. (check-in: c4f5fa78 user: drh tags: trunk)
17:25
Add the btree_sample(INDEX,LOCATION,LIMIT) pragma. (check-in: affc2ef5 user: drh tags: est_count_pragma)
15:36
Fix problems in the est_count pragma for indexes and WITHOUT ROWID tables. (check-in: c39fd9b8 user: drh tags: est_count_pragma)
10:49
Avoid using the "direct overflow read" optimization to read large blobs if the pager layer has a wal file open - even if the database header indicates that the db is not a wal database. (check-in: b54c15f1 user: dan tags: trunk)
10:09
If SQLITE_DIRECT_OVERFLOW_READ is defined at compile time, add "DIRECT_OVERFLOW_READ" to the list of symbols returned by "PRAGMA compile_options". (check-in: b7d9727b user: dan tags: trunk)
2016-10-20
22:02
Experimental est_count pragma. (check-in: 340822af user: drh tags: est_count_pragma)
18:20
Add the ability for the PRAGMA statement to accept multiple arguments. Currently all arguments other than the first are ignored. (Leaf check-in: fd81d8a4 user: drh tags: multi-arg-pragma)
11:48
Add test case to incrblob4.test. No changes to code. (check-in: fd9c6ff0 user: dan tags: trunk)
2016-10-18
16:36
Minor simplification of the comparison opcodes. (check-in: 56474ebc user: drh tags: trunk)
15:29
Fix harmless uninitialized variable warnings in speedtest1.c. (check-in: 2aa7a03b user: drh tags: trunk)
14:35
Add the --repeat option to speedtest1.c and speed-check.sh. (check-in: 6ca0e06b user: drh tags: trunk)
2016-10-17
18:44
Escape non-ASCII character from an ICU extension comment. (check-in: 5ec02ecf user: mistachkin tags: trunk)
18:33
Corrections to a couple recently added 'filectrl.test' results. (check-in: 3d89dc45 user: mistachkin tags: trunk)
15:28
Changes to support interrupting a checkpoint using sqlite3_interrupt(). (check-in: c88d36e2 user: dan tags: trunk)
00:48
Small size reduction and performance increase in the string duplicator. (check-in: cda998f0 user: drh tags: trunk)
2016-10-15
18:37
Add documentation for the OP_SorterInsert opcode, formerly omitted by mistake. No changes to code. (check-in: 16d88a90 user: drh tags: trunk)
08:56
Add test case to hook.test. No changes to non-test code. (check-in: 2674d779 user: dan tags: trunk)
2016-10-14
21:27
Add the SQLITE_FCNTL_WIN32_GET_HANDLE file control. (check-in: 16039be3 user: mistachkin tags: trunk)
21:16
Bump the version number in anticipation of the next development cycle. (check-in: aa6de9c4 user: drh tags: trunk)
10:20
Version 3.15.0 (check-in: 70787558 user: drh tags: trunk, release, version-3.15.0)
2016-10-13
12:56
Take care to avoid integer overflow when doing the initial page cache allocation with an excessively large cache_size setting. (check-in: 4d66ac98 user: drh tags: trunk)
2016-10-12
18:55
Avoid reading the -1-th element of an array in the query planner. Fix to a bug introduced by check-in [8e2b25f9b8a7] from earlier today. Curiously, the problem only appeared on 32-bit systems. (check-in: 443913d5 user: drh tags: trunk)
18:26
Add to sqlite3_analyzer command-line options --version and --tclsh, and also the undocumented --debug option. (check-in: e87d02d2 user: drh tags: trunk)
15:15
New testcase() macros to ensure coverage of the ORDER BY LIMIT optimization code in where.c. (check-in: 61f05269 user: drh tags: trunk)
14:48
Fix a problem affecting queries that match the pattern (... WHERE ipk IN (....) ORDER BY ? LIMIT ?). Fix for [96c1454c]. (check-in: 8e2b25f9 user: dan tags: trunk)
2016-10-10
14:48
Remove a stray line from session4.test causing a memory leak. No changes to SQLite code. (check-in: 6624c496 user: dan tags: trunk)
14:34
When handling ORDER BY expressions, do not assume all values of an indexed expressions are distinct. Fix for [4766f444]. (check-in: aebe429e user: dan tags: trunk)
13:29
Make sure indexes on expressions skip over initial NULL values in the index. Fix for ticket [4baa464912129477f3c9] (check-in: 71797ba4 user: drh tags: trunk)
10:06
Fix a case in which the rtree module was ignoring an OOM while generating an error message. (check-in: 788f86be user: dan tags: trunk)
2016-10-08
16:10
Update test/trace3.test to account for the fact that casting a pointer to a 64-bit signed integer might produce a negative value. (check-in: 010ec225 user: dan tags: trunk)
13:34
Corrections to Lemon documentation. No SQLite changes. (check-in: c568ae3e user: drh tags: trunk)
11:55
Add missing "static" to internal function whereRangeVectorLen(). (check-in: cbed3d75 user: dan tags: trunk)
09:21
Updates to the tool/warnings.sh script to make it easier to use on OpenBSD. (check-in: dbe2997f user: dan tags: trunk)
2016-10-05
20:14
Fix some compiler warnings in fts5. (check-in: b066637b user: dan tags: trunk)
20:10
Fix harmless compiler warnings. (check-in: 2dde3375 user: drh tags: trunk)
15:04
Add a test case for the fix in [a596b7c6cc]. (check-in: b2ef39b1 user: dan tags: trunk)
15:02
Do not set MEM_Blob unless the appropriate fields of the Mem object have been correctly initialized. (check-in: a596b7c6 user: drh tags: trunk)
01:33
Fix the backup_malloc.test script so that it works on Windows. (check-in: d080012d user: drh tags: trunk)
2016-10-04
12:20
Fix a problem with the lempar.c parser template. (check-in: 3a9d802f user: drh tags: trunk)
00:47
