Many hyperlinks are disabled.
Use anonymous login
to enable hyperlinks.
380 check-ins using file src/delete.c version 5c0d89b3ef
2007-08-15
| ||
20:41 | Enhancements and smoke testing of the new memory allocation subsystem. Have not yet cut it over to the core, though. (CVS 4230) (check-in: 1dad2c0a1f user: drh tags: trunk) | |
19:16 | Test infrastructure for the new memory subsystem. (CVS 4229) (check-in: 9e50665672 user: drh tags: trunk) | |
17:08 | Modify the crash-recovery test code in test6.c for 3.5. Also change some other code to use the new sqlite3_io_methods interface. Lots of things are broken now. (CVS 4228) (check-in: af9503daf3 user: danielk1977 tags: trunk) | |
17:07 | Add a debugging memory allocator. (CVS 4227) (check-in: 8d2d1c4ff9 user: drh tags: trunk) | |
13:04 | Add initial implementations of mutex and memory subsystem modules. (CVS 4226) (check-in: c0fa376979 user: drh tags: trunk) | |
11:28 | Revisions to the interface design for 3.5. (CVS 4225) (check-in: 174116f7c0 user: drh tags: trunk) | |
2007-08-14
| ||
18:03 | Clarify documentation on sqlite3_interrupt(). (CVS 4224) (check-in: 0b5b526c9d user: drh tags: trunk) | |
17:42 | Add comments describing upcoming changes to crash-test module (test6.c). No code changes. (CVS 4223) (check-in: 63631c1737 user: danielk1977 tags: trunk) | |
13:20 | Back-port critical soft heap limit fixes to version 3.3.17. The following check-ins were back-ported: (4031), (4037), (4097), (4202), (4203), (4207), (4208), (4214). (CVS 4222) (Leaf check-in: f0029da32f user: drh tags: branch_3_3_17) | |
01:58 | Add a first cut at the new version 3.5 interface definitions to sqlite.h.in. (CVS 4219) (check-in: 0a57f3dd4d user: drh tags: trunk) | |
2007-08-13
| ||
20:31 | Version 3.4.2 (CVS 4218) (check-in: 64989904d4 user: drh tags: trunk, release) | |
16:15 | Update version number of documentation in preparation for version 3.4.2. (CVS 4217) (check-in: 1ed51e3ca0 user: drh tags: trunk) | |
15:28 | Invalidate prepared SQL statements when rolling back a transaction that includes schema modifications. (CVS 4216) (check-in: a07143fc05 user: danielk1977 tags: trunk) | |
15:18 | Disable tests that require 80-bit long doubles when the host only provides 64-bit long doubles. Ticket #2570. (CVS 4215) (check-in: 4f4f8ef416 user: drh tags: trunk) | |
14:56 | Fix a page leak in btree balance routines that follows a malloc failure. This problem detected by the new soft-heap-limit testing. (CVS 4214) (check-in: bad661bb0b user: drh tags: trunk) | |
14:41 | When loading a database schema, do not clear the SQLITE_InternChanges flag. This fixes the problem illustrated by (4211). (CVS 4213) (check-in: 540a7b28bc user: danielk1977 tags: trunk) | |
12:58 | Fix the test labels in mallocC.test. (CVS 4212) (check-in: 7d4cdff444 user: drh tags: trunk) | |
12:39 | Add a new test cases for a problem with malloc failure during the parsing of CREATE TABLE statements in autovacuum mode. (CVS 4211) (check-in: be50387e65 user: drh tags: trunk) | |
11:10 | Remove a memory allocation from sqlite3VdbeIOTraceSql(). This means that when doing I/O tracing, the text of an SQL statement is truncated to the first 1000 characters. But it also means tracing works in an out-of-memory situation. (CVS 4210) (check-in: 578da2476e user: drh tags: trunk) | |
2007-08-12
| ||
20:07 | Updates to test scripts for better support of soft-heap-limit testing. (CVS 4209) (check-in: 990f621f22 user: drh tags: trunk) | |
2007-08-11
| ||
00:26 | Backout check-in (4206) and replace it with a much better and more general fix for the problem identified in ticket #2565. (CVS 4208) (check-in: 7961a73850 user: drh tags: trunk) | |
2007-08-10
| ||
23:56 | Fix some incorrect asserts() in the pager - problems brought to light by the new soft-heap-limit testing apparatus of check-in (4202). (CVS 4207) (check-in: 51f3e01b74 user: drh tags: trunk) | |
23:54 | Fix the corruption problem of ticket #2565 as demonstrated by the test added in (4204). There may yet be other instances of similar problems lurking in the code. (CVS 4206) (check-in: 7ed2f59e70 user: drh tags: trunk) | |
23:47 |
Convert fts2 to use sqlite3_prepare_v2() to prevent certain logic
errors around SQLITE_SCHEMA handling. This also allows
sql_step_statement() and sql_step_leaf_statement() to be replaced with
sqlite3_step().
Also fix a logic error in flushPendingTerms() which was clearing the term table in case of error. This was wrong in the face of SQLITE_SCHEMA. Even though the change to sqlite3_prepare_v2() should cause us not to see SQLITE_SCHEMA any longer, it was still a logic error... (CVS 4205) (check-in: 16730cb137 user: shess tags: trunk) | |
19:46 | Add a test case to reproduce the database corruption problem reported by ticket #2565. (CVS 4204) (check-in: f267ce8094 user: drh tags: trunk) | |
19:46 | Update tests so that they work properly even if the soft-heap-limit is set low. (CVS 4203) (check-in: e01eb99edf user: drh tags: trunk) | |
16:41 | Add the --soft-heap-limit=N option to the test scripts. Set this value to enforce a soft heap limit across all tests. This changes exposes several new bugs. (CVS 4202) (check-in: 8be8459816 user: drh tags: trunk) | |
2007-08-09
| ||
00:00 | Fix documentation typos. Tickets #2168 and #2560. (CVS 4201) (check-in: 4333b231af user: drh tags: trunk) | |
2007-08-08
| ||
12:11 | More tweaking of linkage. Ticket #2554. (CVS 4200) (check-in: 3759a38fe8 user: drh tags: trunk) | |
01:04 | Additional refinements on the amalgamation: Give a couple of constants file scope and add the SQLITE_API prefix to some interfaces that need it. Ticket #2554. (CVS 4199) (check-in: c09ec102c4 user: drh tags: trunk) | |
2007-08-07
| ||
17:13 | Remove all tabs from source code files. Ticket #2556. (CVS 4198) (check-in: 7550dd1d59 user: drh tags: trunk) | |
17:04 | Revise the amalgamation so that all symbols have file scope except those that begin with "sqlite3_". Ticket #2554. (CVS 4197) (check-in: 73db545289 user: drh tags: trunk) | |
2007-08-05
| ||
23:52 | Fix some compiler warnings. (CVS 4196) (check-in: 6cc15409ad user: drh tags: trunk) | |
2007-08-03
| ||
08:18 | Change some (English language) grammar in faq.tcl. Ticket #2480. No code changes. (CVS 4195) (check-in: e526817f15 user: danielk1977 tags: trunk) | |
07:33 | Avoid reloading the schema when an expired statement is finalized or reset. This is only necessary when the in-memory schema does not match the file-system schema. (CVS 4194) (check-in: 61de5b52da user: danielk1977 tags: trunk) | |
2007-07-30
| ||
23:02 | Fix a typo in limits.html. (CVS 4193) (check-in: c709140b51 user: drh tags: trunk) | |
20:41 | Make excess output from the CLI appear inside SQL comments so as not to interfere with generated SQL. Ticket #2544. (CVS 4192) (check-in: 49a2e85511 user: drh tags: trunk) | |
18:55 | Fix ticket #2439: the FTS1 and FTS2 extensions use the non-standard, unportable and highly deprecated <malloc.h> header on all platforms except Apple Mac OS X. The <malloc.h> actually is never required on any OS with an at least partly POSIX-conforming API as the malloc(3) & friends functions officially live in <stdlib.h> since over 10 years. Under some platform like FreeBSD the inclusion of <malloc.h> since a few years even causes an "#error" and this way a build failure. So, just get rid of the bad <malloc.h> usage in FTS1 and FTS2 extensions at all and stick with <stdlib.h> there only. (CVS 4191) (check-in: 3f9a666143 user: rse tags: trunk) | |
18:31 | "extern" declarations inside function bodies are not every compiler's favorite, so move to global scope. Additionally, at least under Unix environment use <unistd.h> to get the proper prototype instead of using a K&R-style own declaration. (CVS 4190) (check-in: 5955a77d6c user: rse tags: trunk) | |
18:26 | Get rid of the following compile-time warnings: tool/mkkeywordhash.c:353: warning: array subscript has type 'char' tool/mkkeywordhash.c:354: warning: array subscript has type 'char' (CVS 4189) (check-in: c01d794479 user: rse tags: trunk) | |
18:24 | Get rid of the following compile-time warning: "src/shell.c:112: warning: 'iotracePrintf' defined but not used" (CVS 4188) (check-in: 271e27fd30 user: rse tags: trunk) | |
