SQLite

Timeline
Login

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

200 most recent check-ins using file test/lookaside.test version 90052e87

2022-09-27
17:00
Fix a potential division by zero error in this version of RBU that might occur if an external client checkpoints a wal file created by a suspended RBU process. (Leaf check-in: 73edd93a user: dan tags: branch-3.14)
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)
19:29
Fix a buffer overread that could occur when running fts5 prefix queries inside a transaction. (Leaf check-in: b584fd36 user: dan tags: branch-3.9)
18:43
Fix a buffer overread that could occur when running fts5 prefix queries inside a transaction. (Leaf check-in: 0770363b user: drh tags: branch-3.19)
17:55
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: 34cd2d92 user: drh tags: branch-3.19)
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)
17:39
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: 7b4583f9 user: drh tags: branch-3.9)
2018-12-19
17:05
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. Cherrypick from commit [8201f4e1] on branch-3.18. (check-in: 553a923c user: dan tags: branch-3.19)
16:03
Add 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. Cherrypick from commit [8201f4e1] on branch-3.18. (check-in: 4cb67252 user: dan tags: branch-3.9)
14:44
Fix a crash that can follow an OOM in fts3 on this branch. (check-in: de078148 user: dan tags: branch-3.19)
01:57
Add extra defenses against strategically corrupt databases to fts3/4. (check-in: 882ef4e3 user: drh tags: branch-3.9)
01:44
Add extra defenses against strategically corrupt databases to fts3/4. (check-in: 4f315e4a user: drh tags: branch-3.19)
01:38
Add extra defenses against strategically corrupt databases to fts3/4. (check-in: 4bb21d82 user: drh tags: branch-3.18)
2018-03-16
07:49
Fix a problem in test script thread001.test causing a spurious "-1 files were left open" error when run separately. Cherrypick of [1774f1c3b]. (check-in: 6cf8172d user: dan tags: branch-3.19)
2018-03-06
11:49
Avoid running a couple of tests in crash8.test that depend on the presence of the journal file if running on an F2FS file-system that does not require a journal file. Cherrypick of [797e02e0]. (check-in: eeedfc74 user: dan tags: branch-3.19)
2018-02-02
16:46
Enable the ".wheretrace" and ".selecttrace" options in the shell when configure is run using --enable-debug. This mirrors a similar change that has already occurred on trunk. (Leaf check-in: c4f798b2 user: drh tags: branch-3.9.2)
08:14
In SQLITE_ENABLE_BATCH_ATOMIC_WRITE builds on F2FS file-systems, invoke SQLITE_FCNTL_ROLLBACK_ATOMIC_WRITE if an SQLITE_FCNTL_COMMIT_ATOMIC_WRITE call fails. Also, do not use an atomic transaction to create the initial database. This is because if an error occurs while writing to the db file, any changes to the file-size do not seem to be rolled back automatically. The only time this matters is when the file was 0 bytes in size to start with. Cherrypick of [b3122db154]. (check-in: 22a228ed user: dan tags: branch-3.19)
08:08
Update a couple of test scripts so that they work on F2FS file-systems that support atomic transactions. Cherrypick of [49e58e645e]. (check-in: 1b394679 user: dan tags: branch-3.19)
2018-01-13
14:07
Fix a typo in test file crash8.test. (check-in: 4951d91d user: dan tags: branch-3.19)
13:21
Fix various problems in test scripts preventing "make test" from passing on F2FS file-systems with the "atomic-write" feature. Cherrypick of [56d93d07]. (check-in: 6a296d4d user: dan tags: branch-3.19)
2017-11-01
19:30
Fix the SQLITE_NoopUpdate #define so that it occurs under the correct conditions. (check-in: bdf791f9 user: drh tags: begin-concurrent-pnu)
18:48
The "PRAGMA noop_update" command now requires SQLITE_ENABLE_NOOP_UPDATE and no longer requires SQLITE_DEBUG. (check-in: 81baf67c user: drh tags: begin-concurrent-pnu)
2017-10-11
20:10