Do not bother to initialize unused fields in the BtreePayload object. (check-in: b10d0f93 user: drh tags: trunk)
2016-10-03
18:13
Allow deterministic functions in the WHERE clause of a partial index. (check-in: c6e9b942 user: drh tags: trunk)
16:33
Size and performance optimizations on the sqlite3ExprAssignVarNumber() routine. (check-in: 109852e5 user: drh tags: trunk)
15:28
Avoid unnecessary strlen() calls in sqlite3ExprAssignVarNumber() by passing in the token length from the parser. (check-in: d15ae2e5 user: drh tags: trunk)
14:44
Avoid unnecessary memset() calls in the query optimizer. (check-in: 5bb7cde0 user: drh tags: trunk)
12:56
Fix an obsolete comment on a variable definition in expr.c. No code changes. (check-in: 012ff6dd user: drh tags: trunk)
02:59
Make sure the sqlite_sequence table is updated when an optional xfer optimization is used. Fix for ticket [7b3328086a5c1]. (check-in: 81e4d4f8 user: drh tags: trunk)
01:21
Allocate Parse objects off of the stack where appropriate for a substantial performance increase and a size reduction. (check-in: ea8affa9 user: drh tags: trunk)
2016-10-01
23:55
Add SQLITE_USE_ALLOCA to the --lean configuration in speed-check.sh. (check-in: fe07609e user: drh tags: trunk)
21:43
Rearrange fields in the Parse object to reduce the amount of initialization required. (check-in: 361940b4 user: drh tags: trunk)
20:43
Reduce the width of a memset() to avoid double-initializing some variables. (check-in: 34edbfd1 user: drh tags: trunk)
19:32
Remove an unnecessary memset() call. (check-in: a76bff74 user: drh tags: trunk)
19:21
Use sqlite3DbMallocRawNN() where appropriate, instead of sqlite3DbMallocRaw(). (check-in: 54a449a4 user: drh tags: trunk)
16:53
Make sure deleting an unused prepared statement does not reference uninitialized fields in the structure. (check-in: 7983eef0 user: drh tags: trunk)
11:39
Avoid accessing Vdbe.pc if it is uninitialized. Check Vdbe.magic first. (check-in: 6ac6e446 user: drh tags: trunk)
00:37
Avoid unnecessary zeroing of fields in the Vdbe object when it is allocated. (check-in: 1e21bbe8 user: drh tags: trunk)
2016-09-30
22:24
Avoid initializing the column-cache section of the Parse object, since entries in the cache will be initialized as they are used, and avoiding the initial memset() saves many CPU cycles. (check-in: 63cf7eaf user: drh tags: trunk)
21:20
Fix an always-true conditional left over from the previous commit. (check-in: ab12fce3 user: drh tags: trunk)
20:22
Fix the Parse.aColCache column cache so that all of the valid entries are in the first Parse.nColCache slots. (check-in: 60285020 user: drh tags: trunk)
19:14
Another simplification to the Parse object. (check-in: c1419727 user: drh tags: trunk)
18:35
Remove unnecessary fields from the Parse object. (check-in: 814e41da user: drh tags: trunk)
17:46
Avoid unnecessary Mem initializations when generating a new sqlite3_stmt object. (check-in: 47ae1cda user: drh tags: trunk)
2016-09-29
20:28
Small size and performance optimization to sqlite3VdbeChangeP5(). (check-in: 3c93c8f5 user: drh tags: trunk)
19:50
Remove the peep-hole optimization of removing OP_Close opcodes that come before OP_Halt, as the extra work of removing those opcodes uses more cycles than just running them. (check-in: 984a96d7 user: drh tags: trunk)
19:27
Size and performance optimization to sqlite3ValueFromExpr() (check-in: 945f82bc user: drh tags: trunk)
15:53
Clarification and typo-fixes in comments. No changes to code. (check-in: e2cd79aa user: drh tags: trunk)
2016-09-28
20:42
Use Knuth multiplicative hashing for the symbol table. (check-in: cc29ddd6 user: drh tags: trunk)
16:15
Merge recent enhancements from trunk. (check-in: 06014163 user: drh tags: apple-osx)
16:05
Two more typo fixes in comments. (check-in: 40c0fb0a user: drh tags: trunk)
16:04
Fix typos in comments. No code changes. (check-in: 5bbd071d user: drh tags: trunk)
2016-09-27
00:09
Update requirements marks due to improvements in the wording of documentation. No changes to code. (check-in: bf903b2e user: drh tags: trunk)
00:03
Fix FTS3 test cases due to the flattener change of the previous check-in. (check-in: 0c8b9b21 user: drh tags: trunk)
2016-09-26
14:39
When flattening a query of the form "SELECT * FROM (SELECT * FROM tbl WHERE x=?) WHERE y=?", ensure that the final WHERE clause is "x=? AND y=?" instead of "y=? AND x=?". Although it is still not guaranteed, this makes the order in which WHERE clause terms are processed comport more closely to users expectations. (check-in: cf7f9e6d user: dan tags: trunk)
12:38
Inline the relevent parts of sqlite3ExprAlloc() into spanExpr(), for a performance improvement. (check-in: fe89225e user: drh tags: trunk)
2016-09-24
17:42
Omit the LikeOp object from the parser. Change more sqlite3PExpr() calls into sqlite3ExprAlloc() calls. (check-in: 795454a3 user: drh tags: trunk)
01:41
Add -DSQLITE_MAX_EXPR_DEPTH=0 to the --lean option on speed-check.sh. (check-in: a8cb1390 user: drh tags: trunk)
2016-09-23
21:36
Add the EP_Leaf flag bit to the Expr.flags field to indicate Expr nodes that do not have substructure. Use that bit to avoid unnecessary recursion. (check-in: 8a6ea455 user: drh tags: trunk)
20:59
Use sqlite3ExprAlloc() instead of sqlite3PExpr() for leaf nodes in the expression tree, where appropriate. This is both smaller and faster. (check-in: afac0709 user: drh tags: trunk)