14:40 | Fix a bug with explicit collation sequences attached to a column reference expression that is on the right-hand-side of a binary comparison operator. (CVS 4187) (check-in: 7b69968618 user: danielk1977 tags: trunk) | |
2007-07-26
| ||
10:16 | Test that an explicit collation sequence overrides an implicit one attached to a 'new' reference (it does). No code changes. (CVS 4186) (check-in: a443b07ed6 user: danielk1977 tags: trunk) | |
06:50 | Fix two obscure memory leaks that can follow a malloc() failure in sqlite3_set_auxdata(). Ticket #2534. (CVS 4185) (check-in: b88af1827b user: danielk1977 tags: trunk) | |
2007-07-25
| ||
00:56 | Implement xRename() for fts1 so that it is possible to rename fts1 tables. See http://www.sqlite.org/cvstrac/chngview?cn=4143 (CVS 4184) (check-in: febf75f022 user: shess tags: trunk) | |
00:38 |
Replicates http://www.sqlite.org/cvstrac/chngview?cn=4151 which
modified fts2:
Modify handling of SQLITE_SCHEMA in fts2 code. An SQLITE_SCHEMA error may cause SQLite to reload the internal schema, deleting and recreating v-table objects. So the sqlite3_vtab structure can be deleted out from under a v-table implementation. (CVS 4183) (check-in: f9020cffda user: shess tags: trunk) | |
00:28 | Sorry, previous check-in included a last-minute "Did it really work?" change :-). (CVS 4182) (check-in: 5db25e369a user: shess tags: trunk) | |
00:25 | Apply change 4095 to fts1. Fix snippet generation when the left-most column of an fts table is used in the MATCH clause. Fix for ticket #2429. (CVS 4181) (check-in: c2ba3cc0f7 user: shess tags: trunk) | |
2007-07-24
| ||
10:22 | Remove the comment in the documentation that says that incremental vacuum mode is not persistent because as of 3.4.1 it is persistent. Ticket #2528. (CVS 4180) (check-in: 6c4040941e user: drh tags: trunk) | |
2007-07-23
| ||
22:51 | Detect nested aggregate functions even if the inner aggregate function is aliased using an AS clause. Ticket #2526. (CVS 4179) (check-in: de000280c6 user: drh tags: trunk) | |
19:39 | The REPLACE conflict resolution falls back to FAIL when on a CHECK constraint violation. Ticket #2525. (CVS 4178) (check-in: b213614abf user: drh tags: trunk) | |
19:31 | Move prototypes for LEMON-generated parser functions into sqliteInt.h. Ticket #2521. (CVS 4177) (check-in: bc02c7f162 user: drh tags: trunk) | |
19:26 | Check the return value of sqlite3PagerWrite() when autovacuuming. Ticket #2524. (CVS 4176) (check-in: b4a5c62b85 user: drh tags: trunk) | |
19:12 | Rework the UTF8 reader logic in order to avoid the use of malloc(). Ticket #2523. (CVS 4175) (check-in: 9a059cb6bc user: drh tags: trunk) | |
2007-07-22
| ||
19:10 | Fix a bad sizeof in vdbe.c. Ticket #2522. (CVS 4174) (check-in: 77ebc3feb0 user: drh tags: trunk) | |
2007-07-21
| ||
19:41 | Work around a bug in GCC. Ticket #2469. (CVS 4173) (check-in: 1924ba5207 user: drh tags: trunk) | |
2007-07-20
| ||
20:25 | Work around an AFP file system byte-range locking error caused by opening and closing a second file descriptor on the database file as RDWR which clears the exclusive byte-range lock and results in an fcntl error when the AFP locking code attempts to clear it afterwards. This error is caused by the test harness and isn't a flaw in the sqlite or its implementation of AFP file locking. (CVS 4172) (check-in: b2ba3e81fb user: aswift tags: trunk) | |
13:42 | Updates to build scripts. Added scripts for building FTS2 shared libraries. (CVS 4171) (check-in: e269f6e9f6 user: drh tags: trunk) | |
10:56 | Version 3.4.1 (CVS 4170) (check-in: 81a4dd07c1 user: drh tags: trunk) | |
10:48 | Keep sqlite3ext.h backwards compatible. This really should have been part of the previous check-in. (CVS 4169) (check-in: 7cf15a5f65 user: drh tags: trunk) | |
10:33 | Add new APIs to the end of the sqlite3_apis structure in loadext.c, in order to preserve backwards compatibility. Fix an incompatibility introduced by (4108). (CVS 4168) (check-in: 639993490e user: drh tags: trunk) | |
01:17 | Documentation change in preparation for version 3.4.1. (CVS 4167) (check-in: cf227766ac user: drh tags: trunk) | |
00:35 | Fix a memory leak in the Rename method of the echo test virtual table. No changes to the core. (CVS 4166) (check-in: e5c132fff3 user: drh tags: trunk) | |
00:33 | Cleanup unused function in pager.c. This is the conclusion of the fix to ticket #2518. We believe the fix is complete and correct. (CVS 4165) (check-in: 31dba1e933 user: drh tags: trunk) | |
2007-07-19
| ||
22:30 | Disable tests that use the progress handler when the progress handler is disabled at compile-time. Followup to ticket #2497. (CVS 4164) (check-in: cbc56fd4ec user: drh tags: trunk) | |
16:35 | Preliminary fix for ticket #2518. Make sure the VACUUM command increments the change counter. (CVS 4163) (check-in: 75263797e2 user: drh tags: trunk) | |
12:41 | Clarify the documentation on the nByte parameter to sqlite3_prepare(). Make it clear that nByte is a maximum string length. Ticket #2516. (CVS 4162) (check-in: d1ae3de461 user: drh tags: trunk) | |
2007-07-18
| ||
18:17 | Get the min/max optimization working with descending indices. Ticket #2514. (CVS 4161) (check-in: a80a3c9d0a user: drh tags: trunk) | |
18:16 | Additional diagnostics added to the "out" file generated by lemon. (CVS 4160) (check-in: 7ef2aaf72a user: drh tags: trunk) | |
2007-07-17
| ||
17:22 | Fix annoying emacs warning about suspicious line. The change removes a leading space (emacs wants Makefiles to use leading tabs there). (CVS 4159) (check-in: 4daadf659a user: shess tags: trunk) | |
2007-07-13
| ||
10:36 | Add an extra integrity_check to the transaction test script. (CVS 4158) (check-in: 5ea43b2db1 user: drh tags: trunk) | |
10:35 | Update documentation to make it clear that VACUUM is free to change implicit rowids. Ticket #2382. (CVS 4157) (check-in: d111b1daa5 user: drh tags: trunk) | |
10:26 | "extern" prototypes cause problems for the Digital Mars compiler. So remove them. Ticket #2502. (CVS 4156) (check-in: f35fbf8070 user: drh tags: trunk) | |
2007-07-12
| ||
13:18 | Fix for #2497. Set the database error code to the return value of sqlite3_step(). (CVS 4155) (check-in: b01fda5f50 user: danielk1977 tags: trunk) | |
2007-07-03
| ||
23:37 | Fix compilation warnings shown when using strict GCC options for os_os2.c. (CVS 4154) (check-in: 22277b7e1b user: pweilbacher tags: trunk) | |
05:31 | Fix a memory leak in shell.c. Reported on the mailing list. This has only been informally tested. (CVS 4153) (check-in: a008905b39 user: danielk1977 tags: trunk) | |
2007-07-02
| ||
19:31 | Fix a comment typo reported on the mailing list. (CVS 4152) (check-in: 25e6df9cdd user: drh tags: trunk) | |
10:16 | Modify handling of SQLITE_SCHEMA in fts2 code. An SQLITE_SCHEMA error may cause SQLite to reload the internal schema, deleting and recreating v-table objects. So the sqlite3_vtab structure can be deleted out from under a v-table implementation. (CVS 4151) (check-in: dee1a0fd28 user: danielk1977 tags: trunk) | |
2007-07-01
| ||
21:18 | Remove an unnecessary temporary variable for clarity of presentation in a loop that GCC 4.2.0 is miscompiling. Ticket #2469. GCC 4.2.0 miscompiles this loop regardless of whether or not the temporary variable is used, but by removing the variable, we hope to make it easier to explain the problem to GCC maintainers. The error only appears if -ftree-vrp is used (which is turned on by -O2). (CVS 4150) (check-in: 35ae398bd3 user: drh tags: trunk) | |
15:41 | Change OS/2 version of sqlite3Os2FullPathname() for cross-compiler compatibility: - allocate zBuff on demand (restricted stack space on old compilers) - 2 bytes in zDrive in include '\0' - pass drive number to DosQueryCurrentDir() instead of 0 to make EMX work - zFull does not need to be preallocated (CVS 4149) (check-in: cc21051765 user: pweilbacher tags: trunk) | |
2007-06-30
| ||
16:30 | Including os2safe.h from os_os2.c is no longer necessary (now that it's including from os.h) (CVS 4148) (check-in: 6432bbe1c1 user: pweilbacher tags: trunk) | |
15:24 | old emx/gcc sets __EMX__ (with double underscores) as builtin #define (CVS 4147) (check-in: d69204fec7 user: pweilbacher tags: trunk) | |
2007-06-29
| ||
12:04 | Set FD_CLOEXEC on all open files under Unix. Ticket #2475. (CVS 4146) (check-in: f1e5fed8eb user: drh tags: trunk) | |
2007-06-27
| ||
