Many hyperlinks are disabled.
Use anonymous login
to enable hyperlinks.
50 check-ins occurring around 2677848087c9c090.
|
2010-10-26
| ||
| 11:56 | Add missing header comments for changes related to the experimental sqlite3_blob_reopen() API. (check-in: d1cc5c93f0 user: dan tags: experimental) | |
| 07:14 | More coverage tests for fts3.c. (check-in: 7a2f286400 user: dan tags: experimental) | |
|
2010-10-25
| ||
| 19:01 | Further improvements to coverage of fts3.c. Fixes for bugs revealed by the same. (check-in: 918b609290 user: dan tags: experimental) | |
| 12:47 | Test coverage improvements for fts3.c. (check-in: a8b1d99899 user: dan tags: experimental) | |
| 09:01 | Add test for matchinfo when a phrase includes some common tokens. (check-in: 80a54ebc41 user: dan tags: experimental) | |
|
2010-10-23
| ||
| 19:07 | Fixes for the matchinfo() function related to FTS4 common token handling. (check-in: deb80eac91 user: dan tags: experimental) | |
|
2010-10-22
| ||
| 19:03 | Add new test file fts3defer2.test. (check-in: 5a4d5bfcae user: dan tags: experimental) | |
| 16:44 | Add missing comments and fix compiler warnings in new FTS3/4 code. Other minor fixes too. (check-in: 1c9c70fec3 user: dan tags: experimental) | |
| 13:55 | Prevent an assert from failing when opening a zero-length database file with an apparently hot journal with locking_mode=exclusive set. (check-in: f000ac1e52 user: dan tags: trunk) | |
|
2010-10-21
| ||
| 22:58 | Make sure the estimated row count for ephemeral tables is initialized so that automatic indices can be used on those tables. (check-in: d30f7b2def user: drh tags: trunk) | |
| 15:49 | Merge trunk changes into experimental branch. (check-in: fd1e5cade0 user: dan tags: experimental) | |
| 15:12 | Fix some segfaults that could occur in obscure circumstances where error messages contained characters that could be mistaken for printf format specifiers. (check-in: f91471e723 user: dan tags: trunk) | |
| 12:34 | Fix a typo-bug that prevented --disable-amalgamation from working in Makefile.in. Also fix an overly long line in Makfile.in. (check-in: 2c3c4ba035 user: drh tags: trunk) | |
| 03:13 | Add new WHERETRACE macros for better diagnostics of the query planner. Added a new test case for the performance regression fixed by the previous check-in. (check-in: 1e0db99797 user: drh tags: trunk) | |
| 02:05 | Fix the query planner so that it uses the multi-index OR-clause solution if that is the lowest cost estimate. A prior bug cause the multi-index solution to be ignored in some circumstances. (check-in: 28ba625528 user: drh tags: trunk) | |
|
2010-10-20
| ||
| 18:56 | Updates to FTS4 to improve performance and make more accurate cost estimates for prefix terms. (check-in: d0a450ce78 user: dan tags: experimental) | |
|
2010-10-19
| ||
| 14:08 | Experimental changes to fts4 to try to selectively avoid loading very large doclists. (check-in: 5ae0ba447a user: dan tags: experimental) | |
|
2010-10-15
| ||
| 14:45 | Cherry-pick the change at [2d4505510032bf9] into the 3.6.1 branch. (check-in: ecb1419e4b user: drh tags: branch-3.6.1) | |
|
2010-10-14
| ||
| 01:17 | Avoid taking locks on unused database connections when committing a read transaction. (check-in: c0ee614fd9 user: drh tags: trunk) | |
| 01:16 | Check-in [d3c95e3a4e08d0] was incorrect. This is the correct fix for sqlit3_bind_blob(). Ticket [860399cc408f2dd5f4] (check-in: ea8c2f5f8a user: drh tags: trunk) | |
|
2010-10-12
| ||
| 02:13 | Fix the sqlite3_bind_blob() interface (and its kin) so that the destructor is invoked if the binding fails. Ticket [860399cc408f2dd5f41aed44b] Update the documentation to explain which interfaces invoke their destructors on failure and which do not. (check-in: d3c95e3a4e user: drh tags: trunk) | |
|
2010-10-11
| ||
| 17:58 | Make sure the sqlite3_create_collation() interfaces always return an error code if they fail. Ticket [a04e42a3fcacaffa3133436]. (check-in: 6464276ec1 user: drh tags: trunk) | |
| 17:57 | Further changes to extension loading to support compile-time options. (check-in: f818fb6f73 user: drh tags: trunk) | |
| 13:12 | Make APIs added since version 3.6.0 accessible to loadable extensions. (check-in: 76c64a3556 user: drh tags: trunk) | |
|
2010-10-08