18:13
New test case for the ORDER BY LIMIT optimization. (check-in: 9a5a489d user: drh tags: trunk)
2016-09-22
21:37
Fix a potential null-pointer dereference and crash in the case where one thread is calling sqlite3_column_text() and another thread is calling sqlite3_step() on the same prepared statement at the same instant. (check-in: ee1382a3 user: drh tags: trunk)
18:53
Remove the internal sqlite3CodeOnce() interface, replacing it with a direct call to sqlite3VdbeAddOp0(v,OP_Once). Slightly smaller and faster. (check-in: c3774c6a user: drh tags: trunk)
18:46
Makefile changes to support building winsqlite3.dll using STDCALL rather than CDECL. (check-in: 5e892d60 user: mistachkin tags: trunk)
2016-09-21
23:58
Fix speedtest1.c so that it works with SQLITE_OMIT_DEPRECATED. Add the --lean and --cachesize options to speed-check.sh. (check-in: 7785b3a2 user: drh tags: trunk)
19:43
Fix a problem with the fts5 auxiliary function API causing a crash when a query contained a phrase of zero terms. (check-in: 0741812d user: dan tags: trunk)
19:00
Do not run vacuummem.test if ENABLE_MEMORY_MANAGEMENT is defined. (check-in: 4a613d85 user: dan tags: trunk)
18:18
Fix obsolete requirements marks. Add new tests for VACUUM of attached databases. (check-in: 2b44f9aa user: drh tags: trunk)
17:53
Avoid running vacuum5.test in auto-vacuum mode. (check-in: be184bef user: dan tags: trunk)
17:47
Do not run test "delete_db.test" with either the journaltest or inmemoryjournal permutations. Ensure that the multiplexor tests in delete_db.test are performed in non-autovacuum mode. (check-in: 46b7d19e user: dan tags: trunk)
14:41
Update the fts5vocab table to handle "ORDER BY term" efficiently. (check-in: d4928fb5 user: dan tags: trunk)
2016-09-20
22:04
Improved implementation of 64-bit signed integer multiply that correctly detects overflow (and promotes to floating-point) in some corner cases. Fix for ticket [1ec41379c9c1e400] (check-in: db3ebd7c user: drh tags: trunk)
17:49
Fix harmless compiler warning. (check-in: 72429063 user: mistachkin tags: trunk)
17:38
Fix typo in comment. (check-in: bd3ecbb5 user: mistachkin tags: trunk)
01:46
Further performance enhancements in zeroblob() handling. (check-in: 21b0e3b7 user: drh tags: trunk)
01:19
Avoid unnecessary calls to ExpandBlob() for smaller and faster code. (check-in: 5e196fd1 user: drh tags: trunk)
2016-09-19
23:39
Very small optimization in the bytecode engine. (check-in: 46002511 user: drh tags: trunk)
11:00
Fix a segfault introduced by the row-value enhancement that comes up on a skip-scan where the first term of the index is unconstrained and the second term is of the form "columm IN (SELECT...)". (check-in: 2401ea5a user: drh tags: trunk)
10:24
Improved performance in sqlite3ExprCodeTarget(). (check-in: 75146165 user: drh tags: trunk)
02:19
Small performance optimization in the expression walker. (check-in: c6e6afb9 user: drh tags: trunk)
2016-09-18
16:08
Revise the implementation of OP_Once so that it is smaller, faster, and uses less memory. This also fixes an obscure bug introduced 3 days ago by check-in [5990a1bdb4a073]. (check-in: 6bf5ba10 user: drh tags: trunk)
2016-09-17
22:46
Minor enhancement to the TclKit batch tool: it should download the TclKit EXE alone when the TCLKIT_NOENV environment variable is set. (check-in: baceb988 user: mistachkin tags: trunk)
19:34
Small performance optimization in the freeSpace() routine of btree.c. (check-in: 63962e2a user: drh tags: trunk)
17:39
Improved description of the ElseNotEq opcode in comments. No changes to code. (check-in: e011435e user: drh tags: trunk)
2016-09-16
20:39
Fix typo in the name of a Win32 string conversion routine. Replace several uses of raw_printf() with utf8_printf() in the shell. (check-in: 6fc95dd7 user: mistachkin tags: trunk)
19:42
If no KEY is specified on ATTACH, copy the main schema key if the key size is non-zero. In other words, allow for -1 as a valid key size in support of the --textkey to SEE. (check-in: 03573887 user: drh tags: trunk)
19:11
Fix problems with the fileRead() function in the command-line shell. Fix a harmless memory leak in the command-line shell, to make validation testing easier. (check-in: e660402e user: drh tags: trunk)
18:53
Remove the #ifdef SQLITE_DEBUG from around the testcase_glob() routine in the command-line shell. (check-in: 9885dac4 user: drh tags: trunk)
17:50
Merge recent changes from trunk. (check-in: e3d9efa2 user: drh tags: apple-osx)
16:30
Fix a problem causing incorrect code to be generated for IN constraints like "a IN (1, 2, 3)" where column "a" is a rowid column with an extra UNIQUE index created on it. Ticket [0eab1ac759]. (check-in: a92aee55 user: dan tags: trunk)
15:42
Replace a faulty assert() with a testcase() to assure the condition is tested. Ticket [0eab1ac7591f]. (check-in: a49bc0a8 user: drh tags: trunk)
14:50
Add WHERETRACE macros showing every entry and exit from whereLoopAddBtreeIndex(). (check-in: e7653dec user: drh tags: trunk)
11:53
Fix SQLITE_OMIT_AUTHORIZATION so that it compiles cleanly. (check-in: a3e3b3e1 user: drh tags: trunk)
02:52
In the shell, make ".testcase" and ".check" available in all configurations, not just SQLITE_DEBUG. All testcases to be named. Show the result of each testcase on stdout. (check-in: 792afefa user: drh tags: trunk)
