Many hyperlinks are disabled.
Use anonymous login
to enable hyperlinks.
688 check-ins using file src/func.c version 8f72e88c
|
2022-10-17
| ||
| 16:29 | Fix a potential call to strlen() with a null argument in the command-line shell following an OOM error. forum post 9c4f2ebe22. (check-in: 67bb8eda user: drh tags: branch-3.39) | |
|
2022-10-14
| ||
| 15:46 | Handle recovering databases with reserved bytes at the end of each page. (check-in: 52d0235e user: dan tags: recover-extension) | |
| 13:26 | Remove a resolved TODO comment. (check-in: 32fd4ac3 user: stephan tags: trunk) | |
| 02:00 | Cancel IndexExpr objects when they go out of scope. (check-in: 09635193 user: drh tags: index-expr-opt) | |
|
2022-10-13
| ||
| 21:08 | This experimental branch attempts to use columns for an index-on-expression in place of the expression that is being indexed. This particular check-in mostly works, but there are still issues. (check-in: 2e8d4fd4 user: drh tags: index-expr-opt) | |
| 20:06 | Add tests for detecting page size of databases. (check-in: 31f9e236 user: dan tags: recover-extension) | |
| 16:48 | Move the rest of testing1.js into tester1.js and eliminate the dependency on jaccwabyt_test.c. Extend the list of default config-related #defines in sqlite3-wasm.c and reorganize them for maintainability. (check-in: 4e2a8aff user: stephan tags: trunk) | |
| 15:09 | Optimize the IS NULL and IS NOT NULL operators so that they avoid loading large strings or blobs off of disk if all it needs to know is whether or not the string or blob is NULL. (check-in: cb943501 user: drh tags: trunk) | |
| 14:54 | Improvements to the description of the OPFLAG_TYPEOFARG option to OP_Column. (Closed-Leaf check-in: 5e9c67ba user: drh tags: isnull-opt) | |
| 14:35 | Adjust the implementation of sqlite3VdbeTypeofColumn() to make it easier to test (and slightly smaller). (check-in: 79fdd021 user: drh tags: isnull-opt) | |
| 14:01 | Fix a typo in the documentation of the OP_Column opcode. Forum post a2b5bd6d43. (check-in: 043e76e6 user: drh tags: trunk) | |
| 12:47 | Proposed optimization to the IS NULL and NOT NULL operators that avoids loading the entire content of larges strings and BLOBs. Response to forum post 3c08d4715dc05b00. (check-in: 45f17156 user: drh tags: isnull-opt) | |
| 08:03 | Port the first 180-odd unit tests from testing1.* into the new tester1.*. Fix a stray-keystroke-induced typo which broke pstack.allocChunks(). (check-in: ef689e33 user: stephan tags: trunk) | |
|
2022-10-12
| ||
| 18:40 | Updates to the fuzzer query invariant checker - tracking changes made over in dbsqlfuzz. (check-in: 4ca16a30 user: drh tags: trunk) | |
| 18:30 | Changes to Makefile.msc to fix the build on Windows following the previous merge. (check-in: 368fa6b2 user: drh tags: trunk) | |
| 18:13 | Merge fiddle-opfs branch into trunk. (check-in: 98f95494 user: stephan tags: trunk) | |
| 16:35 | Merge trunk into fiddle-opfs as a preliminary step for a clean merge in the other direction. (Closed-Leaf check-in: 5e1848ce user: stephan tags: fiddle-opfs) | |
| 15:54 | Add initial infrastructure for setting up function/regression tests for the JS/WASM APIs. (check-in: 7f5db982 user: stephan tags: fiddle-opfs) | |
| 15:40 | Minor doc cleanups and corrections in sqlite3-wasm.c (check-in: 5144c122 user: stephan tags: fiddle-opfs) | |
| 14:39 | Correct mismatched H1 tags in test code. Minor CSS tweaks. (check-in: 4d8eb90a user: stephan tags: fiddle-opfs) | |
| 12:49 | Add the sqlite3_value_encoding() interface. (check-in: d6d44997 user: drh tags: trunk) | |
|
2022-10-11
| ||
| 13:57 | Enhance the OP_IsType opcode so that it is slightly smaller and faster and so that it works correctly with invalid serial-type codes 10 and 11. (check-in: 846f863e user: drh tags: trunk) | |
| 12:02 | Improved the ability of the CLI to handle very long input lines. Potentially a fix for the bug reported by forum post fa4bb2941a. (check-in: d0e107ee user: drh tags: trunk) | |
|
2022-10-10
| ||
| 23:54 | Enhance the "PRAGMA integrity_check" statement so that it verifies datatype constraints on non-STRICT tables: (1) Columns with TEXT affinity should not contain numeric values, and (2) columns with NUMERIC affinity should not contain TEXT values that could be converted into numbers. (check-in: aa6e9086 user: drh tags: trunk) | |
| 21:21 | Code clean-up for the integrity_check enhancement. (Closed-Leaf check-in: a1401731 user: drh tags: integrity_check_datatypes) | |
| 19:38 | Fix corner-case bugs in the new integrity_check logic. All tests pass now. (check-in: dbab9d52 user: drh tags: integrity_check_datatypes) | |
| 18:25 | An attempt to enhance PRAGMA integrity check so that it does data type checking on non-STRICT tables. Specifically: (1) Columns with TEXT affinity should not contain numeric values, and (2) columns with numeric affinity should not contain text values that can be converted to numeric. (check-in: 8b1e7f05 user: drh tags: integrity_check_datatypes) | |
| 12:10 | Improved detection of database corrupting when moving pages on an autovacuumed database when creating a new btree. dbsqlfuzz 9a55397eae13cec64aebf1fd35489c3a90bdaac5 (check-in: 835e1912 user: drh tags: branch-3.39) | |
| 12:02 | Improved detection of database corrupting when moving pages on an autovacuumed database when creating a new btree. dbsqlfuzz 9a55397eae13cec64aebf1fd35489c3a90bdaac5 (check-in: 327965bc user: drh tags: trunk) | |
|
2022-10-09
| ||
| 17:08 | Remove some extraneous debug output. (check-in: 56ff4205 user: stephan tags: fiddle-opfs) | |
| 17:03 | Remove a now-unnecessary --cachesize flag for the OPFS sqlite3_vfs speedtest1-worker.html invocation. (check-in: 08b516d6 user: stephan tags: fiddle-opfs) | |
| 15:12 | Minor cleaups in the post-init async phase of sqlite3 module initialization. (check-in: 2e024a6b user: stephan tags: fiddle-opfs) | |
| 13:35 | Remove ext/wasm/kvvfs.make which was inadvertently resurrected in [92b500da70a3]. (check-in: 2e7fa085 user: stephan tags: fiddle-opfs) | |
| 13:33 | Cherrypick [ea370b9b05f7ed7eaa] and part of [92b500da] into kv-vfs branch. (Leaf check-in: 81263803 user: stephan tags: kv-vfs) | |
| 13:26 | Refactor kvvfs JS bits to make use of [ea370b9b05f7ed7eaa]. At main-thread startup, if kvvfs is available, replace the kvvfs I/O methods with JS impls. Checkin part 2 of 2, to account for cherrypicking [ea370b9b05f7ed7eaa] into the kv-vfs branch. (check-in: a9047e02 user: stephan tags: fiddle-opfs) | |
| 13:19 | Refactor os_kv.c so that the kvvfs read/write/delete methods can be swapped out at runtime by JS implementations. This eliminates the kvvfs dependency on Emscripten. Checkin part 1 of 2, to account for cherrypicking. (check-in: ea370b9b user: stephan tags: fiddle-opfs) | |
| 11:42 | Add -DSQLITE_OS_KV_OPTIONAL to speedtest1 flags for kvvfs testing purposes. (check-in: 92b500da user: stephan tags: fiddle-opfs) | |
|
2022-10-08
| ||
| 17:27 | When casting an odd-length BLOB into a TEXT with encoding UTF16, omit the last byte. dbsqlfuzz c5df68b5bbdf6c9f3c1851e41e88f49ac05c9969 (check-in: 1e2796b3 user: drh tags: trunk) | |
|
2022-10-07
| ||
| 19:11 | Fix a problem causing the seek-scan optimization to skip over valid rows that could occur when it is used with expressions of the form (a IN (?,?..) AND b >= ?). dbsqlfuzz ab1db6dc0efb04cba1cd3431ee6da4894fdc4520. (check-in: b6be4ce6 user: drh tags: branch-3.39) | |