23:52 | Add a test case to verify that ticket #2470 has been fixed. (CVS 4145) (check-in: b3f442698e user: drh tags: trunk) | |
17:09 | Test error handling when renaming fts2 tables. (CVS 4144) (check-in: 57840eba6a user: danielk1977 tags: trunk) | |
16:26 | Implement xRename() for fts2 so that it is possible to rename fts2 tables. (CVS 4143) (check-in: 488474fde7 user: danielk1977 tags: trunk) | |
15:53 | Allow ALTER TABLE on virtual tables. (CVS 4142) (check-in: 37d1f9f37e user: danielk1977 tags: trunk) | |
15:01 | respect $LDFLAGS from env/configure (CVS 4141) (check-in: 9c13fc0f4b user: vapier tags: trunk) | |
11:10 | Fix for compiling with OMIT_VIRTUALTABLE defined. (CVS 4140) (check-in: 850822e290 user: danielk1977 tags: trunk) | |
10:21 | Fix typo in comments of the sqlite.h.in header file - comments from which documentation is automatically extracted and generated. Ticket #2468. (CVS 4139) (check-in: 53b526203c user: drh tags: trunk) | |
10:20 | Make sure auto_vacuum is turned off before running the tests in pragma2.test. (CVS 4138) (check-in: fc85c8aa16 user: drh tags: trunk) | |
00:36 | Set the connection error code when sqlite3_blob_read() or sqlite3_blob_write() fail. Ticket #2464. (CVS 4137) (check-in: 3bd990be3c user: drh tags: trunk) | |
00:08 | Update the FAQ to describe how moving sqlite3 handles across threads is not safe with SQLITE_ENABLE_MEMORY_MANAGEMENT. Tickets #2357 and #2463. (CVS 4136) (check-in: 2640f518de user: drh tags: trunk) | |
2007-06-26
| ||
22:55 | Make sure the TCL bindings always use Tcl_GetWideIntFromObj() even if the reported type is "int" because on x86-64 and "int" type is 64-bits. Ticket #2465. (CVS 4135) (check-in: 5c93324b93 user: drh tags: trunk) | |
22:42 | Fix a test case so that it (hopefully) works on x86-64. Ticket #2465. (CVS 4134) (check-in: 3daf7cae18 user: drh tags: trunk) | |
22:10 | Remove vestigal code from Makefile.linux-gcc. Ticket #2466. (CVS 4133) (check-in: 338c187910 user: drh tags: trunk) | |
12:54 | Reorganize comments in fts2_tokenizer.h. No code changes. (CVS 4132) (check-in: b331e30395 user: danielk1977 tags: trunk) | |
12:52 | Use (((i64)1)<<63) instead of just (1<<63) to compute the most negative 64-bit integer. (CVS 4131) (check-in: be2570c061 user: danielk1977 tags: trunk) | |
11:13 | Fix for #2444: Avoid SIGFPE on 64-bit platforms when evaluating expressions like ((1<<63)/-1). (CVS 4130) (check-in: c6dfd9e434 user: danielk1977 tags: trunk) | |
10:56 | Remove the unused EXTSRC variable from the non-configure makefile. (CVS 4129) (check-in: bbdcf372c6 user: danielk1977 tags: trunk) | |
10:55 | Fix an unitialized variable in fts2. (CVS 4128) (check-in: c349cf9425 user: danielk1977 tags: trunk) | |
10:38 | Allow virtual tables to mark columns as 'hidden'. Hidden columns do not show up in "*" expansion, the implicit column list of an INSERT statement, or the results of a "PRAGMA table_info" statement. (CVS 4127) (check-in: 7f887a6a00 user: danielk1977 tags: trunk) | |
01:04 | Try to work around an MSVC compiler bug. Ticket #2457. (CVS 4126) (check-in: 4a7079a19a user: drh tags: trunk) | |
00:52 | Modify the amalgamation generator to identify every API using the SQLITE_API macro which is normally defined to nothing but which can be overridden on the compiler command-line to be "static" if desired. Ticket #2453. (CVS 4125) (check-in: 474a52347d user: drh tags: trunk) | |
00:37 | Change the name of sqlite3_atoi64 back to sqlite3Atoi64 so that it is clearly and internal API. Modify the amalgamation to use a macro SQLITE_PRIVATE instead of the keyword "static" so the macro can be overridden to by nothing. (CVS 4124) (check-in: 96190cf13d user: drh tags: trunk) | |
2007-06-25
| ||
17:28 | Rework the string to integer conversion logic to address tickets #2344 and #2454. (CVS 4123) (check-in: 5b3a490649 user: drh tags: trunk) | |
16:29 | Fix for #2445. A bug in the lookupName() logic that could cause a crash when a WHERE clause used an alias to refer to an expression in the result-set of the SELECT, and that expression was itself a reference to a table column. (CVS 4122) (check-in: 044ca1c72a user: danielk1977 tags: trunk) | |
14:28 | Modify the non-configure build system to make it easier to build the library with the fts2 or icu extensions linked in. (CVS 4121) (check-in: 02b23c4394 user: danielk1977 tags: trunk) | |
13:50 | Put #ifdefs in fts2_tokenizer so that the build works even when FTS2 is omitted. Add the SQLite blessing to the header comments on all FTS2 source files. (CVS 4120) (check-in: c795e6fd8f user: drh tags: trunk) | |
12:49 | All the use of MySQL-style quoting in the FTS modules. Ticket #2446. (CVS 4119) (check-in: 3be2a6d1c3 user: drh tags: trunk) | |
12:05 | Add a test that calls fts2_tokenizer() with an argument set via C code. (CVS 4118) (check-in: fbcf2d75cd user: danielk1977 tags: trunk) | |
11:24 | Add some tests for the fts2 icu tokenizer. (CVS 4117) (check-in: b79ced3e0a user: danielk1977 tags: trunk) | |
09:52 | Add some documentation for user-defined fts2 tokenizers. (CVS 4116) (check-in: 5a9eee8658 user: danielk1977 tags: trunk) | |
08:16 | Make the auto_vacuum mode peristent in all cases. (CVS 4115) (check-in: 5b0408ddd0 user: danielk1977 tags: trunk) | |
2007-06-24
| ||
16:11 | Fix for #2451. Code comment changes only. (CVS 4114) (check-in: bc61dcbf64 user: danielk1977 tags: trunk) | |
10:14 | Make setting "PRAGMA auto_vacuum = incremental" persistent if it is executed before the database file is created. (CVS 4113) (check-in: c666fad151 user: danielk1977 tags: trunk) | |
08:00 | Add the "(database.)freelist_count" PRAGMA. For querying the size of the database free-list. (CVS 4112) (check-in: 1fb4251a70 user: danielk1977 tags: trunk) | |
06:32 | When expanding '*' in the result set of a SELECT, quote the expanded identifiers. Fix for #2450. (CVS 4111) (check-in: d5b7224f37 user: danielk1977 tags: trunk) | |
2007-06-22
| ||
20:17 | fix linking failure on OS/2 that happens with the amalgamation and the high memory option enabled (CVS 4110) (check-in: f35e20e196 user: pweilbacher tags: trunk) | |
20:04 | fix compilation failure on OS/2 with the amalgamation (CVS 4109) (check-in: 12327ca049 user: pweilbacher tags: trunk) | |
15:21 | Extend fts2 so that user defined tokenizers may be added. Add a tokenizer that uses the ICU library if available. Documentation and tests to come. (CVS 4108) (check-in: 68677e420c user: danielk1977 tags: trunk) | |
2007-06-21
| ||
15:25 | Clarify documentation of the column metadata APIs. Make sure that the new documentation claims are tested. (CVS 4107) (check-in: 2dafe08a91 user: drh tags: trunk) | |
13:30 | Update the URL to mailing list archives. (CVS 4106) (check-in: af4b914a21 user: drh tags: trunk) | |
2007-06-20
| ||
16:20 | Use NAWK everywhere in Makefile.in. Ticket #2437. (CVS 4105) (check-in: 314669b6cc user: drh tags: trunk) | |
16:13 | A column name preceded by a unary "+" operator is considered to be a column name when computing the collating sequence to be used by a comparison operator or ORDER BY clause. (CVS 4104) (check-in: 99a20f6a42 user: drh tags: trunk) | |
15:29 | Remove the dependency on libm for isnan(). Ticket #2436. (CVS 4103) (check-in: 406675bb1c user: drh tags: trunk) | |
15:14 | Remove an unused variable from btree.c. (CVS 4102) (check-in: cb47a8297f user: drh tags: trunk) | |
13:37 | Fix comment typo in analyze.c. Ticket #2434. (CVS 4101) (check-in: df9957b947 user: drh tags: trunk) | |
13:10 | Fix a bug in the command-line shell that prevents the reading of the ".sqliterc" file. Ticket #2433. (CVS 4100) (check-in: 6eaf29f5e1 user: drh tags: trunk) | |
12:18 | Disallow empty GROUP BY clauses. Ticket #2431. (CVS 4099) (check-in: 9581e7a4a4 user: drh tags: trunk) | |
11:56 | Additional changes for ticket #2426. (CVS 4098) (check-in: 79debf95cd user: drh tags: trunk) | |
09:09 | Fix for #2432 and partial fix for #2427 (documentation change only). (CVS 4097) (check-in: f025a74531 user: danielk1977 tags: trunk) | |
08:49 | Change main.mk to make it easier to build testfixture and sqlite3 with the icu or fts2 extensions linked in. (CVS 4096) (check-in: 91719cc8ea user: danielk1977 tags: trunk) | |
06:23 | Fix snippet generation when the left-most column of an fts2 table is used in the MATCH clause. Fix for ticket #2429. (CVS 4095) (check-in: fec56ad2ed user: danielk1977 tags: trunk) | |