00:26
In the command-line shell, add the --new option to the ".open" command. Also, report the current database filename as part of the ".show" command. (check-in: 8e5c9203 user: drh tags: trunk)
2016-09-15
21:35
Add the ".testcase" and ".check" dot-commands in the shell, when compiled using SQLITE_DEBUG. (check-in: 62289f27 user: drh tags: trunk)
19:15
Omit the sqlite3Apis constant object when compiling with SQLITE_OMIT_LOAD_EXTENSION, since it is not used. (check-in: 7b104613 user: drh tags: trunk)
14:54
Remove no-op code (adding an assert() to prove that the code was no-op) in the text affinity logic of the bytecode engine comparison operators. (check-in: b9f5bdbf user: drh tags: trunk)
05:47
Fix a typo in a comment in sqlite.h.in. No changes to code. (check-in: 12df7152 user: dan tags: trunk)
00:32
Slight size reduction and performance increase in sqlite3VdbeHalt(). (check-in: 5990a1bd user: drh tags: trunk)
00:04
Optimization to vdbeRecordCompareInt() makes it slightly smaller and faster. (check-in: 5a2a7712 user: drh tags: trunk)
2016-09-14
18:56
Update comments used to generate API documentation to include the new "[dateof:3.X.Y]" notation for dates of release where appropriate. No changes to code. (check-in: 55f7f8ee user: drh tags: trunk)
2016-09-12
14:23
Fix a bug in sqldiff causing it to confuse blobs zero bytes in size with NULL values. (check-in: 264e9c75 user: dan tags: trunk)
13:32
Add missing "finish_test" command to vacuum5.test. (check-in: 02f40c3b user: dan tags: trunk)
09:28
Have releasetest.tcl delete all test files except for "testfixture", "testfixture.exe", "sqlite3", "sqlite3.exe", "test.log" and "test-out.txt" after each test run is finished. Passing the "--keep" option restores the old behaviour. (check-in: e0e042dc user: dan tags: trunk)
2016-09-10
19:51
Changes to give a warning-free build with SQLITE_OMIT_INCRBLOB and SQLITE_OMIT_SHARED_CACHE. (check-in: 711c5917 user: drh tags: trunk)
14:43
Add test code to delete an SQLite database and its auxiliary files. (check-in: 8a9d1281 user: dan tags: trunk)
2016-09-09
20:26
Make the new file test_delete.c portable to MSVC on Windows. (Closed-Leaf check-in: 0e3ace39 user: mistachkin tags: test_delete)
20:23
Remove an unnecessary call to sqlite3VdbeMemExpandBlob() when casting to text. (check-in: 19e2e595 user: drh tags: trunk)
20:00
Add new file test_delete.c, containing test code for deleting an sqlite database. (check-in: fb0b95e9 user: dan tags: test_delete)
19:33
Performance improvements to the comparison operators in the bytecode engine. (check-in: 37803b19 user: drh tags: trunk)
15:12
Fix a out-of-order variable declaration for some compile-time configurations. (check-in: 6ac932c9 user: drh tags: trunk)
13:23
In releasetest.tcl, the --srcdir option is automatic, so no need to mention it in the --help output. (check-in: 2bdd838e user: drh tags: trunk)
13:17
Fix an obsolete comment in the releasetest.tcl source code. (check-in: afaaa276 user: drh tags: trunk)
12:29
In releasetest.tcl: Improved testing for uncommitted changes. If uncommitted changes are seen, exit immediately unless the --force option is used. (check-in: 9a14e6c4 user: drh tags: trunk)
2016-09-08
23:16
Remove an extra space before function names in the amalgamation. (Closed-Leaf check-in: 20f3c743 user: mistachkin tags: callbackConv)
23:12
Issue a warning and prompt the user to continue if the releasetest.tcl script is run in a checkout with uncommitted changes. (check-in: 30e917ff user: drh tags: trunk)
2016-09-07
19:54
Add support for row-value comparisons, including IN operators, and row-value updates in the UPDATE statement. (check-in: ddb5f055 user: drh tags: trunk)
19:37
Fix a problem handling expressions like "(a, b) IN (SELECT ... ORDER BY 1, 2)" when there is an index on "a" but not "b". (Closed-Leaf check-in: 7f2c5c9e user: dan tags: rowvalue)
18:20
Merge recent enhancements from trunk. (check-in: 00759af0 user: drh tags: apple-osx)
18:11
Add the ext/misc/memvfs.c extension that implements a VFS for read-only database files contained in memory. (check-in: 12b7782a user: drh tags: trunk)
13:30
Simplify the affinity handling logic in codeAllEqualityTerms(). Logically the same, just a little easier to read and understand. (check-in: bbab9621 user: drh tags: rowvalue)
13:12
Merge fixes from trunk. (check-in: 193f036c user: drh tags: rowvalue)
10:10
Fix the ".read" command in the command-line shell so that it understands that the input is not interactive. (check-in: d8451fe8 user: drh tags: trunk)
01:51
The ORDER BY LIMIT optimization is not valid unless the inner-most IN operator loop is actually used by the query plan. Fix for ticket [0c4df46116e90f92]. (check-in: 820644b8 user: drh tags: trunk)
2016-09-06
19:08
Remove obsolete vector-IN test cases. Fix a bad testcase() macro. (check-in: ab3f8f19 user: drh tags: rowvalue)
18:51
Fix corer cases of vector IN operators where the RHS is a compound SELECT that includes an ORDER BY clause. (check-in: 8329ac6f user: drh tags: rowvalue)
17:21
Remove an unnecessary branch from expr.c. (check-in: 7cc9746c user: dan tags: rowvalue)
17:13
Avoid unnecessary memory allocations for aiMap in codeEqualityTerm(). (check-in: 70319c3d user: drh tags: rowvalue)
16:53