| 18:57 | Fix a problem causing the seek-scan optimization to skip over valid rows that could occur when it is used with expressions of the form (a IN (?,?..) AND b >= ?). dbsqlfuzz ab1db6dc0efb04cba1cd3431ee6da4894fdc4520. (check-in: 63d9efe2 user: dan tags: trunk) | |
| 18:06 | Fix a problem that occurred when using a bloom filter to optimize an (ipk = ?) lookup in the case where the RHS of the expression is a TEXT value. First reported by forum post f61a8b7053. (check-in: dc01d9d8 user: drh tags: branch-3.39) | |
| 17:58 | Increase the version number to 3.39.5. (check-in: 8eb0048e user: drh tags: branch-3.39) | |
| 15:55 | Improved tracing output for showing all registers used by the OP_SeekGE opcode. This applies to debugging builds only. (check-in: 0aa6dee7 user: drh tags: trunk) | |
|
2022-10-06
| ||
| 21:00 | Ensure the page-size, auto-vacuum and other settings are copied into the recovered database. (check-in: 078520f2 user: dan tags: recover-extension) | |
| 17:20 | Add code to determine the database page-size by searching for well-formed pages. (check-in: 0dbd0cce user: dan tags: recover-extension) | |
| 14:10 | Fix a problem that occurred when using a bloom filter to optimize an (ipk = ?) lookup in the case where the RHS of the expression is a TEXT value. First reported by forum post f61a8b7053. (check-in: 8e14c351 user: dan tags: trunk) | |
|
2022-10-05
| ||
| 16:58 | Improve handling of corruption in the sqlite_schema table b-tree structure. (check-in: 2be0dba1 user: dan tags: recover-extension) | |
|
2022-10-04
| ||
| 19:47 | Merge latest trunk changes. (check-in: b0feecaa user: dan tags: recover-extension) | |
| 19:43 | Better handling for databases with corrupt headers. (check-in: 17f68d80 user: dan tags: recover-extension) | |
| 17:06 | Replace time-based auto-unlock of opfs sync handles with lock acquisition/release via sqlite3_io_methods::xLock/xUnlock(). (check-in: 2625b7cf user: stephan tags: fiddle-opfs) | |
| 14:50 | Attempt to clarify the operation of the xLock and xUnlock VFS I/O methods. Assert() statements added to prove that they behave as the documentation says. (check-in: 3efa8112 user: drh tags: trunk) | |
| 11:14 | Tweaks to the opfs async wait/relinquish times. (check-in: 35f33c23 user: stephan tags: fiddle-opfs) | |
| 10:35 | Change the name of the autoconf-generated configuration file from "config.h" to "sqlite_cfg.h" to work around a bug in a dependency checker, reported in a tweet from Joey Castillo. (check-in: 897e187c user: drh tags: trunk) | |
| 09:12 | Update ext/wasm/api/README.md to account for recent changes. (check-in: 9d488081 user: stephan tags: fiddle-opfs) | |
| 09:12 | OPFS async proxy: add a wait-and-retry policy to the get-sync-handle step to help account for cross-tab locking. (check-in: 45c48c63 user: stephan tags: fiddle-opfs) | |
| 08:58 | Update the 'start in speedtest kvvfs' link in speedtest1-worker to account for recent changes. (check-in: f4d5ba41 user: stephan tags: fiddle-opfs) | |
| 01:21 | Minor pstack doc update. (check-in: 6b4fb7c4 user: stephan tags: fiddle-opfs) | |
| 01:11 | Minor cleanups and additions in sqlite3.capi.wasm.pstack. (check-in: 97bd670d user: stephan tags: fiddle-opfs) | |
| 00:54 | Add a test/debug mechanism to shut down the OPFS async listener so that it can be inspected (it normally can't be because its tight event-listening loop ties up the thread) and then restarted. (check-in: 7d0bcff4 user: stephan tags: fiddle-opfs) | |
|
2022-10-03
| ||
| 23:13 | Set default page cache size to 16mb in wasm builds. Fix an off-by-one counter in sqlite3_wasm_enum_json(). Minor coding style conformance tweaks. (check-in: 72a9e589 user: stephan tags: fiddle-opfs) | |
| 22:51 | Add a --cachesize flag to the speedtest1-worker-opfs link in index.html because opfs is much faster with that. (check-in: 5b8f8e33 user: stephan tags: fiddle-opfs) | |
| 22:38 | Minor JS doc updates and typo fixes. (check-in: 3cfcc14d user: stephan tags: fiddle-opfs) | |
| 18:23 | In os_kv.c, every xWrite method call on a database should record the page size. (check-in: e9411c74 user: drh tags: kv-vfs) | |
| 18:07 | Add configurable cache_size to batch-runner and speedtest1-worker. Add SQL tracing to demo-123.js just for demonstration's sake. (check-in: 2ab06513 user: stephan tags: fiddle-opfs) | |
| 18:05 | Omit WAL mode from the wasm build. (check-in: 88efe2a6 user: stephan tags: fiddle-opfs) | |
| 15:22 | Adjust the --help output for speedtest1 to clarify that the argument to the --cachesize option is in pages, not bytes. (check-in: d7c2be6b user: drh tags: trunk) | |
| 14:01 | Merge the speedtest1 enhancements from trunk (and fiddle-opfs) into the kv-vfs branch. (check-in: 3d350aa0 user: drh tags: kv-vfs) | |
| 13:56 | Copy all of the fiddle-opfs changes to speedtest1 into trunk. (check-in: f3d31c99 user: drh tags: trunk) | |
| 13:46 | Resolve/remove a TODO. (check-in: 2ea35fea user: stephan tags: fiddle-opfs) | |
| 13:45 | Add the --vfs option to speedtest1. (check-in: aeb88402 user: drh tags: trunk) | |
| 13:24 | Merge the latest trunk enhancements into the kv-vfs branch. (check-in: f4aefe63 user: drh tags: kv-vfs) | |
| 13:03 | Export sqlite3_trace_v2() to wasm and use it to ensure that the new per-VFS post-open SQL support in the DB ctor works. Default opfs vfs to journal_mode=truncate, as it's faster in that mode. Add 't' DB open-mode flag to enable SQL tracing to console.log(). (check-in: 508f7f6d user: stephan tags: fiddle-opfs) | |
| 11:42 | Reimplement fiddle_reset_db() so that it works with all VFSes. (check-in: 18462052 user: stephan tags: fiddle-opfs) | |
| 11:33 | Experimentally relinquish the OPFS VFS sync access handle when the db is idle and reacquire it on demand, the goal being to help alleviate cross-tab locking issues. (check-in: 2703ac98 user: stephan tags: fiddle-opfs) | |
| 11:23 | Add journal=MODE to the list of supported URL flags for speedtest1-worker.html. (check-in: 5c43e8d2 user: stephan tags: fiddle-opfs) | |
| 09:21 | OPFS VFS now lazily opens its sync access handle, as a step towards experimenting with relinquishing it during idle times to help avoid cross-tab and page-reload locking issues. (check-in: a984e1ba user: stephan tags: fiddle-opfs) | |
| 08:30 | Partial revert of [a82e6faaa642] to remove the 'I' alias for 'j' wasm function signature letter. In hindsight, that change seems premature. (check-in: dcd46af9 user: stephan tags: fiddle-opfs) | |
| 08:21 | Minor JS API tweaks prompted by documenting them. (check-in: a82e6faa user: stephan tags: fiddle-opfs) | |
|
2022-10-02
| ||
| 22:50 | More cleanups in the UDF argument and result handling, in particular int64. Consolidate some duplicate int64/bigint range checking code. Expose the UDF low-level utilities (arg/result conversion) to client code. Add the sqlite3_context pointer to the JS-side UDF wrappers for API consistency. (check-in: 10ab77af user: stephan tags: fiddle-opfs) | |
| 20:13 | Minor doc typo fix in sqlite.h.in. (check-in: 372802aa user: stephan tags: trunk) | |
| 20:08 | JS: clean up create_function() wrapper and add support for create_window_function(). Eliminate an extraneous blob copy when a UDF returns a blob. Make use of newfound JS-fu to clean up how sqlite3ApiBootstrap() config is initialized. (check-in: d3bad934 user: stephan tags: fiddle-opfs) | |