Change some internal details to bring this branch closer to the code on trunk. (check-in: 58e42cfd user: dan tags: schemalint)
2017-09-12
20:09
Add the highly-experimental "PRAGMA noop_update=TRUE" command. (check-in: afe45271 user: drh tags: begin-concurrent-pnu)
2017-08-29
17:52
Add a Tcl script to run the performance tests described in README-server-edition.html. (check-in: 1b3df8ff user: dan tags: server-process-edition)
2017-08-28
14:33
Backport changes to shell.c into shell.c.in. (check-in: 2348045f user: drh tags: trunk)
11:12
Fix the ".dump", ".schema", and ".fullschema" commands of the command-line shell so that they work even if PRAGMA empty_result_callbacks is enabled. Fix for ticket [02f0f4c54f281]. (check-in: cf0d3715 user: drh tags: trunk)
2017-08-27
10:55
Instrumentation and changes in an effort to reduce thread contention. (Leaf check-in: c93298d5 user: drh tags: reduced-thread-contention)
2017-08-25
19:51
Fix a few over-length source code lines. No functional changes. (check-in: 1783c54c user: drh tags: trunk)
15:43
Size and performance optimization for sqlite3Init(). (check-in: 776d9128 user: drh tags: trunk)
13:34
Small performance optimization in sqlite3WhereBegin(). (check-in: 39f708d1 user: drh tags: trunk)
13:02
Convert a branch made unreachable by [59560d07] into an assert(). (check-in: 2738b8db user: drh tags: trunk)
11:44
Improve the internal mechanism used to keep track of what kind of syncing to do for WAL transaction commits and checkpoint operations. Use the checkpoint-style of syncing to sync the header of a new or restarted WAL file. (check-in: e8d23afe user: drh tags: trunk)
09:17
Avoid returning duplicate rows in experimental pragmas "pragma_list", "module_list" and "function_list". (check-in: b79cc8dc user: dan tags: trunk)
01:14
Update the mechanism used to keep track of what kind of syncing to do for WAL transaction commits and checkpoint operations. Use the checkpoint-style of syncing to sync the header of a new or restarted WAL file. (Closed-Leaf check-in: bf65dae8 user: drh tags: wal-sync-refactor)
2017-08-24
20:54
Fix an incorrect hyperlink in a comment. (check-in: 25292b9a user: drh tags: trunk)
16:21
Version 3.20.1 (Leaf check-in: 8d3a7ea6 user: drh tags: release, branch-3.20, version-3.20.1)
15:43
Fixes to documentation about SQLITE_OPEN_URI. (check-in: 7ec72314 user: drh tags: trunk)
14:08
Increase the version number to 3.20.1. (check-in: 70338c7c user: drh tags: branch-3.20)
14:04
Make sure the sqlite3_result_pointer() interface does not leave a VM register in an inconsistent state. Fix for ticket [7486aa54b968e9b5]. Test cases are in TH3. (check-in: abc7adfd user: drh tags: branch-3.20)
13:55
Make sure the sqlite3_result_pointer() interface does not leave a VM register in an inconsistent state. Fix for ticket [7486aa54b968e9b5]. Test cases are in TH3. (check-in: d2f9230c user: drh tags: trunk)
10:10
Test BEGIN CONCURRENT transactions that consist entirely of read-only statements. (check-in: c3fe1f4b user: dan tags: begin-concurrent)
2017-08-23
23:09
Performance optimization to pageFindSlot() in the b-tree layer. (check-in: 59560d07 user: drh tags: trunk)
20:36
Size and performance optimization to dropCell() and freeSpace(). (check-in: bc1ec123 user: drh tags: trunk)
14:45
Smaller and faster implementation of the fillInCell() routine. (check-in: 77074990 user: drh tags: trunk)
2017-08-22
21:23
Update documentation to make it clear that SQLITE_SOURCE_ID and sqlite3_sourceid() might changes if the source code is edited. (check-in: e69c0c87 user: drh tags: trunk)
21:07
Less dramatic changes to the source-id following an edit. Modify the way that the amalgamation is constructed to give it the opportunity to detect changes and modify the source-id. (check-in: 564c7340 user: drh tags: trunk)
19:54
Modify the SQLITE_SOURCE_ID if the source code has changed in any way since the previous check-in. (check-in: 515d6a83 user: drh tags: trunk)
19:49