Avoid a NULL pointer deref in codeAllEqualityConstraints() following an OOM. (check-in: c6e528c1 user: drh tags: rowvalue)
16:33
Simplify the fix in commit [7d9bd22c]. (check-in: bd5a3420 user: dan tags: rowvalue)
15:25
Fix the header comment on codeEqualityTerm(). (check-in: b7e710e4 user: drh tags: rowvalue)
14:58
Fix a problem handling (a, b) IN (SELECT ...) expressions when there is an index on just one of "a" or "b". (check-in: 231c72d9 user: dan tags: rowvalue)
14:37
Enhance the sqlite3GetTempRange() and sqlite3ReleaseTempRange() internal routines so that they use sqlite3GetTempReg() and sqlite3ReleaseTempReg() when nReg==1. (check-in: 4071da2f user: drh tags: rowvalue)
12:04
Fix a typo in a comment in expr.c. (check-in: 288e934f user: dan tags: rowvalue)
2016-09-05
22:50
Simplified logic to extract a column from a row value. (check-in: e8f105c3 user: drh tags: rowvalue)
20:46
Sync up the MSVC makefiles. (check-in: f66d0d8c user: mistachkin tags: callbackConv)
20:40
Merge updates from trunk. (check-in: ea635bf9 user: mistachkin tags: callbackConv)
19:57
Fix an assert() so that it does C-compiler does not combine an assert() conditional with a production code conditional and thereby confuse the mutation testing script. (check-in: 2fa5288a user: drh tags: rowvalue)
15:02
Simplifications to the SQLITE_KEEPNULL flag on VDBE comparison operators. (check-in: 96269f01 user: drh tags: rowvalue)
12:12
Do vector comparison size checking early - at name resolution time - to forestall future problems. (check-in: ae127bcc user: drh tags: rowvalue)
12:02
Catch vector size mismatch problems during name resolution to avoid later problems. (Closed-Leaf check-in: 56562a03 user: drh tags: early-vector-size-check)
09:44
Fix a crash that could occur under certain circumstances if the vectors on either side of a comparison operator were of a different size. (check-in: 42670935 user: dan tags: rowvalue)
2016-09-03
19:52
Fix a problem causing the affinity of sub-select row-value elements to be ignored in some contextes. (check-in: 7d9bd22c user: dan tags: rowvalue)
16:24
Merge the fuzzershell enhancement from trunk. (check-in: ed206048 user: drh tags: rowvalue)
16:23
Build the generate_series(START,END,STEP) table-valued function into fuzzershell, to make it easier to construct compact test cases that contain actual data. (check-in: 672c21bc user: drh tags: trunk)
15:31
Consider the affinity of "b" when using an "a IN (SELECT b ...)" expression with an index on "a". Fix for [199df416]. (check-in: f5e49855 user: dan tags: rowvalue)
01:46
Performance optimizations. (check-in: f1d06c49 user: drh tags: rowvalue)
2016-09-02
23:56
Merge recent changes from trunk. (check-in: c7271fbd user: drh tags: rowvalue)
21:34
Add a test case for the OOM handled by the previous commit. (check-in: 9bdf7ca1 user: dan tags: trunk)
21:17
Correctly detect an OOM occurring in the setDestPgsz() subroutine of backup. (check-in: d9d8a048 user: drh tags: trunk)
17:27
The previous commit could miss an SQLITE_NOMEM error. This commit catches it. (Closed-Leaf check-in: 3a9f4750 user: dan tags: mistake)
17:18
Within a backup operation, ensure that a read-transaction is opened on the source database before its page size is read. This ensures the page-size used to write to the backup database is the same as its actual page-size, which is important for ZipVFS databases. (check-in: 7908fc60 user: dan tags: trunk)
2016-09-01
14:03
Have "sqldiff --rbu" ignore rows with NULL values in primary key fields. RBU can't handle such rows and the documentation already says sqldiff ignores them. Because the code now uses "=" instead of "IS" to filter on primary key columns, diffs on virtual tables are faster now too. (check-in: f4ba894a user: dan tags: trunk)
09:35
If SQLITE_ENABLE_ZIPVFS is defined, journal_mode=off is configured and a savepoint or statement rollback is attempted, move the pager into the error state to prevent the transaction from being committed. This makes it safe to use journal_mode=off with zipvfs under some conditions. (check-in: 38d31e18 user: dan tags: trunk)
2016-08-29
14:18
Use some of the example code from the sessions documenatation in the sessions test cases. (check-in: 6602974d user: dan tags: trunk)
2016-08-27
20:21
Fix typos in comments. No changes to running code. (check-in: a07269f2 user: drh tags: trunk)
18:35
Fix some comments in sqlite3session.h. No changes to code. (check-in: 78cd64e2 user: dan tags: trunk)
18:34
Fixes to fts5 snippet() function. (check-in: d464a7b1 user: dan tags: trunk)
14:13
Merge updates from trunk. (check-in: 082fd5f8 user: drh tags: rowvalue)
14:05
Fix the extra comments (added with -DSQLITE_ENABLE_EXPLAIN_COMMENTS) so that the sense of <, <=, >, and >= tests is correct and so that the SQLITE_STOREP2 version is shown correctly. These changes are already in the rowvalue branch but are added here since they are technically unrelated to rowvalue. (check-in: 4d43c469 user: drh tags: trunk)
01:41
Fix the "Synopsis" on the OP_Lt, OP_Le, OP_Gt, and OP_Ge opcodes, which has been backwards for time out of mind. (check-in: f51248dc user: drh tags: rowvalue)
2016-08-26
22:09
Fix a minor problem in sqlite3FindInIndex() related to rowids being used as part of the index. (check-in: 829f802b user: drh tags: rowvalue)
21:15