| 18:47 | js: implement a hand-written wrapper for sqlite3_create_function_v2() which converts, if necessary, JS-function-type args to WASM function wrappers. Replace DB.createFunction() impl with the new one. (check-in: 435ab333 user: stephan tags: fiddle-opfs) | |
| 03:14 | Doc typo fixes. (check-in: e528675d user: stephan tags: fiddle-opfs) | |
| 03:11 | More fleshing out of sqlite3.capi.wasm.pstack. (check-in: eb572667 user: stephan tags: fiddle-opfs) | |
| 01:48 | Document the roles of the new (this past week) JS files added to the build process. (check-in: 8b3bc731 user: stephan tags: fiddle-opfs) | |
| 00:09 | General internal cleanups in the oo1 API. (check-in: f9db664f user: stephan tags: fiddle-opfs) | |
|
2022-10-01
| ||
| 20:27 | Performance optimization in sqlite3_prepare() (discovered while working on the stmt-cache branch). (check-in: cd494240 user: drh tags: trunk) | |
| 19:21 | Correct fiddle db export breakage caused by a post-testing API change made in [1fa019c88dea]. (check-in: e73cc44e user: stephan tags: fiddle-opfs) | |
| 18:55 | Pedantic constness tweak. (check-in: c8a173cf user: stephan tags: fiddle-opfs) | |
| 18:47 | wasm: correct a memleak caused by a shadowed var in the previous checkin. Add a stack-like allocator, sqlite3.capi.wasm.pstack, as a faster way of managing short-lived pointers (like the one which got shadowed). (check-in: 1fa019c8 user: stephan tags: fiddle-opfs) | |
| 16:01 | Fiddle: fix makefile dependency issue and duplicate inclusion of post-js.js. Reimplement db export using sqlite3_serialize(). (check-in: 29db7de7 user: stephan tags: fiddle-opfs) | |
| 13:45 | Merge trunk into fiddle-opfs branch. (check-in: 64ebcbe4 user: stephan tags: fiddle-opfs) | |
| 13:38 | Tweak sqlite3-worker1.js to be able to load either sqlite3.js or sqlite3-wasmfs.js, noting that the latter still does not load in a Worker because of an Emscripten loader bug. (check-in: 000ef705 user: stephan tags: fiddle-opfs) | |
| 13:28 | Merge the Parse.prepFlags change from trunk. (Leaf check-in: c8d77f39 user: drh tags: stmt-cache) | |
| 13:17 | Replace the Parse.disableVtab field with Parse.prepFlags for increased generality, a small size reduction, and a small performance increase. (check-in: b7da0bcd user: drh tags: trunk) | |
| 12:04 | Note two wasm-related potential TODOs. (check-in: 5636e828 user: stephan tags: fiddle-opfs) | |
| 12:00 | Merge the latest trunk enhancements into the stmt-cache branch. (check-in: 7812aea6 user: drh tags: stmt-cache) | |
|
2022-09-30
| ||
| 23:49 | Remove a couple of weird quirks of the Worker1 and Promiser APIs. The Worker1 (now undocumented) unlink capability needs to be reevaluated to work equivalently for all storage backends. (check-in: 9f2b331a user: stephan tags: fiddle-opfs) | |
| 23:02 | Tweaks to the Worker1 and Promiser APIs prompted by documenting them. (check-in: c68b9aa1 user: stephan tags: fiddle-opfs) | |
| 22:40 | Remove one of the NEVER() macros added by [5166acf3f0040459] because it is reachable after all. (check-in: bd8ae5fe user: drh tags: trunk) | |
| 20:59 | New test cases in fuzzdata8.db (check-in: aa367a3c user: drh tags: trunk) | |
| 20:53 | Fix a bug in the demoDelete fix from [d76a301e22e05722]. (check-in: 7e761a02 user: drh tags: trunk) | |
| 20:48 | Add NEVER() to two error-detection tests that were made unreachable by the previous check-in. (check-in: 5166acf3 user: drh tags: trunk) | |
| 20:35 | Add JS wrapper for sqlite3_exec() which knows how to handle a JS callback. Add some console.error() reporting of module-load failures, as they otherwise often get silently swallowed up by the loader's mechanisms. Add 'flexible-string' JS-to-WASM argument converter which performs more X-to-string conversions than the 'string' arg converter does. (check-in: 96818aa8 user: stephan tags: fiddle-opfs) | |
| 20:15 | Better handle a case of database corruption where a b-tree page is one of its own ancestor pages. (check-in: a24e4680 user: dan tags: trunk) | |
| 16:49 | Add sqlite3.version object. Add more state to the Worker #1 config-get response, including sqlite3.version. (check-in: 711f458d user: stephan tags: fiddle-opfs) | |
| 15:46 | Add oo1.JsStorageDb.clearStorage/storageSize() methods, copies of capi.sqlite3_web_kvvfs_clear/size(). (check-in: 1e09efe7 user: stephan tags: fiddle-opfs) | |
| 15:24 | batch-runner.js: force WebSQL batches to not abort for a failed statement (necessary for apples-to-apples-ish benchmark comparisons). (check-in: 60f0c5cb user: stephan tags: fiddle-opfs) | |
| 14:35 | Improvements to the "demoDelete" method in the demovfs, suggested by forum post cad18961c2. (check-in: d76a301e user: drh tags: trunk) | |
| 14:14 | Merge recent trunk enhancements into the reuse-schema branch. (check-in: b119a5e8 user: drh tags: reuse-schema) | |
| 13:54 | Merge recent trunk enhancements into the wal2 branch. (check-in: c22c7c87 user: drh tags: wal2) | |
| 12:27 | Merge version 3.39.4 changes into the reuse-schema-3.39 branch. (Leaf check-in: 81025928 user: drh tags: reuse-schema-3.39) | |
| 12:10 | Expose sqlite3_msize() to wasm. (check-in: aa6ad34f user: stephan tags: fiddle-opfs) | |
| 11:01 | Add oo1.JsStorageDb() as a convenience wrapper for oo1.DB(...,'kvvfs'). Minor doc cleanups. (check-in: 8a799870 user: stephan tags: fiddle-opfs) | |
| 10:55 | wasm: expose sqlite3_de/serialize(), sqlite3_malloc/free() and friends, noting that the former explicitly lies on use of the latter for memory management so is not generically safe for use in wasm. (check-in: fbc0edb5 user: stephan tags: fiddle-opfs) | |
|
2022-09-29
| ||
| 22:08 | Add JS infrastructure to ostensibly allow us to customize the wasm imports, which will hypothetically allow us to eliminate the dependency on EM_JS(), but the corresponding Emscripten glue-level feature currently breaks fatally with WASMFS builds so it's disabled. (check-in: 88d9253b user: stephan tags: fiddle-opfs) | |
| 17:35 | Ensure that the wasm builds always, even when passing -Ox flags from the CLI, use -g3 to avoid problems caused by symbol minification with optimization levels -O2 and higher. (check-in: 27d24d5d user: stephan tags: fiddle-opfs) | |
| 16:54 | Finish eliminating explicit Emscripten module dependencies in test code and fiddle. The only remnant in public code is the Emscripten-generated module load/init interface. (check-in: 7be78dd4 user: stephan tags: fiddle-opfs) | |
| 15:55 | Version 3.39.4 (check-in: a29f9949 user: drh tags: release, branch-3.39, version-3.39.4) | |
| 13:17 | Rework the Emscripten-emitted module loader/init function such that it passes on the sqlite3 module, instead of the Emscripten module, to the first then() of sqlite3InitModule()'s returned Promise. This eliminates any need to mention the Emscripten module object in client-side code unless they want to configure it in advance for loading-status reports. (check-in: 0dbaa0e2 user: stephan tags: fiddle-opfs) | |
|
2022-09-28
| ||
| 19:19 | Fix misuse of the sqlite3_set_auxdata() interface in the ICU extension. (check-in: 6607dd01 user: drh tags: branch-3.39) | |
| 19:14 | Fix misuse of the sqlite3_set_auxdata() interface in the ICU extension. (check-in: f25cf634 user: drh tags: trunk) | |
| 18:10 | Wasm: expose sqlite3_exec() and use it to simplify the db-reset logic in batch-runner.js a bit. (check-in: 2e2821f7 user: stephan tags: fiddle-opfs) | |