2007-06-19
| ||
23:01 | Change the TCL bindings so that @aaa always treats the variable aaa as a bytearray and binds the value as a BLOB. This change is backwards compatible since the $ behavior is unchanged and @ was not accepted until the current round of changes. (CVS 4094) (check-in: 6f7d55aced user: drh tags: trunk) | |
17:48 | Cleanup the TCL interface source code and add documentation for recently added methods on the SQLite object of TCL. (CVS 4093) (check-in: d88b79818a user: drh tags: trunk) | |
17:15 | In the TCL bindings, if a TCL variable has a bytearray representation and the host parameter starts with @ instead of $, then always store the content as a BLOB not as a string even if a string representation is also available. (CVS 4092) (check-in: dcb104bd41 user: drh tags: trunk) | |
15:23 | Change the name of the "limits.h" source file to "sqliteLimit.h". Ticket #2428. (CVS 4091) (check-in: 64bcc41f12 user: drh tags: trunk) | |
10:58 | Do not segfault even if sqlite is misused by requesting query results after the query has been reset. ticket #2426. (CVS 4090) (check-in: 783f19be38 user: drh tags: trunk) | |
10:50 | Cast the 2nd parameter of ftruncate to off_t to work around bugs in some unix implementations. Ticket #2425. (CVS 4089) (check-in: 0b20a69609 user: drh tags: trunk) | |
2007-06-18
| ||
17:44 | Another attempt to fix the build process for TCL windows bindings. (CVS 4088) (check-in: 1fd2a358d6 user: drh tags: trunk) | |
17:25 | Bring the pager_datahash() macro into sync with its function definition. Ticket #2422. (CVS 4087) (check-in: ba967bbb62 user: drh tags: trunk) | |
16:06 | Fix the build scripts so that they generate Tcl bindings for windows correctly. (CVS 4086) (check-in: c7be8b419f user: drh tags: trunk) | |
13:33 | Version 3.4.0 (CVS 4085) (check-in: 2647980fba user: drh tags: trunk) | |
12:22 | Documentation updates prior to release 3.4.0. Enhanced support for soak testing in the test scripts. (CVS 4084) (check-in: eec3871038 user: drh tags: trunk) | |
2007-06-16
| ||
18:39 | Fix a faulty assert() in the pager. (CVS 4083) (check-in: 4cff47b061 user: drh tags: trunk) | |
14:43 | Modify the tests in exclusive2.test to account for the fact that the pager cache may be very small by default. These tests are only valid if the entire database fitting in the cache. (CVS 4082) (check-in: 04a13fe43e user: danielk1977 tags: trunk) | |
14:19 | Add another scenario to the soak1.tcl script. (CVS 4081) (check-in: 10e6822d4f user: drh tags: trunk) | |
11:17 | A minor logic correction in the previous check-in. Also added a lengthy comment describing the meanings of various flags in the {quote: PgHdr} structure. (CVS 4080) (check-in: 57bf8204cd user: drh tags: trunk) | |
04:42 | Fix a database corruption problem that can occur in auto-vacuum mode when a malloc() failure causes a statement rollback, additional statements are run in the same transaction, then the total transaction rolls back. (CVS 4079) (check-in: c9dcf2b926 user: drh tags: trunk) | |
03:06 | Additional debugging instrumentation added to the pager. (CVS 4078) (check-in: dcdb20f81a user: drh tags: trunk) | |
2007-06-15
| ||
20:29 | Make arrangements to optionally print a debug message at the point of a simulated malloc() failure when SQLITE_MEMDEBUG is defined. (CVS 4077) (check-in: 7d3c1f08a3 user: drh tags: trunk) | |
18:59 | Fix more documentation typos. Ticket #2416. (CVS 4076) (check-in: b55318b12d user: drh tags: trunk) | |
18:53 | Remove a C++ism from the test harness. Get the TCL interface working with older versions of TCL. (CVS 4075) (check-in: c8beb8674e user: drh tags: trunk) | |
18:43 | Add a script for "soak" testing. (CVS 4074) (check-in: d4ea1287a8 user: drh tags: trunk) | |
17:50 | Fix a race condition in test_server.c/test7.c (test changes only). (CVS 4073) (check-in: d0de32e4c6 user: danielk1977 tags: trunk) | |
17:04 | Fix typo in change comments for 3.4.0. Ticket #2416 (CVS 4072) (check-in: 1f6f033000 user: drh tags: trunk) | |
17:03 | Fix a memory leak in the parser that can occur following a malloc failure. (CVS 4071) (check-in: d0b6e9a07e user: drh tags: trunk) | |
16:37 | Fix an obscure memory leak in the SQL compiler.. (CVS 4070) (check-in: d4ab94288b user: danielk1977 tags: trunk) | |
15:31 | Fix a memory leak that was discovered by the tests that were added to cover ticket #2339. (CVS 4069) (check-in: 1d10a48934 user: drh tags: trunk) | |
15:08 | Fix a test suite bug. To accommadate the fact that handles returned by (db incrblob) change for the second test suite iteration of full.test. (CVS 4068) (check-in: 32218834b8 user: danielk1977 tags: trunk) | |
14:53 | Fix for #2415. The progress handler should abandon only the current query, not all active queries. (CVS 4067) (check-in: 115e19fe22 user: danielk1977 tags: trunk) | |
13:57 | Do not try to run the fuzz_malloc tests if SQLITE_MEMDEBUG is not enabled. (CVS 4066) (check-in: 452115959f user: drh tags: trunk) | |
12:41 | Clean up a variable in tkt2409.test that would cause subsequent tests to fail. (CVS 4065) (check-in: 0f944cb9fb user: drh tags: trunk) | |
12:16 | Documentation updates in preparation for 3.4.0. (CVS 4064) (check-in: 5232fd29d0 user: drh tags: trunk) | |
12:06 | Update the btree file format description in btreeInt.h. (CVS 4063) (check-in: 1ac160d3dc user: drh tags: trunk) | |
2007-06-14
| ||
20:57 | The C-api reference documentation is now generated directly from comments in the sqlite3.h header file. (CVS 4062) (check-in: d93c41e9a9 user: drh tags: trunk) | |
2007-06-13
| ||
16:49 | Have queries interrupted by the progress-handler return SQLITE_INTERRUPT. Rollback any active transaction if a DML statement returns SQLITE_INTERRUPT. (CVS 4061) (check-in: 33454b5691 user: danielk1977 tags: trunk) | |
15:22 | Fix for #2409. Return SQLITE_IOERR_BLOCKED instead of SQLITE_BUSY in cases where failure to obtain a database lock leaves the cache in an inconsistent state. See additional information at CorruptionFollowingBusyError. (CVS 4060) (check-in: ce2c9925d0 user: danielk1977 tags: trunk) | |
2007-06-12
| ||
18:50 | In the "transaction" command of the TCL interface, if a COMMIT fails finish it with a rollback. (CVS 4059) (check-in: 6da39fa442 user: drh tags: trunk) | |
18:20 | Minor comment edits from my prefix development client. No code changes. (CVS 4058) (check-in: 6953cd0935 user: shess tags: trunk) | |
2007-06-11
| ||
23:23 | Fix typo in the limits.html document. Ticket #2410. (CVS 4057) (check-in: d3fe186c33 user: drh tags: trunk) | |
12:56 | Fix a bug in the LIKE optimization that was causing the pattern string to be dequoted twice. Ticket #2407. (CVS 4056) (check-in: abf64d1d0a user: drh tags: trunk) | |
08:00 | Add a README.txt file for the ICU extension. (CVS 4055) (check-in: 7b6927829f user: danielk1977 tags: trunk) | |
2007-06-10
| ||
22:57 | Define isnan() on windows. Ticket #2399. (CVS 4054) (check-in: fed9373e27 user: drh tags: trunk) | |
2007-06-09
| ||
09:53 | Documentation updates, in particular in the new limits.html file is added. (CVS 4053) (check-in: 4ca6cdae94 user: drh tags: trunk) | |
2007-06-08
| ||
18:27 | Use C-sylte comments exclusively, never C++ comments. Ticket #2406. (CVS 4052) (check-in: 8f5b789fea user: drh tags: trunk) | |
08:43 | Additional test cases for comparisons against NULL in the WHERE clause and elsewhere in a SELECT. (CVS 4051) (check-in: 72612a0373 user: drh tags: trunk) | |
08:39 | Make sure rowid comparisons against NULL work correctly. Ticket #2404. (CVS 4050) (check-in: 11ee8ea43f user: drh tags: trunk) | |
00:20 | Fix the query optimizer so that it correctly handles constant expressions in the ON clause of a LEFT JOIN. Ticket #2403. (CVS 4049) (check-in: 46fdd19548 user: drh tags: trunk) | |
2007-06-07
| ||
19:08 | Make sure zeroblob does reasonable things with a negative argument or an argument that is larger than the maximum blob size. (CVS 4048) (check-in: f40218434e user: drh tags: trunk) | |
15:45 | Fix an error in test file sqllimits1.test that was causing a test to fail when SQLITE_MAX_EXPR_DEPTH was defined. (CVS 4047) (check-in: e66aa2c34a user: danielk1977 tags: trunk) | |
10:55 | Add the SQLITE_MAX_COMPOUND_SELECT compile-time parameter for limiting the number of terms in a compound select statement. Set the default limit to 100. (CVS 4046) (check-in: 0d71ad4591 user: drh tags: trunk) | |