Fix the sqlite3FindInIndex() to ensure that it always uses a prefix of the index and uses no repeated columns. Enhanced comments. (check-in: b9fc89e4 user: drh tags: rowvalue)
19:54
Add an EXPLAIN QUERY PLAN line for when a index is used to implement an IN operator. (check-in: 171aa833 user: drh tags: rowvalue)
19:47
Add test cases to rowvalue2.test. (check-in: 078bb69e user: dan tags: rowvalue)
19:31
Enhance sqlite3FindInIndex() so that it is able to make use of the primary keys at the end of an index. (check-in: 4b589fbf user: drh tags: rowvalue)
18:17
Fix a post-OOM crash in updateRangeAffinityStr(). Add several ALWAYS() macros on unreachable branches. (check-in: 87d40195 user: drh tags: rowvalue)
17:54
Fix a problem with affinity changes and vector range comparisons. (check-in: b34413ac user: dan tags: rowvalue)
13:19
Allow ROWID values in indexed vector comparisons. (check-in: b0cc6be4 user: drh tags: rowvalue)
12:00
Fix a problem in internal function sqlite3OpenTableAndIndices causing an operand of an unrelated VM instruction to be overwritten. Fix for [ef360601]. (check-in: 70901479 user: dan tags: trunk)
04:32
Rename a test procedure in 'speed3.test' to avoid a name collision. (check-in: 84de17bc user: mistachkin tags: trunk)
03:42
Comment improvements. Put ALWAYS and NEVER macros on three unreachable branches. (check-in: 39761700 user: drh tags: rowvalue)
02:10
Adapt the special case of '$' in item names for test 'vtabH-3.1' as well. (check-in: e4aeaa2b user: mistachkin tags: trunk)
01:47
Make the test 'vtabH-3.1' work when there are less than 5 top-level items. (check-in: 328f79bb user: mistachkin tags: trunk)
01:17
Support running the fstree tests in 'vtabH.test' on Windows when not using the system drive. (check-in: 3680f95f user: mistachkin tags: trunk)
01:02
Remove an unreachable branch from sqlite3ExprAffinity() (check-in: 9d96f614 user: drh tags: rowvalue)
2016-08-25
23:24
Fix typos in 'vtabF.test' file. No changes to code. (check-in: a8546fee user: mistachkin tags: trunk)
22:31
Merge recent changes from trunk. (check-in: 5789aab8 user: drh tags: rowvalue)
22:06
Fix the sqlite3_trace_v2() interface so that it goes disabled if either the callback or mask arguments are zero, in accordance with the documentation. (check-in: 37e6c54b user: drh tags: trunk)
21:14
Refactor the sqlite3ExprCodeIN() routine for improved maintainability. (check-in: b56705ae user: drh tags: rowvalue)
17:47
Another fix in the IN-operator algorithm description. (check-in: f474aeac user: drh tags: rowvalue)
17:40
Further refinement of the in-operator.md documentation. (check-in: df064837 user: drh tags: rowvalue)
15:46
Improvements to IN operator code generator comments. Avoid unnecessary Copy operations on the LHS of the IN operator. (check-in: b6344298 user: drh tags: rowvalue)
14:23
Corrections to the IN-operator notes. (check-in: 25033ee9 user: drh tags: rowvalue)
14:00
Add notes on the implementation of the IN operator. (check-in: d256b2ca user: drh tags: rowvalue)
2016-08-24
21:54
Simplified VDBE code for the vector NOT IN null-scanning loop. (check-in: 7ae504e6 user: drh tags: rowvalue)
21:24
Improved extended comments of comparison operators when the SQLITE_STOREP2 flags is set on P5. No changes to non-debug code. (check-in: bbc1b016 user: drh tags: rowvalue)
19:58
Modularize all use of the calling convention macros based on the USE_STDCALL nmake macro. (check-in: e4a56197 user: mistachkin tags: callbackConv)
19:14
Add the sqlite3changebatch_db() API. (check-in: bee44ebc user: dan tags: changebatch)
18:51
In sqlite3FindInIndex(), improve internal comments and avoid an unreachable branch. (check-in: 55945fc1 user: drh tags: rowvalue)
18:50
Fix a bug in the fts5 snippet function causing it to return text containing zero phrase instances. (Closed-Leaf check-in: b174ed2b user: dan tags: fts5-snippet-bias)
18:49
Merge updates from trunk. (check-in: 428fc38d user: mistachkin tags: callbackConv)
18:21
Sync up the MSVC autoconf makefile. (check-in: 77f7c31a user: mistachkin tags: trunk)
17:49
Fix more unreachable branches. (check-in: 6099c180 user: drh tags: rowvalue)
15:37
Add a NEVER() on an unreachable branch in comparisonAffinity(). (check-in: 505a2f20 user: drh tags: rowvalue)
12:22
Fix a buffer overrun in the code for handling IN(...) operators when the LHS of the operator contains indexed columns or expressions. (check-in: f41a0391 user: dan tags: rowvalue)
00:51
The previous OOM fix was bad. Back it out and replace it with a better one. (check-in: 1e3bc369 user: drh tags: rowvalue)
00:25
Avoid a potential null-pointer dereference following an OOM. (check-in: 25f6ed8d user: drh tags: rowvalue)
2016-08-23
19:02
Simplify the row value misuse error message. (check-in: 838c50a5 user: drh tags: rowvalue)
19:02
Add further tests for changebatch. And a fix to prevent a changeset from conflicting with itself. (check-in: 506d6ff4 user: dan tags: changebatch)
18:30
Fix a problem with rowvalue UPDATE when the rowvalue is not the left-most and the RHS is a multi-column subquery. (check-in: e149e6b9 user: drh tags: rowvalue)
18:09
Fix harmless compiler warnings in changebatch. (check-in: a721a738 user: drh tags: changebatch)
17:30
Fix an uninitialized variable in CASE expression code generation. (check-in: c8ffae05 user: drh tags: rowvalue)