| 17:52 | More work on batch-runner.html/js to facilitate speed comparisons between various VFSes and WebSQL. (check-in: 3bd1bc24 user: stephan tags: fiddle-opfs) | |
| 17:10 | Prototype implementation of "PRAGMA reset_database". This pragma differs from SQLITE_DBCONFIG_RESET_DATABASE in that the pragma only works if the database is reasonably well-formed, whereas the dbconfig works regardless. (Leaf check-in: cf0999d4 user: drh tags: reset-database) | |
| 13:38 | Remove a NEVER macro in defragmentPage() that dbsqlfuzz discovered can be true. crash-32d9312f145cdce41613573f6431d9a3e439e3d7 (check-in: a4342fc0 user: drh tags: branch-3.39) | |
| 13:01 | Correct duplicate copies of sqlite3-api.js being embedded in the wasmfs-based builds. (check-in: bbfcfba2 user: stephan tags: fiddle-opfs) | |
| 07:53 | Work around broken -Os wasm builds by adding the -g3 flag. Unrelated documentation tweaks. (check-in: f5d6bf86 user: stephan tags: fiddle-opfs) | |
|
2022-09-27
| ||
| 17:03 | Minor doc correction. (check-in: 093f6e4b user: stephan tags: fiddle-opfs) | |
| 16:35 | Provide the SQLITE_MAX_ALLOCATION_SIZE compile-time option for limiting the maximum memory allocation size. (check-in: 584de6a9 user: drh tags: trunk) | |
| 14:31 | Comment the -Os wasm builds as being broken (mysteriously missing JS bits). (check-in: f61475ef user: stephan tags: fiddle-opfs) | |
| 13:40 | WASM API renaming. Reworked JS API bootstrap's async post-init into a generic mechanism, no longer OPFS-specific. (check-in: c42a8cb0 user: stephan tags: fiddle-opfs) | |
| 09:17 | wasm/js: rename /persistent to /opfs to account for potential future persistent storage options. Minor flag-handling cleanups in the speedtest1 pages. Minor API tweaks in oo1. (check-in: 4dc972a3 user: stephan tags: fiddle-opfs) | |
| 02:04 | Enhance defensive mode so that it disallows CREATE TRIGGER statements if the statements within the trigger attempt to write on a shadow table. Also make the legacy FTS3 code more robust against integer overflow during memory allocation. (check-in: c41f25e6 user: drh tags: branch-3.39) | |
| 01:53 | Make the legacy FTS3 code more robust against integer overflow during memory allocation. (check-in: 5517bc50 user: drh tags: trunk) | |
| 00:56 | Enhance defensive mode so that it disallows CREATE TRIGGER statements if the statements within the trigger attempt to write on a shadow table. (check-in: 3283bbd1 user: drh tags: trunk) | |
|
2022-09-26
| ||
| 20:51 | Modify implementation of recovery to break recovery of the lost_and_found table into multiple steps. (check-in: adedfd04 user: dan tags: recover-extension) | |
| 13:55 | Fiddle: replace db export routine with a C-side one which works for both Emscripten FS-hosted and OPFS-hosted db files. Minor code-adjacent cleanups. (check-in: 3579a8d6 user: stephan tags: fiddle-opfs) | |
| 11:38 | Get fiddle db export working for OPFS VFS. Add root dir handle to the main OPFS VFS worker to enable creation of certain utility functions without delegating to the async worker. Add sqlite3.capi.sqlite3_wasm_rc_str() to map integer result codes back to their SQLITE_xxx counterparts. Minor doc touchups. (check-in: 9b2244e1 user: stephan tags: fiddle-opfs) | |
| 11:34 | wasm: change StructBinder signature for sqlite3_file::pMethods from 'P' to 'p' to eliminate an unnecessary and inconsistent level of magic. (check-in: 85f2e877 user: stephan tags: fiddle-opfs) | |
|
2022-09-24
| ||
| 19:54 | Add the SQLITE_RECOVER_SLOWINDEXES option, for specifying that indexes should be created and populated along with tables, instead of separately at the end of the recovery operation. (check-in: ad9dba9d user: dan tags: recover-extension) | |
| 19:17 | Add sqlite3_recover_step() to header file sqlite3recover.h. Update the API docs in this file. (check-in: 47f41615 user: dan tags: recover-extension) | |
| 18:05 | Update things to use sqlite3_recover_step() internally. (check-in: f4b15aad user: dan tags: recover-extension) | |
| 11:32 | Fiddle: move, rather than copy, data buffers between the threads. Fix the case of an uploaded db failing to install because its filename is the same as the opened db. (check-in: bcec4f96 user: stephan tags: fiddle-opfs) | |
| 10:15 | Resolve "already configured" warnings from shell's main() when starting it up in fiddle mode. (check-in: 114ef355 user: stephan tags: fiddle-opfs) | |
| 10:12 | Reworked out the OPFS async proxy metrics are fetched so that they play more nicely with the tight event-polling loop. (check-in: ef503ced user: stephan tags: fiddle-opfs) | |
| 07:36 | Refactoring towards getting fiddle to support OPFS as a first-class citizen. Certain operations, e.g. import, export, and unlink, are not OPFS-aware. (check-in: 1b923ed6 user: stephan tags: fiddle-opfs) | |
|
2022-09-23
| ||
| 21:01 | Internal changes to the recover extension towards a step() style interface. (check-in: 73926d5c user: dan tags: recover-extension) | |
| 11:40 | Fix various compiler warnings in new code on this branch. (check-in: ae49e9ef user: dan tags: recover-extension) | |
| 11:30 | Merge latest trunk changes into this branch. (check-in: e87fa70a user: dan tags: recover-extension) | |
|
2022-09-21
| ||
| 20:24 | Reformulate some JS to work around a buggy/broken code transformation in one of the Emscripten-driven code optimizers. (check-in: e1249369 user: stephan tags: fiddle-opfs) | |
| 19:51 | Put pieces in place for fiddle to support opfs, but more cleanup is required in the sqlite3.js/fiddle connection. bigIntEnabled now defaults to whether the Emscripten's module has bigint enabled. Add wasm-sensible defaults for several SQLITE_ENABLE/OMIT flags in sqlite3-wasm.c. (check-in: 7c7fb753 user: stephan tags: fiddle-opfs) | |
| 18:21 | Reduced and improved comments on the constraintCompatibleWithOuterJoin() function. (check-in: 38aaf26e user: drh tags: trunk) | |
| 16:21 | shell.c.in: when building in fiddle mode, define _POSIX_SOURCE (ifndef) so that emcc's string.h reveals strdup(). (check-in: fb85b269 user: stephan tags: fiddle-opfs) | |
| 14:02 | Correct mistyped --shrink-memory flag in speedtest1-worker. Minor OPFS proxy cleanups. (check-in: 86e2b55e user: stephan tags: fiddle-opfs) | |
| 12:27 | Doc cleanups and additions. Add a way for the OPFS async runner to propagate exception text to the calling thread. (check-in: 5c5e8065 user: stephan tags: fiddle-opfs) | |
| 12:25 | Added some structure to the wasm demo link list. (check-in: 777077c4 user: stephan tags: fiddle-opfs) | |
| 09:54 | The RIGHT JOIN logic added to version 3.39.0 added a new restriction on the query flattener which resulted in a performance regression for some queries, as documented by forum post 96b9e5709cf47cda. This change removes that restriction (thus restoring performance) and fixes the problem that RIGHT JOIN was having in a different way. (check-in: 501609ed user: drh tags: trunk) | |
| 08:41 | Add a link to fiddle.html in the wasm test app index.html. (check-in: 96c734c0 user: stephan tags: fiddle-opfs) | |
| 08:39 | Move fiddle build rules into the wasm-centric build files. Add rule to push wasm bits to the wasm test server. (check-in: 113f8204 user: stephan tags: fiddle-opfs) | |