Trying again to get out-of-tree builds to work correctly. (Closed-Leaf check-in: a1b3337e user: drh tags: mksourceid)
19:44
mksourceid still does not work right. (Leaf check-in: ca5dcb4a user: drh tags: mistake)
19:43
Attempting to fix the source-id generator so that it works for out-of-tree builds. (check-in: 5a037ac2 user: drh tags: mksourceid)
19:19
Add the "mksourceid" program to the build process. That program changes the SQLITE_SOURCE_ID if the source tree has been modified in any way. (check-in: d4c05e04 user: drh tags: mksourceid)
15:21
Fix error tests in seldom-used compile-time branches of the unix backend. (check-in: 885c2b44 user: drh tags: trunk)
2017-08-21
02:20
Remove an unnecessary conditional. (check-in: 56d19f9f user: drh tags: trunk)
02:05
Minor optimization to sqlite3VdbeMemSetStr(). (check-in: 6538ef7b user: drh tags: trunk)
2017-08-19
20:38
Space and size optimization to the printf implementation. (check-in: d01d2cff user: drh tags: trunk)
15:50
Ensure that write-locks on pages are dropped at the end of each write transaction, even if there is still a read transaction open. (check-in: 2dd36ade user: dan tags: server-process-edition)
2017-08-18
22:30
Another size and performance optimization to readDbPage(). This time we eliminate some unnecessary local variables. (check-in: 745bc8de user: drh tags: trunk)
21:14
Size and performance optimization the readDbPage() routine in the pager. (check-in: ca9e1875 user: drh tags: trunk)
19:28
Version 3.19.4 (check-in: 605907e7 user: drh tags: release, version-3.19.4, branch-3.19)
18:55
Add tests to this branch. (check-in: abb6e076 user: dan tags: server-process-edition)
16:09
For the unix VFS, avoid an unnecessary stat() system call prior to opening any file in the common case where there are no unused file descriptors. (check-in: 3075cfa0 user: drh tags: trunk)
16:04
Do not search for locks to clear when connecting to a db in multi-process mode unless it looks like the previous user of the client-id crashed. (check-in: 66fb9e1c user: dan tags: server-process-edition)
14:34
Combine the OP_CreateTable and OP_CreateIndex opcodes of the bytecode engine into a single OP_CreateBtree opcode. This simplifies the implementation and makes the bytecode programs clearer. (check-in: eb1202b5 user: drh tags: trunk)
08:29
Only use indexes on expressions to optimize ORDER BY and GROUP BY if the collation sequence matches. Possible fix for [e20dd54a]. (check-in: 37e19008 user: dan tags: trunk)
2017-08-17
20:53
Use the __builtin_clzll() function of gcc to improve the performance and reduce the size of the sqlite3LogEst() routine. (check-in: a42a438c user: drh tags: trunk)
19:32
Add support for crash recovery in multi-process mode. And add test cases for the same. (check-in: a8115f95 user: dan tags: server-process-edition)
19:23
Small optimization in the Expr tree walker. (check-in: 264a5e1b user: drh tags: trunk)
18:54
Size optimization in the authorizer error message generation logic. (check-in: 0367a4d5 user: drh tags: trunk)
18:23
In sqlite3ViewGetColumnNames(), return the number of errors, not an error code. (check-in: f13682ea user: drh tags: trunk)
18:17
The RTREE extension should return SQLITE_CORRUPT_VTAB, not just SQLITE_CORRUPT when it encounters incorrectly formatted shadow tables. (check-in: 0712f057 user: drh tags: trunk)
14:47
Remove an unnecessary branch from the [be436a7f4587ce517ddc] fix. (check-in: fb6ca99b user: drh tags: trunk)
14:12
Add test cases for ticket [be436a7f4587ce517] using virtual table modules fts5 and rtree. (check-in: 2101b420 user: dan tags: trunk)
02:26
Defer schema resets when the query planner is running. Proposed fix for ticket [be436a7f4587ce517]. (check-in: a7bc7752 user: drh tags: trunk)
2017-08-16
19:20
Avoid a test for CURTYPE_BTREE in sqlite3VdbeCursorMoveto() in order to reduce the size and improve the performance of OP_Column. (check-in: f078deb2 user: drh tags: trunk)
17:06
Update this branch with the latest changes from trunk. (check-in: 380a7b7a user: dan tags: server-process-edition)