2007-06-02
| ||
07:54 | Fix a vdbe stack leak that could occur where one side of a WHERE clause inequality evaluated to SQL null. (CVS 4045) (check-in: 17152bf1a2 user: danielk1977 tags: trunk) | |
2007-05-31
| ||
08:20 | Extend out-of-memory testing with fuzzily generated sql some. One fix for a problem found by the same. (CVS 4044) (check-in: d2282e64f1 user: danielk1977 tags: trunk) | |
2007-05-30
| ||
10:36 | Add some extra tests for malloc failure during expression parsing and execution using fuzzily generated SQL. (CVS 4043) (check-in: 7522d2fb32 user: danielk1977 tags: trunk) | |
08:18 | Add the start of the soak-test infrastructure. (CVS 4042) (check-in: 5d0b247ca1 user: danielk1977 tags: trunk) | |
06:19 | Fix a problem with zeroblob() and CAST(...) expressions that could cause an assert() to fail. (CVS 4041) (check-in: 03750a2a6b user: danielk1977 tags: trunk) | |
2007-05-29
| ||
12:11 | Consider explicit collate clauses when matching WHERE constraints to indices. Fix for #2391. (CVS 4040) (check-in: f9a95e92df user: danielk1977 tags: trunk) | |
2007-05-24
| ||
10:18 | Add some tests to verify that the parameter to the incremental_vacuum pragma is working. It is. (CVS 4039) (check-in: 1abf784fe2 user: danielk1977 tags: trunk) | |
09:44 | Fix a typo in comments in sqlite.h.in. No code changes. Ticket #2384. (CVS 4038) (check-in: e8ae6214bf user: danielk1977 tags: trunk) | |
09:41 | Remove an assert() statement that may not be true if an IO or malloc error has occured. (CVS 4037) (check-in: 16c8e9c20c user: danielk1977 tags: trunk) | |
09:20 | Fix a page reference leak that could occur after an IO error. (CVS 4036) (check-in: d00c8242c6 user: danielk1977 tags: trunk) | |
07:47 | Remove an incorrect assert() statement that was being triggered by ioerr tests run in utf-16 mode. (CVS 4035) (check-in: c6575173be user: danielk1977 tags: trunk) | |
07:22 | Fix an assert() failure occuring in corrupt.test. (CVS 4034) (check-in: 3c54cddf23 user: danielk1977 tags: trunk) | |
2007-05-23
| ||
16:23 | Add some code to MemTranslate() to prevent the READ_UTF8() macro from overreading a buffer. (CVS 4033) (check-in: 0595319cd7 user: danielk1977 tags: trunk) | |
13:50 | Make sure the database schema has been read before compiling an incrmental_vacuum pragma. (CVS 4032) (check-in: efd7bcb34c user: danielk1977 tags: trunk) | |
13:34 | Fix a leaked page reference that could occur after an IO error in auto-vacuum databases. Also modify incrvacuum.test to work with DEFAULT_AUTOVACUUM=1 builds. (CVS 4031) (check-in: e691f2fa3d user: danielk1977 tags: trunk) | |
09:52 | Fix a database corrupting bug in the optimistic overflow chain. Exposed by running bigrow.test with auto-vacuum enabled. (CVS 4030) (check-in: 10ee2d50cb user: danielk1977 tags: trunk) | |
07:20 | Additional comments ono the SQLITE_MIXED_ENDIAN_64BIT_FLOAT issue. (CVS 4029) (check-in: dd48a31e67 user: drh tags: trunk) | |
06:31 | Make sure zeroblob() works with the concatenation operator. Ticket #2379. (CVS 4028) (check-in: 6e19ef7582 user: drh tags: trunk) | |
06:25 | Increase the number of repititions in crash.test. (CVS 4027) (check-in: f3c3412afa user: danielk1977 tags: trunk) | |
2007-05-21
| ||
21:59 | Fix overzealous fts2 assertions WRT rowid 0 or lower. Only check that docids are ascending if there was a prior docid set for the doclist, ignore the initial docid of 0. (CVS 4026) (check-in: ed3a131f1d user: shess tags: trunk) | |
2007-05-19
| ||
11:50 | Add the larger SQLite icon to the repository. (CVS 4025) (check-in: d7539c6e8b user: drh tags: trunk) | |
06:48 | Add a few extra words to the sqlite3_finalize() documentation. (CVS 4024) (check-in: 315dc71b92 user: danielk1977 tags: trunk) | |
2007-05-17
| ||
18:28 | Change the way overflow pages are allocated from the free-list to increase the chance that they are a contiguous block. Much more can still be done. (CVS 4023) (check-in: b739189204 user: danielk1977 tags: trunk) | |
16:38 | Bring the autoconf based build system up to date with the recent changes. #2366, #2367 and #2368. (CVS 4022) (check-in: ae644bc7aa user: danielk1977 tags: trunk) | |
16:34 | Fix a problem with casting blobs to numeric types. (CVS 4021) (check-in: f2ce727c4d user: danielk1977 tags: trunk) | |
14:45 | Add speed3.test file. For testing performance issues related to overflow pages. (CVS 4020) (check-in: 684f765df3 user: danielk1977 tags: trunk) | |
06:44 | Modify test scripts to allow for the fact that zeroblob() is not available when compiled with OMIT_INCRBLOB. Ticket #2365. (CVS 4019) (check-in: 2ee959be4a user: danielk1977 tags: trunk) | |
2007-05-16
| ||
18:23 | Remove the SKIP_UTF16 macros (they are no longer in use). (CVS 4018) (check-in: 73e654fbdc user: danielk1977 tags: trunk) | |
18:11 | Fix handling of utf-16 encoding of code point 0xE000. (CVS 4017) (check-in: bfc35ce867 user: danielk1977 tags: trunk) | |
17:50 | Avoid passing a negative value to isspace() in a couple places. (CVS 4016) (check-in: d5db8be368 user: danielk1977 tags: trunk) | |
17:28 | Change a few selected functions to macros to speed things up. (CVS 4015) (check-in: 93f811ec74 user: danielk1977 tags: trunk) | |
14:23 | Omit some extra code when OMIT_INCRBLOB is defined. (CVS 4014) (check-in: 1d89be287d user: danielk1977 tags: trunk) | |
13:55 | Add a --nostatic option to mksqlite3c.tcl. With this option turned on, the extra "static" storage class markers are not inserted into the amalgamation. (CVS 4013) (check-in: 57e17c7cda user: drh tags: trunk) | |
11:55 | Keep the full precision of integers if possible when casting to "numeric". Ticket #2364. (CVS 4012) (check-in: 2ac985a380 user: drh tags: trunk) | |
2007-05-15
| ||
18:35 | Additional tests for malformed UTF-8. (CVS 4011) (check-in: 448d3ef670 user: drh tags: trunk) | |
16:51 | Make sure pParse->rc gets set whenever sqlite3ErrorMsg() is called. This is added insurance that parsing will stop quickly after an error. This change did make the parser stop faster in some cases, which required some revisions to tests. (CVS 4010) (check-in: f84d9dab11 user: drh tags: trunk) | |
14:40 | Remove the alternative UTF-16 implementation of SUBSTR(). It does not work anymore and the UTF-8 version can do everything that the UTF-16 version does. (CVS 4009) (check-in: 9b91502ef0 user: drh tags: trunk) | |
14:34 | Relax the restriction on using bytes 0x80 through 0xbf as the first character of an identifier. Enhancements to ALTER TABLE tests for tables with strange names or stange column names. (CVS 4008) (check-in: 262a3e6339 user: drh tags: trunk) | |
14:17 | Change the documentation to clearly state that the result of sqlite3_column_type() is undefined following a type conversion. (CVS 4007) (check-in: 960be575e2 user: drh tags: trunk) | |
14:10 | Fix to check-in (4005). A call to sqlite3_column_blob() should not make subsequent calls to sqlite3_column_type() return SQLITE_BLOB. Sqlite3_column_type() returns the initial type. (CVS 4006) (check-in: b5e85deb5a user: drh tags: trunk) | |
13:27 | Clarification on the best practices for using the _bytes() APIs. Change sqlite3_value_blob() to force the representation to be purely a BLOB and not a dual BLOB/String. Ticket #2360. (CVS 4005) (check-in: cf2dd45b58 user: drh tags: trunk) | |
11:55 | A new approach for UTF-8 translation. (CVS 4004) (check-in: 6c8ad2790e user: drh tags: trunk) | |
09:00 | Do not allow bytes 0x80 through 0xBF to be the first character of an identifer because no valid UTF-8 character can begin with those bytes. If we allowed an identifier to begin with one of those bytes, then the substr() function in ALTER TABLE will not work correctly. (CVS 4003) (check-in: 252810424d user: drh tags: trunk) | |
07:14 | Fix some test script errors in fuzz.test. (CVS 4002) (check-in: 51eeae7b7a user: danielk1977 tags: trunk) | |
07:00 | A fix for queries that used correlated, compound sub-queries in the HAVING clause. Also update fuzz.test some more. (CVS 4001) (check-in: da0c1ab4de user: danielk1977 tags: trunk) | |
03:56 | Fix problems in ALTER TABLE that occur when column or table names are invalid UTF encodings. (CVS 4000) (check-in: 9ff382f458 user: drh tags: trunk) | |