| 00:16 | Break out the decision of whether or not a constraint term is usable by an outer join into a subroutine: constraintCompatibleWithOuterJoin(). (Closed-Leaf check-in: c7fbc9b0 user: drh tags: flattener-omit-restriction-29) | |
|
2022-09-20
| ||
| 22:19 | Do not allow EP_InnerON terms to be used on a outer join. (check-in: f47aa745 user: drh tags: flattener-omit-restriction-29) | |
| 21:07 | Add new test cases that (current) get an incorrect answer. (check-in: fe5c3773 user: drh tags: flattener-omit-restriction-29) | |
| 19:45 | New test cases for query flattening when there are ON clauses and outer joins. (check-in: db55387b user: drh tags: flattener-omit-restriction-29) | |
| 19:22 | Modify the OP_IfNotOpen opcode so that the jump is taken if the cursor is open but was previously set to a NULL row using OP_NullRow. (check-in: 1292d68c user: drh tags: flattener-omit-restriction-29) | |
| 18:18 | Remove code that has been previously commented out. (check-in: 852b385a user: drh tags: flattener-omit-restriction-29) | |
| 17:43 | Merge the fuzzcheck fix from trunk into the flattener improvement branch. (check-in: 930a8c15 user: drh tags: flattener-omit-restriction-29) | |
| 17:21 | Update the fuzz invariant checker to conform to the latest changes in dbsqlfuzz. (check-in: 8d8c1243 user: drh tags: trunk) | |
| 16:57 | Is the query flattener restriction 29 (do not allow both EP_InnerON and EP_OuterON constraints on the same FROM clause term) really necessary? This branch explores what happens if we omit that restriction. This is an effort to address the performance regression reported by forum post 96b9e5709cf47cda that is caused by flattener restriction 29. (check-in: 19270577 user: drh tags: flattener-omit-restriction-29) | |
| 16:20 | Correct a too-strict is-opfs-available check. (check-in: 1b5f1b4a user: stephan tags: fiddle-opfs) | |
| 16:10 | Merge kv-vfs-magic-names branch into fiddle-opfs branch and make some kvvfs-relevant tweaks. (check-in: e3d36dcd user: stephan tags: fiddle-opfs) | |
| 14:52 | Export the sqlite3_uri_...() family of functions to wasm. (check-in: 72bebc84 user: stephan tags: fiddle-opfs) | |
| 14:39 | Merge opfs-proxy-atomics branch into the fiddle-opfs branch. (check-in: 25a36920 user: stephan tags: fiddle-opfs) | |
| 14:36 | When compiled with SQLITE_OS_KV_OPTIONAL, the magic names ":localStorage:" and ":sessionStorage:" are recognized and converted to use the kv-vfs. (Leaf check-in: c5db9262 user: drh tags: kv-vfs-magic-names) | |
| 14:21 | Remove an obsolete file. (Closed-Leaf check-in: 1e438554 user: stephan tags: opfs-proxy-atomics) | |
| 13:25 | Speed up de/serialization of func args and return values in the OPFS VFS proxy. (check-in: 5bf235bb user: stephan tags: opfs-proxy-atomics) | |
| 10:47 | OPFS proxy: remove one sanity-checking-only level of proxy function to shave off a few microseconds. (check-in: b534831f user: stephan tags: opfs-proxy-atomics) | |
| 10:11 | Remove extraneous result-code SAB slots. Doc additions. (check-in: 88de2071 user: stephan tags: opfs-proxy-atomics) | |
| 08:27 |
An alternative messaging strategy for the OPFS VFS proxy which uses only SharedArrayBuffer and Atomics, instead of worker messages, for communication (only the initial one-time handshake during initialization uses worker messages). | |
| 03:31 | OPFS VFS: further internal refactoring towards experimenting with a new comms model. (check-in: 5ca412ce user: stephan tags: fiddle-opfs) | |
| 01:28 | OPFS VFS: moved i/o buffer from per-file to the VFS, and related refactoring, in prep for experimentation with a new inter-worker comms model. (check-in: d4d63e45 user: stephan tags: fiddle-opfs) | |
|
2022-09-19
| ||
| 19:25 | Clarify comments that describe query flattener restrictions associated with outer joins. No code changes. (check-in: 2647d73b user: drh tags: trunk) | |
| 18:22 | Further metrics and buffer-copy optimizations in the OPFS proxy, but with little effect. (check-in: d1f1fe6f user: stephan tags: fiddle-opfs) | |
| 17:41 | Fix a minor comment typo. No changes to code or documentation. (check-in: 8032db4a user: drh tags: trunk) | |
| 17:09 | Cut the speedtest1 runtime of the OPFS VFS proxy by approximately 3/4ths via xRead/xWrite buffer-copying optimizations. Still slower than the WASMFS impl by approx. 1/5th. (check-in: fb7f2873 user: stephan tags: fiddle-opfs) | |
| 14:56 | Remove an obsolete file. Update wasmfs build for today's changes. Fix the case that OPFS.xTruncate() is passed a BigInt instead of a Number. (check-in: 97605392 user: stephan tags: fiddle-opfs) | |
| 13:44 | Add sqlite3_web_vfs_list() to JS API. Corrected OPFS VFS's registering itself as the default VFS. speedtest1-worker now uses the xDelete() of both the default VFS and OPFS, to avoid that it starts up with a persistent OPFS test db (the native app calls unlink(), but that unlink call operates on a different virtual filesystem than the OPFS VFS). (check-in: 2ec7e091 user: stephan tags: fiddle-opfs) | |
| 13:16 | Merge kv-vfs branch into fiddle-opfs. Adjust various JS APIs and apps to deal with the new method of handling kvvfs. Adjust speedtest1 JS build to include sqlite3-api.js so that it can use kvvfs and opfs VFSes. Permit passing of the vfs as a URL parameter to certain demo/test apps. Milestone: speedtest-worker.html?vfs=opfs runs with the standalone OPFS impl. (check-in: ec09f32f user: stephan tags: fiddle-opfs) | |
| 11:47 | speedtest1: add --vfs NAME flag and eliminate code duplication in argc size verification for all flags which take a value. (check-in: 281d0986 user: stephan tags: fiddle-opfs) | |
| 10:13 | Minor doc improvements in demo-123.js. (check-in: 9370ce94 user: stephan tags: fiddle-opfs) | |
| 09:58 | Generic cleanups in the OPFS VFS proxies. (check-in: 7d5f6adc user: stephan tags: fiddle-opfs) | |
| 09:25 | Correct OPFS VFS xRead() to copy the result buffer if the result code is SQLITE_IOERR_SHORT_READ. (check-in: 56668f99 user: stephan tags: fiddle-opfs) | |
| 05:19 | Replace OPFS VFS xSleep() impl with a more efficient one (no Worker round-trip needed). (check-in: b9773f16 user: stephan tags: fiddle-opfs) | |
| 03:57 | Rename demo-oo1.* to demo-123.* and add demo-123-worker.html, which runs the same demo via a Worker. Doc typo fixes. (check-in: 2e4a005b user: stephan tags: fiddle-opfs) | |
| 00:40 | Disable build of wasmfs-using components on aarch64 (ARM), as wasmfs won't build there. (check-in: 9a9eeebc user: stephan tags: fiddle-opfs) | |
|
2022-09-18
| ||
| 17:59 | Remove a NEVER macro in defragmentPage() that dbsqlfuzz discovered can be true. crash-32d9312f145cdce41613573f6431d9a3e439e3d7 (check-in: 90403759 user: drh tags: trunk) | |
| 17:32 | Numerous cleanups in the JS bits. Removed some now-defunct wasm test files. Expose sqlite3.opfs object containing various OPFS-specific utilities. (check-in: 26e625d0 user: stephan tags: fiddle-opfs) | |
| 03:05 | Move the OPFS VFS bits back into api/sqlite3-api-opfs.js. Refactor the OPFS VFS init process to use a Promise-returning function which the client must call, as that eliminates any uncertainty about when the VFS (necessarily activated asynchronously) actually becomes available to the client. Rename x-sync-async.* to test-opfs-vfs.* Milestone: first successful test of OPFS without WASMFS. (check-in: b2abf60d user: stephan tags: fiddle-opfs) | |