16:52
Enhance this branch to support page-level-locking (without MVCC) for multi-process deployments. (check-in: 04e0cb57 user: dan tags: server-process-edition)
14:16
Remove an unnecessary local variable from OP_Column, for a small size reduction and performance increase. (check-in: 39543903 user: drh tags: trunk)
11:04
Performance improvement in the OP_Column opcode. (check-in: dc98a92f user: drh tags: trunk)
2017-08-15
14:14
Small size and performance improvement in the OP_Column opcode. (check-in: 2cf3f3de user: drh tags: trunk)
03:46
Btree optimization: New field BtCursor.pPage that points to the current page, saving a single pointer dereference on each access. (check-in: 373b71d1 user: drh tags: trunk)
01:07
Fix a problem allowing a conflicting transaction to be committed in the case where more than one 32KB shared-memory page has been written to since the transaction was started. (check-in: 346a710d user: drh tags: begin-concurrent-branch-3.19)
2017-08-14
23:53
Minor size and performance optimization to sqlite3BtreeCloseCursor(). (check-in: 16969338 user: drh tags: trunk)
18:13
Change the internal btree routine moveToRoot() to return SQLITE_EMPTY if the table is empty or if it has pgnoRoot==0. This simplifies the callers, making them smaller and faster. The SQLITE_EMPTY result code is intercepted and changed into SQLITE_OK before surfacing in an API. (check-in: 240d5714 user: drh tags: trunk)
17:03
Fix harmless indentation error. (check-in: 25e92baa user: drh tags: trunk)
14:53
Sometimes a TK_COLUMN Expr node can have Expr.pTab==0 if it is a reference to an expression column in an index on an expression. Fix for ticket [aa98619ad08ddcab]. (check-in: d0da791b user: drh tags: trunk)
07:16
Remove code for PLL support in wal mode from this branch. (check-in: 8e1b28ed user: dan tags: server-process-edition)
06:55
Update this branch with latest trunk changes. (check-in: bc2498d6 user: dan tags: server-process-edition)
01:33
Properly dequote column names in tables constructed by an aggregate SELECT. (check-in: 7e0d3e9c user: drh tags: trunk)
2017-08-12
18:31
Add new test file fts5vocab2.test. (check-in: 02174842 user: dan tags: trunk)
14:06
Add a more rigorous test case for the bug fixed by the previous commit on this branch. (check-in: 42560723 user: dan tags: begin-concurrent)
02:16
Update the speed-check.sh test script to append log output to the end of the cout-NAME.txt file. (check-in: 14d262d6 user: drh tags: trunk)
02:01
Remove the zBase field from the StrAccum object. Resulting code is slightly smaller and faster. (check-in: 6e52fa5f user: drh tags: trunk)
01:09
Size and performance micro-optimization on sqlite3SrcListIndexedBy(). (check-in: 28a5aec1 user: drh tags: trunk)
2017-08-11
21:16
Fix a problem allowing a conflicting transaction to be committed in the case where more than one 32KB shared-memory page has been written to since the transaction was started. (check-in: 38dd9b50 user: dan tags: begin-concurrent)
20:22
Add another PAGERTRACE() macro to show when pages are added to Pager.pAllRead. No impact on production builds. (check-in: 11054cf5 user: drh tags: begin-concurrent)
19:31
Add a PAGERTRACE() macro to log when frames are written into the WAL file. This is for analysis only and is a no-op for production builds. (check-in: 39f39e3d user: drh tags: begin-concurrent)
19:16
Merge the latest enhancements from trunk. (check-in: 47e71695 user: drh tags: begin-concurrent)
18:59
Fix compiler warnings that arise if the PAGERTRACE macro is turned on. This changes does not affect production builds. (check-in: 831156a4 user: drh tags: trunk)
18:55
Improved comments in the header of the LSM1 vtab module. No code changes. (check-in: aa49926d user: drh tags: trunk)
13:51
New test cases for LSM1. (check-in: cb0c49cb user: drh tags: trunk)
12:49
Convert the LSM1 virtual table to be WITHOUT ROWID and get UPDATE and DELETE operations working on it. (check-in: 2164031b user: drh tags: trunk)