17:02
Add a new method to sessions - sqlite3sessions_fullchangeset() - to return a changeset that always contains values for all old.* fields. Update changebatch to use these values to more reliably detect multi-column UNIQUE constraint violations. (check-in: efa761b2 user: dan tags: changebatch)
14:42
Fix the SQLITE_USER_AUTHENTICATION login check so that the SQLITE_AUTH_USER error code correctly propagates and is not overwritten by SQLITE_ERROR. (check-in: c640489d user: drh tags: trunk)
2016-08-22
21:01
Add a couple of extra tests to changebatch1.test. (check-in: 207d970b user: dan tags: changebatch)
20:49
Add an experimental module to detect conflicts between sessions changesets. (check-in: 0c9fd6b7 user: dan tags: changebatch)
20:10
Add support for the SQLITE_DBCONFIG_MAINDBNAME configuration option. (check-in: 78395193 user: drh tags: trunk)
14:30
Fix the vector BETWEEN operator so that it only evaluates the left-most vector expression once. Add support for vector comparisons in the CASE operator. (check-in: 07e69f43 user: drh tags: rowvalue)
00:48
Reinstate the mechanism in BETWEEN that avoids evaluating the first expression more than once, but fix the affinity extractor so that it works with this mechanism. The de-duplication of the first expression still does not work for vector expressions, though. (check-in: 2f39987f user: drh tags: rowvalue)
2016-08-20
22:49
The docs promise the in "x BETWEEN y AND z" the x expression is only evaluated once. That is no longer true, and so some tests are failing. This needs to be fixed before merging to trunk. (check-in: e50d264f user: drh tags: rowvalue)
21:11
Further comment enhancements. No changes to code. (check-in: d4562a9e user: drh tags: rowvalue)
21:02
Clarification of code comments in expr.c. Clean up the implementations of sqlite3ExprIsVector() and sqlite3ExprVectorSize() slightly. (check-in: 4fb66d65 user: drh tags: rowvalue)
18:06
Fixes for problems following OOM errors. (check-in: 9041ee4a user: drh tags: rowvalue)
17:23
Fix other minor problems with the fts5 snippet() function. (check-in: b56b2a85 user: dan tags: fts5-snippet-bias)
17:00
Do not duplicate the Expr.pLeft subtree of a TK_SELECT_COLUMN node. (check-in: 8384c77e user: drh tags: rowvalue)
15:01
Fix a segfault that could occur if a query that used a vector comparison contained certain types of syntax errors. (check-in: 203f07c5 user: dan tags: rowvalue)
12:00
Improvements to the vector comparison splitter in exprAnalyze(). (check-in: a3ffd283 user: drh tags: rowvalue)
01:06
Improvements to comments. No code changes. (check-in: 4165d20f user: drh tags: rowvalue)
00:51
Add support for vector assignments in the SET clause of an UPDATE statement. (check-in: f320d47d user: drh tags: rowvalue)
00:07
Change the way TK_SELECT_COLUMN is handled so that the subquery is only generated once even if part of the vector comparison is used for indexing and the other part is now. This change also is a pathway to vector assignment in UPDATE statements. (check-in: d8feea7d user: drh tags: rowvalue)
2016-08-19
19:58
Replace the magic number (-2) with its symbol XN_EXPR in the exprMightBeIndexed() routine. No logic changes. (check-in: d4a5af69 user: drh tags: rowvalue)
19:12
Improved comments on sqlite3CodeSubquery(). No changes to code. (check-in: acea4ee1 user: drh tags: rowvalue)
18:40
Show the WhereTerm.iField value on debugging output, when it is non-zero. (check-in: 931c9535 user: drh tags: rowvalue)
18:37
Register any built-in fts5 module before loading automatic extensions. This allows automatic extensions to register fts5 tokenizers and auxiliary functions. (check-in: b10e31dc user: dan tags: fts5-snippet-bias)
18:28
Fix an obsolete comment on codeEqualityTerm(). No changes to code. (check-in: d07be5af user: drh tags: rowvalue)
15:41
Merge recent enhancements from trunk. (check-in: b1787236 user: drh tags: rowvalue)
15:17
Merge the VACUUM attached database enhancement from trunk. (Closed-Leaf check-in: 12d5e38d user: drh tags: dbconfig_maindbname)
15:15
Enhance the VACUUM command so that it can operate on an attached database. (check-in: 083f9e62 user: drh tags: trunk)
15:12
Disable row counting on queries run while doing a VACUUM. (Closed-Leaf check-in: ad35ef11 user: drh tags: vacuum-attached-db)
14:25
Have the fts5 snippet() function avoid favouring snippets that begin with sentences that do not contain search terms. Add an extra bias in favour of the first sentence in the document. (check-in: d8180af2 user: dan tags: fts5-snippet-bias)
14:20
Add the capability to VACUUM an attached database by specifying the schema name as an argument to the VACUUM command. Since version 2.0, VACUUM has accepted an argument which was silently ignored. Now it has meaning. (check-in: 29d63059 user: drh tags: vacuum-attached-db)
2016-08-18
22:44
Add the SQLITE_DBCONFIG_MAINDBNAME interface. (check-in: d6e3d579 user: drh tags: dbconfig_maindbname)
22:19
Change the name of Db.zName to Db.zDbSName for improved long-term code maintainability. (check-in: cb9865e1 user: drh tags: trunk)
20:56
Fix a zName to zDbSName conversion missed in the previous check-in. (Closed-Leaf check-in: d7cf423c user: drh tags: zDbSName)
19:04
Fix a SQL NULL handling bug in the vector IN operator code generation. (check-in: 936146b1 user: drh tags: rowvalue)