| 02:35 | Move the OPFS VFS bits back into api/sqlite3-api-opfs.js. Refactor the OPFS VFS init process to use a Promise-returning function which the client must call, as that eliminates any uncertainty about when the VFS (necessarily activated asynchronously) actually becomes available to the client. (check-in: 1c660970 user: stephan tags: fiddle-opfs) | |
| 00:16 | Plug OPFS methods in to their sqlite3_vfs/io_methods counterparts. Add URL args to control debug output and running of sanity-checks in the OPFS init code. (check-in: a0e93ed2 user: stephan tags: fiddle-opfs) | |
|
2022-09-17
| ||
| 23:47 | Correct OPFS xRead() and xWrite() impls and add a very basic sanity test for them. (check-in: cd06cc67 user: stephan tags: fiddle-opfs) | |
| 23:29 | Implement OPFS xAccess(), albeit with more limited semantics than the VFS API calls for. Add a way for OPFS xDelete() to optionally recursively remove empty dirs left over after deleting a file. (check-in: c342b5d7 user: stephan tags: fiddle-opfs) | |
| 21:13 | Generic cleanups in the OPFS sync/async proxy. (check-in: f36bddbe user: stephan tags: fiddle-opfs) | |
| 20:50 | Add the remaining vfs/io_methods wrappers to the OPFS sync/async proxy, but most are not yet tested. (check-in: 44db9132 user: stephan tags: fiddle-opfs) | |
| 18:31 | Fix an uninitialized variable in the decoder kv-vfs. (check-in: 354726aa user: drh tags: kv-vfs) | |
| 18:29 | Include the kv-vfs as an optional VFS on unix builds if the SQLITE_OS_KV_OPTIONAL compile-time option is present. (check-in: 852812d1 user: drh tags: kv-vfs) | |
| 17:26 | Merge the latest trunk enhancements into the kv-vfs branch. (check-in: 37c7ca25 user: drh tags: kv-vfs) | |
| 15:08 | Add initial bits of an experimental async-impl-via-synchronous-interface proxy intended to marshal OPFS via sqlite3_vfs API. (check-in: 38da059b user: stephan tags: fiddle-opfs) | |
|
2022-09-16
| ||
| 20:16 | Merge kv-vfs branch into fiddle-opfs branch. Adjust speedtest1 --size flags to account for new size limit. (check-in: afb79050 user: stephan tags: fiddle-opfs) | |
| 15:19 | Corrections to the xRead method for databases in the os_kv.c VFS so that it correctly reads the header, and thus avoids unnecessary cache flushes. This changes also includes an optimization to the header read logic. (check-in: ef54961c user: drh tags: kv-vfs) | |
| 12:49 | When calculating kvvfs sessionStorage/localStorage size, multiply it by 2 to account for JS using 16-bit characters. (check-in: 52d1b185 user: stephan tags: fiddle-opfs) | |
| 11:45 | Merge kv-vfs branch into fiddle-opfs branch for [21915af560b1|synchronous=off fix]. Remove some duplicate debug output in OPFS test code. (check-in: 13899bb9 user: stephan tags: fiddle-opfs) | |
| 11:37 | Fix os_kv.c so that it uses SQLITE_FCNTL_SYNC and hence works even with PRAGMA synchronous=OFF. (check-in: 21915af5 user: drh tags: kv-vfs) | |
| 02:30 | Add batch-runner-kvvfs.html, a kvvfs-specific build of batch-runner.html. Reduce the speedtest1 --size X value for the batch list generation to 5 so that the kvvfs batch runner can handle it. (check-in: d8df2592 user: stephan tags: fiddle-opfs) | |
| 02:28 | Remove --nosync flag from speedtest1-kvvfs.html, as that seems to be the trigger for the 'no such table: z1' error. That app now warns if that flag is passed to it via URL args. (check-in: ad0677e8 user: stephan tags: fiddle-opfs) | |
| 01:08 | Pull the src/os_kv.c part of [13839759f8f4] into the kv-vfs branch. (check-in: e3344499 user: stephan tags: kv-vfs) | |
| 01:05 | Add return value checks to kvstorageWrite(), necessary for recognizing storage-full errors in the JS binding. speedtest1-kvvfs.html runs successfully on localStorage with --size 1 to 6 and starts failing with storage shortage at --size 7. (check-in: 13839759 user: stephan tags: fiddle-opfs) | |
|
2022-09-15
| ||
| 06:42 | More work on the synchronous OPFS experimentation. Numerous wasm/js build tweaks. Add speeedtest-wasmfs.html, the wasmfs/opfs counterpart of speedtest1.html. (check-in: 00ee49a3 user: stephan tags: fiddle-opfs) | |
| 03:54 | Fix windows builds of recover extension. (check-in: abcbb6ab user: dan tags: recover-extension) | |
| 03:16 | Remove some stale comments. Get scratchpad-opfs-worker2.js running again in prep for reuse in experimenting with the OPFS all-synchronous API. (check-in: 4cbbd370 user: stephan tags: fiddle-opfs) | |
| 03:09 | Split wasmfs-enabled build of sqlite3.js/wasm into sqlite3-wasmfs.js/wasm, as enabling wasmfs breaks all tests/demos which run from a Worker thread. (check-in: 08476f3c user: stephan tags: fiddle-opfs) | |
| 02:27 | Clean up some JS documentation. (check-in: 925fdbfc user: stephan tags: fiddle-opfs) | |
|
2022-09-14
| ||
| 20:45 | Include recovery extension in configure/make builds of testfixture. (check-in: d2f46521 user: dan tags: recover-extension) | |
| 18:57 | Extra tests and fixes. (check-in: deed5336 user: dan tags: recover-extension) | |
| 16:37 | Fix recovery of utf-16 databases. (check-in: 5b05be08 user: dan tags: recover-extension) | |
|
2022-09-13
| ||
| 20:40 | Add OOM tests for the recovery extension. (check-in: 9b6b4c71 user: dan tags: recover-extension) | |
| 19:47 | Fix a broken error handling case in the recovery extension. (check-in: 60089547 user: dan tags: recover-extension) | |
| 19:27 | Add/apply various kvvfs-specific utility APIs to the JS layer to assist in testing and analysis. Correct a backwards default arg check for sqlite3ApiBootstrap(). Add exports for sqlite3_db_handle(), sqlite3_file_control(), and the SQLITE_FCNTL_xxx enum. (check-in: 0d789618 user: stephan tags: fiddle-opfs) | |
| 18:08 | Add tests. Deal issues surrounding aborting recovery from within the SQL callback, and avoiding the pending-byte page. (check-in: 4e97dd31 user: dan tags: recover-extension) | |
|
2022-09-12
| ||
| 22:27 | Get testing1.js working with a kvvfs build. (check-in: 333a4572 user: stephan tags: fiddle-opfs) | |
| 20:21 | Remove an extraneous key-copy op in the EM_JS impl of kvstorageDelete(). (check-in: 1c5aeee4 user: stephan tags: kv-vfs) | |
| 20:18 | Merge kv-vfs branch into fiddle-opfs and add speedtest1-kvvfs.html test app (which currently fails early on). (check-in: 8b1608a9 user: stephan tags: fiddle-opfs) | |
| 20:02 | Fix some problems with handling corrupt database in the recovery extension. (check-in: ed318be8 user: dan tags: recover-extension) | |
| 19:33 | Fix the <sys/time.h> include in os_kv.c. (check-in: 878cc93e user: drh tags: kv-vfs) | |
| 19:29 | Add a gettimeofday()-based implementation of xCurrentTimeInt64() to os_kv.c. (check-in: e393ed65 user: drh tags: kv-vfs) | |
| 19:23 | More tests for the recover module. (check-in: 37fb093b user: dan tags: recover-extension) | |
| 18:10 | Fix a debugging/testing edit in the previous check-in. (check-in: 6fc8d34c user: drh tags: kv-vfs) | |
| 17:59 | Add sqlite3.oo1.DB.clearKvvfsStorage(). Add controls to kvvfs1.js demo to reset and query the db without requiring the dev console. (check-in: d845c6c2 user: stephan tags: fiddle-opfs) | |
| 17:44 | Faster text decoder for kv_os.c. (check-in: 3354a2ed user: drh tags: kv-vfs) | |
| 16:09 | Merge kv-vfs branch into fiddle-opfs branch to add kvvfs-based wasm build and demo. (check-in: a7d8b26a user: stephan tags: fiddle-opfs) | |
| 16:07 | New run-time options to the speed-check.sh script. (check-in: ff46bc9d user: drh tags: trunk) | |