02:45 | Fix a typo in a comment. Ticket #2359. (CVS 3999) (check-in: 1fbbc108a9 user: drh tags: trunk) | |
02:34 | Fix a bug in sqlite3_mprintf() which could have caused a buffer overrun if malloc() failed. (CVS 3998) (check-in: 5af49a57d4 user: drh tags: trunk) | |
01:13 | The built-in substr() function applied to a BLOB counts bytes, not characters. (CVS 3997) (check-in: 75d573080d user: drh tags: trunk) | |
00:09 | Make sure the parser aborts quickly following a syntax error. (CVS 3996) (check-in: d07cdd3c09 user: drh tags: trunk) | |
2007-05-14
| ||
16:50 | Fix a problem with ORDER BY and compound SELECT queries. (CVS 3995) (check-in: af76928fc5 user: danielk1977 tags: trunk) | |
15:49 | Fix a bug in "flattening" optimization. Occured if the parent of the flattened sub-query is also the parent of a sub-query that uses a compound op (i.e. UNION, INTERSECT etc.). (CVS 3994) (check-in: 1c33829c9e user: danielk1977 tags: trunk) | |
14:05 | Prevent sub-queries with "LIMIT 0" from leaving an extra value on the vdbe stack. Also updates to fuzz.test. (CVS 3993) (check-in: b1d1b16e98 user: danielk1977 tags: trunk) | |
12:12 | In the windows driver, reacquire the shared lock if an exclusive lock fails. Ticket #2354. (CVS 3992) (check-in: fc489b5382 user: drh tags: trunk) | |
11:34 | Remove terms with operator TK_AS from the expression tree. Ticket #2356. (CVS 3991) (check-in: 5627ff74be user: drh tags: trunk) | |
2007-05-12
| ||
15:00 | Make REINDEX robust in the face of malloc() errors. (CVS 3990) (check-in: dbe417745d user: drh tags: trunk) | |
12:08 | Make the ANALYZE command robust in the face of malloc() failures. (CVS 3989) (check-in: c08658e1f8 user: drh tags: trunk) | |
10:41 | Fix another variant of the "IN (...)" b-tree problem. (CVS 3988) (check-in: 260338c4b2 user: danielk1977 tags: trunk) | |
09:30 | Fix an obscure b-tree bug that applied to transient trees used for IN(...) expressions. (CVS 3987) (check-in: 96c7232f8b user: danielk1977 tags: trunk) | |
06:11 | Bugfixes for assert() failures found by fuzz.test. (CVS 3986) (check-in: 2e80736d57 user: danielk1977 tags: trunk) | |
2007-05-11
| ||
16:58 | Check in some changes to fuzz.test. Still failing. (CVS 3985) (check-in: 6f346a6be9 user: danielk1977 tags: trunk) | |
12:30 | Avoid multiple storage class specifiers ("static extern") in the amalagmation. (CVS 3984) (check-in: 2f70159b1d user: drh tags: trunk) | |
10:10 | Expand the expressions tested by fuzz.test. Fix for (CAST zeroblob() AS text). (CVS 3983) (check-in: 4e1bb41f30 user: danielk1977 tags: trunk) | |
07:08 | Fix a problem with inserting zeroblob() into an indexed column. (CVS 3982) (check-in: ccef5d2daa user: danielk1977 tags: trunk) | |
01:44 | Additional parsing bugs fixed. (CVS 3981) (check-in: d12a8924c6 user: drh tags: trunk) | |
00:20 | Make sure that the REGISTER token generates a valid Expr. REGISTER will cause the tokenizer to abort, but the parser might do several reduce actions prior to that abort and those reduce actions sometimes need a valid Expr. (CVS 3980) (check-in: d146f01a02 user: drh tags: trunk) | |
2007-05-10
| ||
21:14 | When converting literal BLOBs to text using the encoding of the main database. Ticket #2349. (CVS 3975) (check-in: a57afaff42 user: drh tags: trunk) | |
17:38 | Simplify failing case in fuzz.test. (CVS 3974) (check-in: a54c74990c user: danielk1977 tags: trunk) | |
17:32 | Add cases to fuzz.test. Some tests in this file fail at the moment. (CVS 3973) (check-in: 1e8fcb3fbe user: danielk1977 tags: trunk) | |
17:23 | Add some UTF-8 test infrastructure. Treat NaN as NULL. The printf routines print infinity as "Inf" not as "NaN". Ticket #2345. (CVS 3972) (check-in: ffe615a711 user: drh tags: trunk) | |
15:37 | Add new test file fuzz.test, to test SQLite with fuzzily generated SQL. (CVS 3971) (check-in: 1e9e6fe75d user: danielk1977 tags: trunk) | |
13:23 | Fix a C++ism in func.c. (CVS 3970) (check-in: 9c0050a10c user: drh tags: trunk) | |
11:43 | Another test for overflow in strftime(). It was not broken. (CVS 3969) (check-in: 5800d0327e user: danielk1977 tags: trunk) | |
10:46 | Add code to enforce the MAX_EXPR_DEPTH limit. (CVS 3968) (check-in: 2c9c94a24d user: danielk1977 tags: trunk) | |
2007-05-09
| ||
22:56 | Make sure compound queries inside a subquery only return a single result column. Ticket #2347. (CVS 3967) (check-in: 66954bdd81 user: drh tags: trunk) | |
20:35 | Fix a typo in a comment. Ticket #2348. (CVS 3966) (check-in: c0dbac4630 user: drh tags: trunk) | |
20:31 | Fix a problem with strange expressions being fed to an ATTACH or DETACH statement. (CVS 3965) (check-in: 0c91dc9ee0 user: drh tags: trunk) | |
15:56 | Fix an obscure cache corruption that could occur after an SQLITE_FULL error. (CVS 3964) (check-in: e8e879aca6 user: danielk1977 tags: trunk) | |
11:37 | Add further test cases for compile time limits. (CVS 3963) (check-in: 9bf2c594a4 user: danielk1977 tags: trunk) | |
08:24 | Fix enforcement of the LIKE_PATTERN limit. (CVS 3962) (check-in: 8819617b7c user: danielk1977 tags: trunk) | |
2007-05-08
| ||
21:56 | Remove a keyword from the header comment in date.c (CVS 3961) (check-in: 03349ec0be user: drh tags: trunk) | |
21:45 | Remove compiler warnings in the amalgamation. (CVS 3960) (check-in: c5754530c6 user: drh tags: trunk) | |
20:59 | Disable tracing pragmas when not in debug mode. This eliminates an external dependency on stdout. (CVS 3959) (check-in: 5a0fe98548 user: drh tags: trunk) | |
20:37 | Fix the amalgamation generator so that all non-API functions have file scope. (CVS 3958) (check-in: e9f56ead05 user: drh tags: trunk) | |
18:30 | Update the release builder script. (CVS 3957) (check-in: 21299b9178 user: drh tags: trunk) | |
18:04 | Fix a null-pointer dereference that follows a malloc() failure introduced in (3954). (CVS 3956) (check-in: 251ba1981e user: danielk1977 tags: trunk) | |
17:59 | Update the amalgamation builder for the latest changes. (CVS 3955) (check-in: becd64d361 user: drh tags: trunk) | |
17:54 | Add a few more tests to sqllimit1.test. (CVS 3954) (check-in: eeee6b71e5 user: danielk1977 tags: trunk) | |
16:13 | Add a test case that uses a trigger to insert many rows to sqllimits1.test. (CVS 3953) (check-in: 6368222558 user: danielk1977 tags: trunk) | |
15:59 | Add new test file sqllimits1.test. (CVS 3952) (check-in: c897460397 user: danielk1977 tags: trunk) | |
15:46 | Do not allocate so much surplus memory in the implementation of the replace() function. (CVS 3951) (check-in: 0cf518ceeb user: drh tags: trunk) | |
15:34 | Limit the length of the patterns on LIKE and GLOB to avoid problems with deep recursion and N^2 behavior. (CVS 3950) (check-in: 42e6c82699 user: drh tags: trunk) | |
15:15 | Introduce the (experimental) sqlite3_result_error_toobig() API that function implementations can use to signal SQLite that the function result is too big to represent. (CVS 3949) (check-in: 17c4235c49 user: drh tags: trunk) | |
14:51 | Add the max_page_count pragma used to limit the maximum size of a database file. Untested. (CVS 3948) (check-in: b1b74f0668 user: drh tags: trunk) | |
14:39 | Fix a potential buffer overrun in the replace() function. (CVS 3947) (check-in: b0fb4a3cf6 user: danielk1977 tags: trunk) | |
13:58 | Add more code to enforce the limits specified in limits.h. (CVS 3946) (check-in: c59d436095 user: drh tags: trunk) | |
13:57 | Avoid evaluating (INT_MIN/-1) with 32 bit signed numbers. It can cause SIGFPE. (CVS 3945) (check-in: 2f186e916c user: danielk1977 tags: trunk) | |
12:37 | Do not crash in the internal functions sqlite_rename_table() or sqlite_rename_trigger() if they are somehow passed unexpected input. (CVS 3944) (check-in: c2f90b465e user: danielk1977 tags: trunk) | |
12:12 | Begin adding code to explicitly limit the size of strings and blobs. (CVS 3943) (check-in: 031a5915b6 user: drh tags: trunk) | |
11:27 | Fix a bug in the test logic introduced by check-in (3928). (CVS 3942) (check-in: 47d1678d2d user: drh tags: trunk) | |
01:08 | Begin implementing a new system of upper bounds on the capabilities of SQLite. By clearly defining the upper bounds, we make those upper bounds explicitly testable. This check-in is just infrastructure. The upper bounds are not yet enforced. (CVS 3941) (check-in: 93b623b232 user: drh tags: trunk) | |
2007-05-07
| ||