| ||
| 16:09 | Add tests to e_createtable.test. (check-in: 38bec827f1 user: dan tags: trunk) | |
| 02:34 | Version 3.7.3 (check-in: 2677848087 user: drh tags: trunk, release) | |
|
2010-10-07
| ||
| 13:29 | Update the CLI test script to expect a version number of 3.7.3. (check-in: e55ada8924 user: drh tags: trunk) | |
|
2010-10-06
| ||
| 20:25 | Update the configure script and Makefile so that they work with 3.7.3. (check-in: 97b63a1168 user: drh tags: trunk) | |
| 18:55 | Fix long-standing bugs with the handling of LIMIT clausing in compound SELECT statements with FROM clause subqueries. Ticket [38cb5df375078d3f9]. (check-in: b0450120ea user: drh tags: trunk) | |
| 16:42 | Fix minor test suite problems causing errors on OSX. (check-in: 759c954fee user: dan tags: trunk) | |
| 09:57 | Fix a bug in test code that was causing a valgrind error. No changes to production code. (check-in: f29da7bec6 user: dan tags: trunk) | |
|
2010-10-05
| ||
| 18:22 | Do not embedded #if inside an assert() statement. Fix for check-in [dca8763872a] (check-in: d7d4a94fc1 user: drh tags: trunk) | |
| 17:02 | Have the pager change to at least WRITER_CACHEMOD state before marking any pages as dirty (instead of immediately after). Otherwise, if an error occurs, the pager may be left in WRITER_LOCKED state with dirty pages in the cache. (check-in: 471a4efbb7 user: dan tags: trunk) | |
| 15:41 | If walLockExclusive() fails for reasons other than SQLITE_BUSY inside of walRestartLog() then propagate that error back up to the application. (check-in: 04dcba6b33 user: drh tags: trunk) | |
| 12:05 | Avoid all memory allocation (and hence the possiblitity of OOM failure) in sqlite3_value_double() and sqlite3_column_double(). (check-in: 4afdf9705a user: drh tags: trunk) | |
| 11:33 | Fix an assert() failing on OSX. (check-in: dca8763872 user: dan tags: trunk) | |
| 08:13 | Prevent backcompat.test from mistaking directories for binary executables. (check-in: 717a1e50f0 user: dan tags: trunk) | |
|
2010-10-04
| ||
| 23:55 | Fix a performance regression (relative to version 3.6.23.1) caused by the query planner taking into account non-indexable WHERE clause terms to select the outermost join loops when it should be selecting tables for the outermost loop that do not benefit from being in an inner loop. (check-in: ece641eb89 user: drh tags: trunk) | |
| 16:06 | Fix a couple of test script problems. (check-in: dd10690140 user: dan tags: trunk) | |
| 15:47 | Fix memsubsys1.test so that it works with TEMP_STORE>=2. (check-in: 8ad88ee0c1 user: dan tags: trunk) | |
| 14:11 | Warning cleanup from MSVC. (check-in: cde62657d6 user: shaneh tags: trunk) | |
| 11:01 | Fix broken evidence marks in e_expr.test. (check-in: 37ff94cec8 user: dan tags: trunk) | |
| 10:45 | Add tests for table-constraints to e_createtable.test. (check-in: 70f511872e user: dan tags: trunk) | |
|
2010-10-02
| ||
| 01:01 | Backport the very lastest R-Tree in order to take advantage of its enhanced robustness to corrupt databases. (Leaf check-in: 29a681dd7b user: drh tags: apple-osx-3623) | |
| 01:00 | Fix the amalgamation builder so that it works with the rtree updates of the prior check-in. (Leaf check-in: 265b0b29b8 user: drh tags: branch-3.6.23) | |
|
2010-10-01
| ||
| 20:45 | Backport the RTree implementation from the trunk into the 3.6.23 branch. The code for the application-defined query boxes is still present but is disabled. The reason for this backport is to take advantage of recent enhancements to robustness to database corruption. (check-in: 854b23c117 user: drh tags: branch-3.6.23) | |
| 19:04 | Add tests for CHECK and UNIQUE constraints to e_createtable.test. (check-in: fb8db5581d user: dan tags: trunk) | |
| 17:23 | Add #ifndef USE_SYSTEM_SQLITE...#endif around the bundled SQLite library for the TCL bindings. (check-in: 8d85584a4e user: drh tags: trunk) | |
| 15:11 | <i>Not true: For a zeroblob, the Mem.z pointer can be null:</i> The Mem.z pointer can never been NULL for a string or blob, even a zero-length string or blob. Assert this fact. (Closed-Leaf check-in: efda310480 user: drh tags: mistake) | |
| 13:28 | Updates to the showjournal.c utility in order to bring it up to version 3. (check-in: fa97d89546 user: drh tags: trunk) | |