| 15:59 | Fix uninitialized variable in rollback-journal processing in os_kv.c (check-in: e49682c5 user: drh tags: kv-vfs) | |
| 13:05 | Resolve missing kvvfs JS test: sqlite3_initialize() must be called by the client in this build. Add basic sanity test of kvvfs using sessionStorage(). (check-in: 250a935a user: stephan tags: kv-vfs) | |
| 12:41 | Fix the amalgamation so that it builds with the standard unix backend again. (check-in: 0adf158f user: drh tags: kv-vfs) | |
| 12:39 | Initial build of kvvfs in wasm. It loads but cannot find the VFS for as-yet-unknown reasons (sqlite3 shell works fine), and most APIs throw "null function or function signature mismatch" from deep within wasm, presumably as a side effect of the "missing" VFS. (check-in: 1a2f24a0 user: stephan tags: kv-vfs) | |
|
2022-09-11
| ||
| 16:59 | Minor cleanups and documentation in the wasm pieces. (check-in: 4e6ce329 user: stephan tags: fiddle-opfs) | |
| 05:44 | Correct the result code from kvstorageRead() for the OOM case to be a negative value. (check-in: cdbf09fa user: stephan tags: fiddle-opfs) | |
| 05:38 | Minor cleanups and docs for the EM_JS() impls of the kvvfs read/write/delete ops. (check-in: 53b7572e user: stephan tags: fiddle-opfs) | |
| 04:49 | Add EM_JS() impl for kvstorageRead(). (check-in: 06610314 user: stephan tags: fiddle-opfs) | |
| 02:43 | Initial prototype impls of write/delete-key ops for the pending kvvfs which use sessionStorage or localStorage for db page storage. read-key op is pending. (check-in: a4bd96f5 user: stephan tags: fiddle-opfs) | |
|
2022-09-10
| ||
| 20:01 | Rework recover extension code for readability. (check-in: 1a254096 user: dan tags: recover-extension) | |
| 18:38 | Improved comments in os_kv.c. Better names for the key/value name spaces. (check-in: 8e1652a3 user: drh tags: kv-vfs) | |
| 18:20 | Move the vfskv.c extension to src/os_kv.c and make it part of the amalgamation, activated if and only if SQLITE_OS_KV is true. (check-in: f6632e69 user: drh tags: kv-vfs) | |
| 13:50 | Enhance the b-tree page sorting code to ensure that sqlite3PagerRekey() never overloads a page number and uses only the PENDING_BYTE page for temporary storage. (check-in: aa6a5b72 user: dan tags: branch-3.39) | |
|
2022-09-09
| ||
| 20:44 | Further fixes to comments in sqlite3recover.h. Also rework some data structures in sqlite3recover.c. (check-in: 599d1f8e user: dan tags: recover-extension) | |
| 17:50 | Fix harmless compiler warning seen with MSVC. (check-in: 73c4c68d user: mistachkin tags: trunk) | |
| 16:25 | Handle recovery of virtual tables by recovering each shadow table individually, then writing the CREATE VIRTUAL TABLE statement directly into the sqlite_schema table. (check-in: 5f2d5ccd user: dan tags: recover-extension) | |
| 16:16 | Database must be named either "local" or "session". Each has a distinct namespace. (check-in: c8e41279 user: drh tags: kv-vfs) | |
| 14:22 | Separate sqlite3_io_methods objects for database files and journals. (check-in: 74fbf6c2 user: drh tags: kv-vfs) | |
| 11:41 | Simple reading and writing now works. (check-in: 00845ac9 user: drh tags: kv-vfs) | |
| 05:39 | Remove a reference to a completed TODO. (check-in: 757ad696 user: stephan tags: fiddle-opfs) | |
| 04:50 | speedtest1 wasm: add a link in the worker variant which launches the main-thread variant with the selected CLI flags. Append collected stdout/stderr to the main-thread page after execution is finished to avoid having to open the dev tools (which inexplicably slows down wasm execution). (check-in: 02709ee2 user: stephan tags: fiddle-opfs) | |
|
2022-09-08
| ||
| 21:43 | Update comments in sqlite3recover.h. (check-in: 14164047 user: dan tags: recover-extension) | |
| 21:33 | Make use of the --big-transactions flag in the speedtest1 JS apps. (check-in: f2846dcb user: stephan tags: fiddle-opfs) | |
| 21:09 | Add speedtest1 --big-transactions flag to cause its large tests which rely on implicit transactions to be wrapped in BEGIN/COMMIT, per /chat discussion. Added to support of the WASMFS build, which slows down significantly when thousands of implicit transactions are used. (check-in: 51395c00 user: stephan tags: fiddle-opfs) | |
| 20:04 | Minor build cleanups and tweaks in the speedtest1 wasm apps. (check-in: 5240fb4d user: stephan tags: fiddle-opfs) | |
| 19:53 | Experimentally add --nomutex flag to speedtest1 to have it open db with the SQLITE_OPEN_NOMUTEX flag. (check-in: 45ef576b user: stephan tags: fiddle-opfs) | |
| 19:22 | Update the shell to use the recover extension for the .recover command. (check-in: ae832e77 user: dan tags: recover-extension) | |
| 17:42 | Fix problems with recovering the sqlite_sequence table. (check-in: 356d2209 user: dan tags: recover-extension) | |
| 17:12 | Compiles and loads as an extension. Starts to run but quickly hits issues. (check-in: 2e38726f user: drh tags: kv-vfs) | |
| 16:27 | Compiles without error, but untested. (check-in: a329939c user: drh tags: kv-vfs) | |
| 15:48 | Code snapshot. Completely untested. Probably does not compile. (check-in: 51e26769 user: drh tags: kv-vfs) | |
| 15:30 | Add speedtest1-worker.html, an interactive Worker-thread variant of speedtest1.html. Add ext/wasm/index.html to act as a gateway to the various test pages. (check-in: f16c68ee user: stephan tags: fiddle-opfs) | |
| 11:04 | Add new test file recoverclobber.test. (check-in: cb4e950c user: dan tags: recover-extension) | |
|
2022-09-07
| ||
| 20:14 | Enhance an assert() to impose for tighter constraints on the operation of pcache. (check-in: 461511cc user: drh tags: branch-3.39) | |
| 20:11 | Enhance an assert() to impose for tighter constraints on the operation of pcache. (check-in: 42105eb4 user: drh tags: trunk) | |
| 19:58 | Fix the windows build so that it works with -DSQLITE_OMIT_AUTOINIT. (check-in: 1662eb07 user: drh tags: branch-3.39) | |
| 19:50 | Ensure that the Rekey() operation does not overwrite an existing page number. (check-in: cd1e4006 user: drh tags: branch-3.39) | |
| 19:35 | Increase the version number to 3.39.4 (check-in: e98bed12 user: drh tags: branch-3.39) | |
| 19:28 | An improved technique for ensuring that the Rekey() operation does not overwrite an existing page number. This approach does not change the semantics of the underlying pluggable pcache and it is easier to prove correct. This replaces the changes at [aadd38f99a3e5abc] and [81cff9f1955874aa]. (check-in: 7dbddde7 user: drh tags: trunk) | |
| 17:29 | Non-working code towards a VFS for text key/value storage. (check-in: f9c89ee8 user: drh tags: kv-vfs) | |
| 16:41 | Ensure that the recover extension properly escapes CR and NL characters in text mode. Also that it holds transactions open on both input and output databases for the duration of a recovery operation. (check-in: 6cca8913 user: dan tags: recover-extension) | |
|
2022-09-06
| ||
| 23:04 | Add a note about Emscripten's -sSINGLE_FILE flag, why it would be nice, and why we can't use it. (check-in: 5ea06236 user: stephan tags: fiddle-opfs) | |
| 20:17 | Get speedtest1.js working with WASMFS/OPFS. (check-in: 40e60f57 user: stephan tags: fiddle-opfs) | |
| 19:38 | Tests and a fix for the SQLITE_RECOVER_ROWIDS option. (check-in: 1d5000f5 user: dan tags: recover-extension) | |
| 16:47 | Initial build of speedtest1.wasm and speedtest1.html with which to run it. (check-in: 4441535e user: stephan tags: fiddle-opfs) | |