19:31 | Fix an NULL deref in the randomblob() function following a malloc failure. (CVS 3940) (check-in: 011e7db253 user: drh tags: trunk) | |
16:58 | Add a version of the LIKE operator to the icu extension. Requires optimisation. (CVS 3939) (check-in: 3e96105c1f user: danielk1977 tags: trunk) | |
14:58 | Change the name of create_collation_x() to create_collation_v2(). Also add some tests for it. (CVS 3938) (check-in: ddc4e4797f user: danielk1977 tags: trunk) | |
13:11 | Fix typo in Makefile.in. Ticket #2343 (CVS 3937) (check-in: db51f59a7b user: drh tags: trunk) | |
11:53 | Add interface to configure SQLite to use ICU collation functions. (CVS 3936) (check-in: b29a81b4fb user: danielk1977 tags: trunk) | |
11:24 | Change sqlite3_snprintf() so that it does not write a zero-terminator if the buffer size argument is less than 1. Ticket #2341. Added documentation about the sqlite3_snprintf() function. (CVS 3935) (check-in: f3ae4ac5fe user: drh tags: trunk) | |
09:32 | Add the experimental create_collation_x() api. (CVS 3934) (check-in: ff49d48f2f user: danielk1977 tags: trunk) | |
2007-05-06
| ||
21:20 | Fix missing word in the copyright.html web page. (CVS 3933) (check-in: 2d1348dda2 user: drh tags: trunk) | |
20:04 | Do not flatten subqueries that are part of a compound SELECT and which have both an ORDER BY and a LIMIT clause. Ticket #2339. (CVS 3932) (check-in: 9600a99804 user: drh tags: trunk) | |
16:04 | Add the start of the ICU extension. (CVS 3931) (check-in: f473e85267 user: danielk1977 tags: trunk) | |
2007-05-05
| ||
22:45 | Add "Powered By SQLite" logo contributed by Alberto Simoes. (CVS 3930) (check-in: 25cfa7740a user: drh tags: trunk) | |
18:49 | Fix a header problem introduced in the previous check-in. (CVS 3929) (check-in: bece7a6bae user: drh tags: trunk) | |
18:39 | Factor out btree test logic into a separate test_btree.c file. (CVS 3928) (check-in: d51274f1cc user: drh tags: trunk) | |
12:06 | Fix a sprintf() problem introduced by (3916). (CVS 3927) (check-in: 54a1d275aa user: drh tags: trunk) | |
11:48 | Refactoring. Split btreeInt.h off from btree.c. Split malloc.c off from util.c. Expect much more to follow. (CVS 3926) (check-in: ad549a40ed user: drh tags: trunk) | |
11:48 | Refactoring. Split btreeInt.h off from btree.c. Split malloc.c off from util.c. Expect much more to follow. (CVS 3925) (check-in: 16041116aa user: drh tags: trunk) | |
2007-05-04
| ||
19:16 | Add documentation on the incremental vacuum feature. (CVS 3924) (check-in: b3e2cdaed3 user: drh tags: trunk) | |
19:03 | Fix compilation and testing with OMIT_INCRBLOB defined. (CVS 3923) (check-in: a0f8adc692 user: danielk1977 tags: trunk) | |
18:36 | Test interaction of incremental io and other database writes. (CVS 3922) (check-in: 4516416b4d user: danielk1977 tags: trunk) | |
18:30 | Change incremental vacuum to be triggered by a pragma rather than a command. We have a lot to learn about this yet and we do not want to paint ourselves into a corner by commiting to specific syntax too early. (CVS 3921) (check-in: b13e497a32 user: drh tags: trunk) | |
17:07 | Tighter compression of the keyword hash table. (CVS 3920) (check-in: 681216767d user: drh tags: trunk) | |
16:14 | Optional parameter in the INCREMENTAL VACUUM statement specifies how many pages to vacuum from the database. (CVS 3919) (check-in: ed713f9ccb user: drh tags: trunk) | |
14:36 | Get tests working with SQLITE_OMIT_AUTOVACUUM=1. (CVS 3918) (check-in: b94d39e4d9 user: drh tags: trunk) | |
14:14 | Make "incremental" a fallback token so that you can set "auto_vacuum=incremental" without having to quote the "incremental". (CVS 3917) (check-in: 9d7a2e2bb4 user: drh tags: trunk) | |
13:15 | Eliminate all uses of sprintf() and strcpy(). These were not being misused. But getting rid of them removes a library dependency. And it avoids warnings from the OpenBSD compiler. Ticket #2336. (CVS 3916) (check-in: ba4845b32b user: drh tags: trunk) | |
12:05 | Test cases and corrections to IO and malloc() error handling in incremental blob IO functions. (CVS 3915) (check-in: 641e55284e user: danielk1977 tags: trunk) | |
12:01 | Fix a formatting errror in I/O logging. (CVS 3914) (check-in: 2d37687a08 user: drh tags: trunk) | |
11:59 | Add compile-time option -DSQLITE_MIXED_ENDIAN_64BIT_FLOAT=1 that uses mixed-endian doubles. This is needed on ARM7 to make database file formats compatible with all other processors. Tickets #2278 and #2335. (CVS 3913) (check-in: 2a178d0c79 user: drh tags: trunk) | |
08:32 | Rework accessPayload() function for clarity. (CVS 3912) (check-in: 42d07c70ec user: danielk1977 tags: trunk) | |
2007-05-03
| ||
20:06 | fix from Gentoo for allowing TCLLIBDIR to be overridden in env (CVS 3911) (check-in: e54a49e264 user: vapier tags: trunk) | |
18:14 | Test some more incremental IO error cases. (CVS 3910) (check-in: 64705410bd user: danielk1977 tags: trunk) | |
17:18 | Improvements to I/O tracing output. Require -DSQLITE_ENABLE_IOTRACE when compiling shell.c in order to link in the I/O tracing capability. (CVS 3909) (check-in: 8a43e1676f user: drh tags: trunk) | |
16:55 | Get the amalgamation builder working with incremental I/O. (CVS 3908) (check-in: 92b5360165 user: drh tags: trunk) | |
16:31 | Test cases and minor bugfixes for incremental blob APIs. (CVS 3907) (check-in: e12c522383 user: danielk1977 tags: trunk) | |
13:11 | Fix a bug where accessPayload() was calling PagerWrite() on the wrong page handle. Ticket #2332. (CVS 3906) (check-in: cf9eeba7be user: danielk1977 tags: trunk) | |
13:02 | Use memmove() instead of memcpy() when moving between memory regions that might overlap. Ticket #2334. (CVS 3905) (check-in: 678d672b73 user: drh tags: trunk) | |
11:43 | Minor bugfixes for incrblob mode. (CVS 3904) (check-in: b84d597c90 user: danielk1977 tags: trunk) | |
11:43 | Minor bugfixes for incrblob mode. (CVS 3903) (check-in: db54a9466e user: danielk1977 tags: trunk) | |
2007-05-02
| ||
17:54 | Allow CREATE TABLE to occur while other queries are running. DROP TABLE is still prohibited, however, since we do not want to delete a table out from under an running query. (CVS 3902) (check-in: 5b4bf1fce4 user: drh tags: trunk) | |
17:48 | Combine the internal btree functions BtreePutData() and getPayload(). (CVS 3901) (check-in: a100a5304b user: danielk1977 tags: trunk) | |
16:51 | More fixes and improvements to the zeroblob() mechanism. (CVS 3900) (check-in: 83ab25014e user: drh tags: trunk) | |
16:48 | Cache the location of overflow pages in cursors used for incremental blob IO. (CVS 3899) (check-in: 349f1ea789 user: danielk1977 tags: trunk) | |
15:36 | Fix an invalid UTF8 encoding in the tests for the trim function. (CVS 3898) (check-in: 4dbbfff4a7 user: drh tags: trunk) | |
13:30 | Add support for zero-blobs to the OP_MakeRecord opcode. First test cases of zeroblob functionality. (CVS 3897) (check-in: e6d560ddee user: drh tags: trunk) | |
13:16 | Use the pointer-map pages to make the incremental blob API more efficient. (CVS 3896) (check-in: 93a3bf71d5 user: danielk1977 tags: trunk) | |
02:08 | Fix some compiler warnings. Add the (untested) zeroblob() SQL function. (CVS 3895) (check-in: 6f4f8ba7ec user: drh tags: trunk) | |
01:34 | Begin adding the zeroblob API to support incremental blob i/o. (CVS 3894) (check-in: 7a01836dde user: drh tags: trunk) | |
2007-05-01
| ||
18:25 | Enable prefix-search in query-parsing and snippet generation. If the character immediately after the end of a term is '*', that term is marked for prefix matching. Modify term comparison in snippetOffsetsOfColumn() to respect isPrefix. fts2n.test runs prefix searching through some obvious test cases. (CVS 3893) (check-in: 7c4c659240 user: shess tags: trunk) | |
17:49 | First approximation of incremental blob IO API. (CVS 3892) (check-in: c444836e7b user: danielk1977 tags: trunk) | |
17:14 | Modify loadSegmentLeavesInt() to correctly handle prefix searching. The new function docListUnion() is used to accumulate a union of the hits for the matching terms, which will be merged across segments using docListMerge(). (CVS 3891) (check-in: 72c7963073 user: shess tags: trunk) | |
16:59 | The pager takes the sector size to be the larger of the sector size reported by sqlite3OsSectorSize() and the page size. (CVS 3890) (check-in: e5e6af55cc user: drh tags: trunk) | |