03:47
Simplification to the like optimization logic. Remove unnecessary branches. (check-in: 9466d952 user: drh tags: trunk)
2017-08-10
20:43
Allow WITHOUT ROWID virtual tables to be writable as long as they have exactly one column as their PRIMARY KEY. (check-in: c601d128 user: drh tags: trunk)
20:36
Add the "instance" type to the fts5vocab virtual table module. For direct access to the contents of the fts5 term index. (check-in: 34a7bd71 user: dan tags: trunk)
19:12
Fix a typo in csv01.test introduced by the previous commit. (Closed-Leaf check-in: f38ff718 user: dan tags: writable-vtab-without-rowid)
19:10
Add extra tests to this branch. (check-in: b0e3b88a user: dan tags: writable-vtab-without-rowid)
17:53
Test case for writing to a WITHOUT ROWID virtual table. The TCLVAR virtual table is modified to add a "fullname" column which is the primary key, and to accept update operations against the primary key. (check-in: 6997e00c user: drh tags: writable-vtab-without-rowid)
15:19
Experimental changes that allow a WITHOUT ROWID virtual table to be writable as long as it has only a single-column PRIMARY KEY. (check-in: ab9ee4c1 user: drh tags: writable-vtab-without-rowid)
03:27
Add a first draft of the "vtablog" extensions that implements a generic virtual table useful for experimentation using the command-line shell. (check-in: e49279e6 user: drh tags: trunk)
2017-08-09
22:55
Preserve the error code from xConnect or xCreate methods in virtual table implementations when they are encountered during parsing. (check-in: dcdf0913 user: drh tags: trunk)
20:35
Add experimental sqlite3_open_v2() flag SQLITE_OPEN_REUSE_SCHEMA. For sharing identical in-memory schema objects between connections. (check-in: a6256980 user: dan tags: reuse-schema)
19:27
Various bug fixes for the new LSM1 virtual table design. (check-in: 94434a25 user: drh tags: trunk)
18:40
Work toward redesigning the interface to the LSM1 virtual table. (check-in: 313df946 user: drh tags: trunk)
2017-08-08
21:40
The identifier "vsnprintf" appears to be a reserved word in recent versions of XCode, so avoid using it as a structure field. (check-in: 795eede3 user: drh tags: trunk)
21:30
Fix a faulty signed/unsigned character comparison in the LIKE optimization logic. (check-in: f4a4b149 user: drh tags: trunk)
20:03
Enhance the CSV virtual table extension so that it accepts the last row of the CSV file even if the last row omits the closing \n, as long as the last row has a full set of columns. (check-in: 537e3be2 user: drh tags: trunk)
2017-08-07
19:12
Avoid casting a value larger than 2^31 to a (size_t) or (SIZE_T) on systems where it is a 32-bit type. (check-in: b26d7a1c user: dan tags: trunk)
19:06
On Windows, avoid casting a value larger than 2^31 to a (SIZE_T) on systems where it is a 32-bit type. (Leaf check-in: f08d63b4 user: mistachkin tags: mmap-size-limit)
18:54
Fix a problem with handling SQLITE_FCNTL_MMAP_SIZE requests with a negative parameter in os_unix.c. (check-in: 4249fcf7 user: dan tags: mmap-size-limit)
18:27
Update bigmmap.test to account for builds that use "-DSQLITE_MAX_MMAP_SIZE=<integer-constant>LL". (check-in: 7c8b6f1c user: dan tags: mmap-size-limit)
18:13
Avoid casting a value larger than 2^31 to a (size_t) on systems where it is a 32-bit type. (check-in: 46c3085d user: dan tags: mmap-size-limit)
17:28
Fix an out-of-order test number. (check-in: 38f30091 user: mistachkin tags: trunk)
17:14
Add new test file "bigmmap.test". For testing builds with -DSQLITE_MAX_MMAP_SIZE > 2GB. (check-in: 17447062 user: dan tags: trunk)
14:15
Update this branch with latest trunk changes. (check-in: 17bc7ded user: dan tags: server-process-edition)
14:06
Fix a bug on this branch involving mmap mode and readonly transactions. (check-in: 355c594e user: dan tags: server-process-edition)
2017-08-05
16:15