18:09
Display SELECT_COLUMN expressions in the .wheretrace debugging output. (check-in: 3b27a5da user: drh tags: rowvalue)
17:55
Display VECTOR expressions in .wheretrace and .selecttrace debugging output. (check-in: 157347e2 user: drh tags: rowvalue)
15:36
Since the TK_VECTOR token is not actually used by the parser, make it an extra token code added by the tool/addopcodes.tcl script during parser build. (check-in: b40b202c user: drh tags: rowvalue)
15:21
Merge recent enhancements from trunk. (check-in: 4768a106 user: drh tags: rowvalue)
15:15
Simplify the vector comparison code generator logic, and the resulting VDBE code. (check-in: e2ad0b5d user: drh tags: rowvalue)
14:47
Adjust some tests to account for recent changes to the fts5 snippet function. (check-in: 184ecbe9 user: dan tags: fts5-snippet-bias)
14:33
Rename the Db.zName field to Db.zDbSName to make it more descriptive and to distinguish it from all of the other "zName" variables scattered throughout the code. (check-in: 92a22f01 user: drh tags: zDbSName)
2016-08-17
19:05
Bias the fts5 snippet() function to return snippets that look like they start at the start of sentences. (check-in: 60de1594 user: dan tags: fts5-snippet-bias)
11:14
Fix a problem in the fts5 snippet() auxiliary function. (check-in: a861713c user: dan tags: trunk)
2016-08-16
16:46
Fix a bug in destructor processing of Lemon. That has no impact on the SQLite grammar. The bug was introduced by prior work to optimize the Lemon-generated parser used by SQLite. (check-in: f9035b8e user: drh tags: trunk)
2016-08-13
14:30
Questionable changes to support interruptible checkpoint in ZipVFS. (Leaf check-in: c7a9f26d user: dan tags: interruptible-checkpoint)
14:17
Remove an unnecessary stack variable from sqlite3VdbeExec(). (Closed-Leaf check-in: c54bd9c8 user: drh tags: vector-compare)
13:03
Improvements to commits. No code changes. (check-in: 18f5a3be user: drh tags: vector-compare)
12:37
Fix to the vector less-than operator. All legacy tests passing now. (check-in: ec70a67e user: drh tags: vector-compare)
10:34
Fix an fts5 problem with corrupt database handling found by address-sanitizer. (check-in: e22252e1 user: dan tags: trunk)
10:02
Attempt to simplify the logic and generated code for vector comparisons. Basic comparison operators are working, but there are many indexing test failures still to be worked through. (check-in: dfc028cf user: drh tags: vector-compare)
06:38
Fix a buffer overread in fts5. (check-in: fcfbee6c user: dan tags: trunk)
2016-08-12
16:21
Have wal file checkpoints exit early if the sqlite3_interrupt() API function is called. (check-in: 8a5f41c7 user: dan tags: interruptible-checkpoint)
11:25
Add VdbeCoverage() macros on newly added VDBE branch operations. (check-in: 381aa731 user: drh tags: rowvalue)
11:01
Fix a post-OOM memory leak. (check-in: 14009b32 user: drh tags: rowvalue)
2016-08-11
19:12
Merge version 3.14 plus some subsequent patches (including the page-cache performance patch) from trunk. (check-in: d9f8918c user: drh tags: begin-concurrent)
18:05
Add the "modeof=<filename>" URI parameter to os_unix.c - used to specify a file to copy permissions from when a new database is created. Also allow passing NULL as the second parameter to sqlite3rbu_vacuum(). (check-in: ed406d31 user: dan tags: trunk)
12:31
Merge recent changes from trunk. (check-in: 959677b9 user: drh tags: rowvalue)
09:55
Fix a 1 byte buffer overwrite in the "sqldiff --rbu" command. (check-in: ab83d707 user: dan tags: trunk)
01:24
Fix a harmless compiler warning in FTS5. (check-in: 2ea0c8b4 user: drh tags: trunk)
2016-08-10
20:02
Optimization to sqlite3WhereExprUsage(). (check-in: 95493058 user: drh tags: trunk)
19:43
List of MATCH operator names in isMatchOfColumn() should be constant. (check-in: 7319d676 user: drh tags: trunk)
18:56
Performance enhancement in the expression walker. (check-in: ebccbd96 user: drh tags: trunk)
15:02
Tweaks to pcache1TruncateUnsafe() to make it slightly smaller and faster and easier to test. (check-in: 059f4e2e user: drh tags: trunk)
14:40
Rearrange the code inside sqlite3RunParser() routine so that sqlite3Parser() is only called from a single place. This allows sqlite3Parser() to be in-lined, which results in a 0.25% overall performance gain. (check-in: 25d1d02b user: drh tags: trunk)
13:34
Increment the version number to 3.15. (check-in: 8c2701e7 user: drh tags: trunk)
13:30
Improved comments on the lempar.c parser template. Adjust the YY_SHIFT_USE_DFLT constant in the Lemon-generated parser tables so as to guarantee that it is always out of range of the yy_lookahead[] table. (check-in: 83622f3f user: drh tags: trunk)
11:50
Fix pcache1TruncateUnsafe() run faster for the case where iLimit is very close to iMaxKey. (check-in: b07a26df user: drh tags: trunk)
01:43
Performance optimization in the yy_find_shift_action() routine of the Lemon-generated parser. (check-in: ba6663be user: drh tags: trunk)
2016-08-09
21:08
Prototype for the remember(V,PTR) extension function. (Closed-Leaf check-in: f0942c36 user: drh tags: rememberFunc)
21:01
Fix harmless compiler warning. (check-in: 9a5a4f6e user: drh tags: trunk)
19:48
Minor update to the way fts5 column filters are parsed. (check-in: 14864f2b user: dan tags: trunk)