2007-04-30
| ||
22:09 | Propagate prefix flag through implementation of doclist query code. Also implement correct prefix-handling for traversal of interior nodes of segment tree. A given prefix can span multiple children of an interior node, and from there the branches need to be followed in parallel. (CVS 3889) (check-in: cae844a01a user: shess tags: trunk) | |
21:39 | Fix a potential segfault following a malloc() failure during a call to sqlite3_prepare() where the nBytes parameter is positive but less than the length of the input SQL string. (CVS 3888) (check-in: 27bf3fc3cf user: drh tags: trunk) | |
17:52 | Lift docListMerge() call out of loadSegmentLeavesInt() for prefix search. Doclists from multiple prefix matches will need a union merge function, which will have to logically happen across a segment before doclists are merged between segments. (CVS 3887) (check-in: 7ddb826689 user: shess tags: trunk) | |
16:55 | Try to avoid reading pages when moving overflow chains to the free-list. (CVS 3886) (check-in: 8cccec68bd user: danielk1977 tags: trunk) | |
2007-04-28
| ||
15:47 | Add some tests (and 2 resulting bug fixes) to incr vacuum mode. (CVS 3885) (check-in: 89b1b3f897 user: danielk1977 tags: trunk) | |
2007-04-27
| ||
22:02 | Break interior-node and leaf-node readers apart in loadSegment(). Previously, the code looped until the block was a leaf node as indicated by a leading NUL. Now the code loops until it finds a block in the range of leaf nodes for this segment, then reads it using LeavesReader. This will make it easier to traverse a range of leaves when doing a prefix search. (CVS 3884) (check-in: 9466367d65 user: shess tags: trunk) | |
21:59 | Internationalize the TRIM functions. Ticket #2323. (CVS 3883) (check-in: ff1f4e7447 user: drh tags: trunk) | |
21:24 | Lift code to traverse interior nodes out of loadSegment(). Refactoring towards prefix searching. (CVS 3882) (check-in: 25935db738 user: shess tags: trunk) | |
21:02 | Refactor fts2 loadSegmentLeaf() in preparation for prefix-searching. Prefix-searching will want to accumulate data across multiple leaves in the segment, using LeavesReader instead of LeafReader is the first step in that direction. (CVS 3881) (check-in: 22ffdae4b6 user: shess tags: trunk) | |
17:16 | Make sure sqlite3_value_bytes() does not reformat the content after a call to sqlite3_value_blob(). Add documentation to explain this hazard. Add many new tests. Ticket #2321. (CVS 3880) (check-in: e92bd97a37 user: drh tags: trunk) | |
07:55 | Extra tests for incremental vacuum. (CVS 3879) (check-in: 40ba6493e9 user: danielk1977 tags: trunk) | |
07:05 | Modifications to btree integrity check so that it can be run inside a transaction after an incr vacuum. (CVS 3878) (check-in: 4d4180d647 user: danielk1977 tags: trunk) | |
01:18 | The replace() function should return NULL if the second argument is an empty string. Ticket #2324. (CVS 3877) (check-in: e6a0c90dd9 user: drh tags: trunk) | |
2007-04-26
| ||
14:42 | Add largely untested code for the incremental vacuum function. (CVS 3876) (check-in: f6a6d2b887 user: danielk1977 tags: trunk) | |
12:11 | In the pager, load the content of pages which were initialized with noContent==1 if they are subsequently requested with noContent==0. (CVS 3875) (check-in: d0745a43b6 user: drh tags: trunk) | |
2007-04-25
| ||
18:23 | Fix segfaults that can occur if a malloc failure happens just before a built-in function calls sqlite3_value_text(). (CVS 3874) (check-in: 9cb0ed6ee9 user: drh tags: trunk) | |
15:42 | Disable tests in misc7 that do not work on windows due to limitations of the windows file system. (CVS 3873) (check-in: 66646d6fda user: drh tags: trunk) | |
12:07 | Version 3.3.17 (CVS 3872) (check-in: 16979f4525 user: drh tags: trunk) | |
12:06 | Fix a file descriptor leak in the exclusive2 test script. (CVS 3871) (check-in: 2d2e68da74 user: drh tags: trunk) | |
11:42 | Preparing for the release of 3.3.17. (CVS 3870) (check-in: e278c4ef60 user: drh tags: trunk) | |
11:32 | Fix duplicate test IDs in the test suite. No changes to code. Ticket #2319. (CVS 3869) (check-in: 0935cdf82a user: drh tags: trunk) | |
11:28 | Fix a bug in the sqlite3_changes() function reported on the mailing list. (CVS 3868) (check-in: 58ea768c3e user: drh tags: trunk) | |
2007-04-24
| ||
17:35 | This fixes a missed case in check-in (3866). Do not apply patch (3866) without also applying this patch. (CVS 3867) (check-in: 66c2fa0836 user: drh tags: trunk) | |
17:27 | When the write_version flag in the database header is larger than what the library understands, make the database read-only. The old behavior was to make the database unreadable. (CVS 3866) (check-in: 10648e9992 user: drh tags: trunk) | |
2007-04-23
| ||
23:56 | Get the build working with -DSQLITE_OMIT_UTF16=1. Ticket #2314. (CVS 3865) (check-in: 25c1c7aaa8 user: drh tags: trunk) | |
2007-04-21
| ||
16:37 | Add the ability to turn the FTS2 module into an amalgamation. (CVS 3864) (check-in: 94374654cc user: drh tags: trunk) | |
2007-04-20
| ||
12:22 | Remove the ARRAYSIZE macro from where.c to avoid a name conflict with microsoft header files. Ticket #2311 (CVS 3863) (check-in: 9389a15f35 user: drh tags: trunk) | |
2007-04-19
| ||
18:36 | Fix bug in fts2 handling of OR queries. When one doclist ends before the other, the code potentially tries to read past the end of the doclist. http://www.sqlite.org/cvstrac/tktview?tn=2309 (CVS 3862) (check-in: dfac6082e8 user: shess tags: trunk) | |
14:48 | Better fix than (3860) for the same problem. (3860) could leave file-handles open in some circumstances. (CVS 3861) (check-in: 5ad645339b user: danielk1977 tags: trunk) | |
14:28 | Fix a crash that can occur if an error happens in a virtual table xSync() function. (CVS 3860) (check-in: d1afdd8c9c user: danielk1977 tags: trunk) | |
12:30 | Comment changes in the ioerr tests. No changes to code. (CVS 3859) (check-in: b7ed0e1e88 user: drh tags: trunk) | |
11:09 | Fix a bug whereby sqlite3_prepare_v2() could return both an out-of-memory error and a valid statement handle. (CVS 3858) (check-in: 8795d11c3c user: danielk1977 tags: trunk) | |
00:24 | Get the build working with SQLITE_OMIT_SHARED_CACHE. Ticket #2307. (CVS 3857) (check-in: b623538c55 user: drh tags: trunk) | |
2007-04-18
| ||
17:07 | Fix a && where & was intended. This is a real problem, but it would be very difficult to devise a test case where it might lead to an unexpected result. Ticket #2306. (CVS 3856) (check-in: 7f9f0444ed user: drh tags: trunk) | |
17:04 | Better handle malloc() failures in the "echo" virtual table test module. (CVS 3855) (check-in: 6881f9212f user: danielk1977 tags: trunk) | |
16:45 | The sqlite3_step() routine reports SQLITE_ERROR after a commit failure, not SQLITE_DONE. This fix only effects the legacy sqlite3_prepare() interface. It already worked correctly with sqlite3_prepare_v2(). (CVS 3854) (check-in: 165dc72d70 user: drh tags: trunk) | |
15:33 | Version 3.3.16 (CVS 3853) (check-in: 8c6b5adb5c user: drh tags: trunk) | |
14:47 | Fix an assertion failure caused by (3851). (CVS 3852) (check-in: 24683f1d29 user: danielk1977 tags: trunk) | |
14:24 | Minor changes (mainly to assert() macros) to handle malloc failure in vtab operations. (CVS 3851) (check-in: 175156d1fc user: danielk1977 tags: trunk) | |
13:49 | Bump the version number and update documents in preparation for 3.3.16. (CVS 3850) (check-in: 852100a332 user: drh tags: trunk) | |
2007-04-17
| ||
08:32 | Avoid reloading the db schema after a failed OP_VerifyCookie if the in-memory schema cookie already matches the database file. (CVS 3849) (check-in: 61c1d06d10 user: danielk1977 tags: trunk) | |
2007-04-16
| ||
17:07 | Fix a memory leak introduced by (3842) associated with ticket #2296. (CVS 3848) (check-in: cdc7608b8d user: drh tags: trunk) | |
15:49 | Add test cases to make sure virtual tables cannot be used in shared-cache mode. (CVS 3847) (check-in: 66e468adfc user: danielk1977 tags: trunk) | |
15:35 | Make it clear in the docs that virtual tables and shared cache do not play well together. (CVS 3846) (check-in: 62ef2b1127 user: drh tags: trunk) | |
15:06 | Ensure sqlite3_finalize() can be called from within the xDisconnect() method of virtual tables. (CVS 3845) (check-in: 8d6c3bfc4d user: danielk1977 tags: trunk) | |