Fix a segfault in swarmvtab that could occur if there was an error in the SQL statement passed to the constructor. Add other test cases. (check-in: 6ce8b7ca user: dan tags: trunk)
15:49
Microoptimization in pcache1.c makes the code slightly smaller and faster. (check-in: 422cd9f3 user: drh tags: trunk)
15:30
Fix memory leak in swarmvtab. (check-in: 002a9c7b user: dan tags: trunk)
2017-08-04
20:27
Add the swarm virtual table to the existing union virtual table module. (check-in: 0f6f6f03 user: drh tags: trunk)
20:15
Add the optional non-found-callback to the swarm-vtab. (Closed-Leaf check-in: a94e2f60 user: drh tags: union-vtab)
17:39
Add further test cases for swarmvtab. And minor code changes. (check-in: 0f82d3b9 user: dan tags: union-vtab)
16:16
Add test cases and associated fixes for swarmvtab. (check-in: 7ae20eac user: dan tags: union-vtab)
08:24
Add SQLITE_API qualifiers to public API functions in header file sqlite3rbu.h. (check-in: e11e6eb1 user: dan tags: trunk)
08:23
Update script tool/mksqlite3c.tcl to avoid duplicating the SQLITE_API qualifier when copying sqlite3rbu.h into the amalgamation. (Closed-Leaf check-in: 61ab112f user: dan tags: sqlite3rbu.h-fix)
2017-08-03
20:13
Modify swarmvtab to use a separate database connection for each database file. (check-in: 1f05ad29 user: dan tags: union-vtab)
15:50
Add SQLITE_API qualifiers to public API functions in header file sqlite3rbu.h. (check-in: d8637bad user: dan tags: sqlite3rbu.h-fix)
15:43
Fix some problems in RBU test cases. Also update RBU source code to better handle the trivial case where an RBU update is applied to a database zero pages in size. (check-in: 7676b39b user: dan tags: trunk)
00:29
Optimization to the comparison opcodes in the byte-code engine. (check-in: 654935c7 user: drh tags: trunk)
2017-08-02
22:43
In the KeyInfo object, refactor the nField and nXField elements into nKeyField and nAllField, which are more useful and run a little faster. (check-in: aea5990e user: drh tags: trunk)
19:59
Enhance the code in unionvtab.c to also provide the "swarmvtab" virtual table module. There are still several problems on this branch. (check-in: 03d94388 user: dan tags: union-vtab)
19:04
Avoid redundant calls to sqlite3ApiExit() in sqlite3_step(). (check-in: 527974d4 user: drh tags: trunk)
18:28
Only attempt to invoke WAL callbacks when a transaction has committed. (check-in: bcc6dacb user: drh tags: trunk)
18:27
Minor optimization on clearCell() (check-in: f3c39c29 user: drh tags: trunk)
15:50
Optimizations associated with error handling in btree cursors. (check-in: 49ba54e2 user: drh tags: trunk)
12:38
Remove unused token codes. (check-in: c45078c0 user: drh tags: trunk)
11:36
Rearrange some of the added token codes for a very slight size decrease and performance increase. (check-in: cb3742f5 user: drh tags: trunk)
11:04
Rearrange integer token values in the parser and logic in the resolveP2Values() routine for a small size reduction and performance increase. (check-in: 1cad2926 user: drh tags: trunk)
03:21
Show which opcodes are jumps in the comments when generating the opcodes.h header file. (check-in: eef643a3 user: drh tags: trunk)
03:21
Add the "%token" control to the lemon parser. Not currently used by SQLite. (check-in: a6e4c5ae user: drh tags: trunk)
02:46
Faster implementation of resolveP2Values(). (check-in: 82e46fe0 user: drh tags: trunk)
2017-08-01
20:59
Slightly smaller and faster by allocating Parser objects on the stack. (check-in: 436a89b9 user: drh tags: trunk)
19:53
Split the OP_Last opcode into OP_Last and OP_SeekEnd. Use OP_SeekEnd to position a cursor prior to appending. Ticket [cb91bf4290c211d]. (check-in: 3e02474c user: drh tags: trunk)
14:16
Take advantage of atomic-write capabilities in the F2FS filesystem when the database is stored on such a filesystem. This is a compile-time option activated using SQLITE_ENABLE_BATCH_ATOMIC_WRITE. (check-in: 24190b22 user: drh tags: trunk)
13:54
Merge recent enhancements from trunk. (check-in: aafe1fec user: drh tags: begin-concurrent)
13:24
Version 3.20.0 (check-in: 9501e22d user: drh tags: release, branch-3.20, version-3.20.0)
00:20
Allow ATTACH and DETACH inside of a transaction. (check-in: 95e8f316 user: drh tags: trunk)
2017-07-31
20:06
TclKit batch tool updates for MSVC. Cherrypick of [bcec155e0d6c6b17] and [0c77935cf9949099]. (check-in: 96ff1675 user: mistachkin tags: branch-3.20)
19:55
Add documentation file ./README-server-edition.html. (check-in: fbc4f4ad user: dan tags: server-process-edition)
19:39
Recognize 'AMD64' as 'x64' in the TclKit batch tool for MSVC. (check-in: 0c77935c user: mistachkin tags: trunk)
19:34
Sync up the MSVC autoconf makefile. (check-in: 44cebc6e user: drh tags: branch-3.20)
19:26
Sync up the MSVC autoconf makefile. (check-in: b3b89f4f user: mistachkin tags: trunk)
17:51
Fix a windows-compatibility issue in the recently added schema6.test script. (check-in: 14e6c302 user: drh tags: trunk)
17:40
More consistent column names. Cherry-pick of [09834279] and [0c38dde45] as a fix for ticket [de3403bf5ae]. (check-in: be0e24a0 user: drh tags: branch-3.20)
17:06
Prevent OSSFuzz from using debugging pragmas that are disabled in default builds and which generate lots of excess output. (check-in: cdc6494c user: drh tags: trunk)
16:42
Move the generation of output column names earlier, to right after name resolution and before query transformations such as flattening. This prevents the names from getting mangled by query transformations, and obviates hacks in the query flattener that attempt to work around the name mangling. The resulting code is smaller and faster and gives more consistent output. Fix to ticket [de3403bf5ae5f72ed]. (check-in: ade7ddf1 user: drh tags: trunk)
13:22
Use batch atomic writes when running on a recent F2FS filesystem, if compiled with SQLITE_ENABLE_BATCH_ATOMIC_WRITE. (check-in: 2dd0c77d user: drh tags: branch-3.19)
13:11
Move the generation of output column names earlier, to right after name resolution and before query transformations such as flattening. This prevents the names from getting mangled by query transformations, and obviates hacks in the query flattener that attempt to work around the name mangling. The resulting code is smaller and faster and gives more consistent output. This is a fix for ticket [de3403bf5ae5f72ed] and a cherry-pick of check-in [09834279aeca3bda] (check-in: 499942b3 user: drh tags: branch-3.19)
12:19
Correctly handle an "INTEGER PRIMARY KEY UNIQUE" column in a WITHOUT ROWID table. This is a fix for ticket [bc115541132dad136], cherry-picked from check-in [5216bfb73f1a49bdd8] (check-in: 322a2ede user: drh tags: branch-3.19)
12:07
Increase the version number to 3.19.4 (check-in: b77f297d user: drh tags: branch-3.19)
12:04
Merge further enhancements to the batch-atomic-write subsystem. (Leaf check-in: 4be4265d user: drh tags: batch-atomic-write-3.19)
2017-07-30
20:30
Correctly handle an "INTEGER PRIMARY KEY UNIQUE" column in a WITHOUT ROWID table. This is a fix for ticket [bc115541132dad136] cherry-picked from [5216bfb73f1a49bdd]. (check-in: 5c8cb1ff user: drh tags: branch-3.20)
19:50
Add the schema6.test module for demonstrating schemas that generate identical and different content. (check-in: ac1da06a user: drh tags: trunk)
18:40
Correctly handle an "INTEGER PRIMARY KEY UNIQUE" column in a WITHOUT ROWID table. This is a fix for ticket [bc115541132dad136], a problem discovered by OSSFuzz. (check-in: 5216bfb7 user: drh tags: trunk)
2017-07-29
17:10
Merge latest trunk changes with this branch. (check-in: b42c8779 user: dan tags: server-process-edition)
17:02
New test cases for column name generation interacting with the query flattener. (Closed-Leaf check-in: 0c38dde4 user: drh tags: early-column-names)
17:01
Update test program "tserver" to use a native pthreads mutex/condition variable to efficiently manage wal file checkpoints without the wal file growing indefinitely. (check-in: 8299bdb7 user: dan tags: server-process-edition)
16:01
Move the generation of output column names earlier, to right after name resolution and before query transformations such as flattening. This prevents the names from getting mangled by query transformations, and obviates hacks in the query flattener that attempt to work around the name mangling. The resulting code is smaller and faster and gives more consistent output. This is an alternative fix to ticket [de3403bf5ae5f72ed]. (check-in: 09834279 user: drh tags: early-column-names)
14:56
Use the subquery column name, not the original SQL statement text, as the added AS clause in the query flattener. (Closed-Leaf check-in: 5df7f0e6 user: drh tags: flattener-column-names)
03:33
In the query flattener, only add AS clauses to output columns of the outer query that are copied directly from the inner query. Formerly, all columns of the outer query received an AS clause if they did not have one already. This is a proposed fix for ticket [de3403bf5ae5f72]. (check-in: 439cc5c5 user: drh tags: flattener-column-names)
2017-07-28
22:22
Update Tcl version used by the TclKit batch tool for MSVC. (check-in: bcec155e user: mistachkin tags: trunk)
22:13
Fix harmless compiler warning. (check-in: 3286e1a0 user: mistachkin tags: trunk)
21:02
Reduce the number of mallocs required of writers in server mode. (check-in: 60953997 user: dan tags: server-process-edition)
20:49
Disable the writecrash.test module on Windows. (Leaf check-in: 4c0520d4 user: drh tags: batch-atomic-write)
18:16
More precise determination of when an in-memory journal needs to be spilled to disk. (check-in: 6a505bdd user: drh tags: batch-atomic-write)
13:18
Add new extended error codes for the atomic write file-controls. (check-in: 94a20047 user: drh tags: batch-atomic-write)
12:20
Fix a punctuation typo in the documentation for sqlite3_step(). No changes to code. (check-in: c965b757 user: drh tags: branch-3.20)
11:05
Add new test file test/atomic.test. (check-in: f9213e48 user: dan tags: batch-atomic-write)
02:02
Backport of all batch-atomic-write changes through check-in [67bad7fb9b] (check-in: def55027 user: drh tags: batch-atomic-write-3.19)
01:53
Add the SQLITE_ENABLE_BATCH_ATOMIC_WRITE macro to ctime.c (check-in: 67bad7fb user: drh tags: batch-atomic-write)
00:45
Release candidate 3 for 3.20.0 (check-in: d891a2a0 user: drh tags: branch-3.20)
2017-07-27
20:24
Enhance the like optimization so that it works with an ESCAPE clause. (check-in: f5d330f4 user: drh tags: trunk)
19:59
Increase the version number to 3.21.0 in anticipation for changes to go into the next release. (check-in: 0645f25c user: drh tags: trunk)
18:49
Merge the pointer-passing interface changes from the 3.20 branch. (check-in: c63903a4 user: drh tags: trunk)
18:43
Simplified documentation for the pointer passing interface. (check-in: 2dfcd9a8 user: drh tags: branch-3.20)
18:34
Do not set device-capabilities flags SEQUENTIAL or SAFE_APPEND for f2fs file-systems. (check-in: 4477e60c user: dan tags: batch-atomic-write)
17:45
Add destructor functions for the pointer argument (NB: not the type string) to the sqlite3_bind_pointer() and sqlite3_result_pointer() interfaces. Use this new capability to resolve a pointer leak in the custom geometry function mechanism of the RTREE extension. (check-in: 5a2340b8 user: drh tags: branch-3.20)
16:42
In sqlite3_bind_pointer(), invoke the destructor if the bind index is out of range, like sqlite3_bind_blob() does. (Closed-Leaf check-in: d6684d2a user: drh tags: pointer-with-destructor)
15:53
Improved implementation of the destructor on pointer-passing interfaces. (check-in: 601ad679 user: drh tags: pointer-with-destructor)
14:04
A demonstration of how FTS3 cursor pointers can be passed from the MATCH operator to the snippet() function securely and opaquely without having to use the sqlite3_result_pointer() and sqlite3_value_pointer() interfaces. (Closed-Leaf check-in: 49cf31da user: drh tags: pass-pointer-as-blob)