| 16:35 | Minor internal doc clarifications. (check-in: 09796cc2 user: stephan tags: fiddle-opfs) | |
| 09:59 | Merge trunk into fiddle-opfs branch. (check-in: 7f76eaec user: stephan tags: fiddle-opfs) | |
|
2022-09-05
| ||
| 22:54 | Add the tool/omittest-msvc.tcl script that will try to build using MSVC using various compile-time options, to ensure that the compile-time options all build without errors. (check-in: 6b00ecb5 user: drh tags: trunk) | |
| 22:53 | Get the tool/omittest-msvc.tcl script working. (Closed-Leaf check-in: 82b89f8a user: drh tags: msvc-omit-test) | |
| 21:22 | Fix a problem with script mode and lost-and-found tables. (check-in: 09ec588d user: dan tags: recover-extension) | |
| 21:00 | Add a mode to output SQL statements instead of populating a database to the recover extension. (check-in: 73058416 user: dan tags: recover-extension) | |
| 20:42 | Add a prototype for a script, similar to tool/omittest.tcl, that is designed to test builds using MSVC to ensure they work with various OMIT options. (check-in: b1e37a0b user: drh tags: msvc-omit-test) | |
| 18:17 | Fix the windows build so that it works with -DSQLITE_OMIT_AUTOINIT. (check-in: f74a5ea8 user: drh tags: trunk) | |
| 15:56 | Add the SQLITE_RECOVER_ROWIDS option. To specify that rowid values that are not also explicit INTEGER PRIMARY KEY values should be preserved. (check-in: 69cc9aba user: dan tags: recover-extension) | |
| 13:36 | Merge patches for version 3.39.3 into the reuse-schema-3.39 branch. (check-in: efd54164 user: drh tags: reuse-schema-3.39) | |
| 13:24 | Minor cleanups in OO API #1. Add Sudoku SQL to batch-runner.js's list. (check-in: 4488cb57 user: stephan tags: fiddle-opfs) | |
| 11:02 | Version 3.39.3 (check-in: 4635f4a6 user: drh tags: release, branch-3.39, version-3.39.3) | |
|
2022-09-03
| ||
| 21:08 | Add an option to assume the freelist is corrupt when recovering data. (check-in: 253e498f user: dan tags: recover-extension) | |
| 20:31 | Take the freelist into account when recovering data that is not linked in to any tree associated with a schema entry. (check-in: dbd1f1ef user: dan tags: recover-extension) | |
| 20:07 | Further work on making the recover extension compatible with the .recover command. (check-in: f2ac3158 user: dan tags: recover-extension) | |
| 11:41 | wasm: minor text and build cleanups. (check-in: 36ceef94 user: stephan tags: fiddle-opfs) | |
|
2022-09-02
| ||
| 21:21 | Ensure the Pager.journalOff variable is zeroed if an error occurs while writing the initial header to the journal file. (check-in: fe547035 user: dan tags: branch-3.39) | |
| 21:19 | Ensure the Pager.journalOff variable is zeroed if an error occurs while writing the initial header to the journal file. (check-in: da7af290 user: dan tags: trunk) | |
| 19:18 | Remove a NEVER() that is sometimes true. (check-in: 85a8a117 user: drh tags: branch-3.39) | |
| 17:26 | Fix an assert() associated with the dbsqlfuzz error at [8372468bb5d8922c]. (check-in: f448480e user: drh tags: branch-3.39) | |
| 17:25 | Fix an assert() associated with the dbsqlfuzz error at [8372468bb5d8922c]. (check-in: 54251f81 user: drh tags: trunk) | |
| 16:45 | When an OOM occurs and sets the Parse.nErr value, also set the Parse.nErr value for all outer Parse objects. dbsqlfuzz d33f60aaa67733aa700cd69dacf8e0e23a327a29 (check-in: 073621af user: drh tags: branch-3.39) | |
| 15:53 | When an OOM occurs and sets the Parse.nErr value, also set the Parse.nErr value for all outer Parse objects. dbsqlfuzz d33f60aaa67733aa700cd69dacf8e0e23a327a29 (Leaf check-in: b17a2503 user: drh tags: branch-3.39-pcache) | |
| 15:19 | If sqlite3PcacheMove() moves a page on top of another, swap the two pages, moving the other back to the page number of the original. (check-in: 04eb9f97 user: drh tags: branch-3.39-pcache) | |
| 15:09 | If sqlite3PcacheMove() really performs a swap, then be sure to renumber the second page. (check-in: 81cff9f1 user: drh tags: trunk) | |
| 14:29 | Enhance the pcache1 implementation so that during an xRekey operation if another page already exists at the destination, that other page gets moved to the source key. (check-in: aadd38f9 user: drh tags: trunk) | |
| 12:12 | Mutex protect access to the sqlite3_test_directory and sqlite3_data_directory global variables. See forum thread 719a11e1314d1c70. (check-in: 34c6ec39 user: drh tags: branch-3.39) | |
| 11:45 | Mutex protect access to the sqlite3_test_directory and sqlite3_data_directory global variables. See forum thread 719a11e1314d1c70. (check-in: 8e6ad3a3 user: drh tags: trunk) | |
| 11:12 | Improved mutex protection of the sqlite3_temp_directory and sqlite3_data_directory global variables. (Closed-Leaf check-in: ebbe9634 user: drh tags: tempdir-mutex) | |
| 00:36 | Experimental changes to put sqlite3_temp_directory behind a mutex. (check-in: 5ee3515f user: drh tags: tempdir-mutex) | |
|
2022-09-01
| ||
| 21:00 | Further work on making recovery extension compatible with the shell tool ".recover" code. (check-in: 8df7c7d0 user: dan tags: recover-extension) | |
| 14:33 | Defer deleting a transient SELECT statement associated with a flattening of one arm of a compound SELECT until after the parse has completed. (check-in: 8403c0d5 user: drh tags: branch-3.39) | |
| 13:51 | Defer deleting a transient SELECT statement associated with a flattening of one arm of a compound SELECT until after the parse has completed. This is a follow-up and enhancement to check-in [6e6b3729e0549de0] in response to an assertion fault reported by Chromium. (check-in: 1c4157c7 user: drh tags: trunk) | |
| 10:29 | In the query planner, add a heuristic that will reduce the estimated cost of a full table scan for a materialized view or subquery if the full scan is the outer-most loop. This is shown to speed up some queries. (check-in: 609fbb94 user: drh tags: trunk) | |
|
2022-08-31
| ||
| 20:45 | Add new files for an extension to recover data from corrupted databases. (check-in: f8298eeb user: dan tags: recover-extension) | |
| 15:04 | Enhance the b-tree page sorting code to ensure that sqlite3PagerRekey() never overloads a page number and uses only the PENDING_BYTE page for temporary storage. (check-in: 50077428 user: drh tags: trunk) | |
| 14:40 | Do not attempt the OP_Count optimization on queries with HAVING clauses. This fixes a problem exposed by [2cf373b10c9bc4cb]. (check-in: 5f93239a user: drh tags: branch-3.39) | |
| 00:44 | Checked in the wrong change by mistake. (Closed-Leaf check-in: 984a07d7 user: drh tags: mistake) | |
|
2022-08-30
| ||
| 17:57 | batch-runner.js: add rudimentary metrics export to CSV. Add a toggle to reverse the log output mode (in normal order or most recent first). (check-in: b26e2bc0 user: stephan tags: fiddle-opfs) | |
| 16:54 | Improved comments in pcache1.c. No changes to code. (check-in: dd017bb1 user: drh tags: trunk) | |
| 14:13 | Any function call can abort. Take this into account when deciding if a DML statement needs a statement journal. See forum thread 9b9e4716c0d7bbd1 for more information. (check-in: 5bf4c5fd user: drh tags: branch-3.39) | |
| 10:26 | batch-runner.js: re-enable unlink-before-run so that OPFS speedtest1 batch can work. (check-in: ba08f7a7 user: stephan tags: fiddle-opfs) | |
| 10:04 | batch-runner.js: move generated SQL files into ./sql and start adding hand-written ones (like the Mandelbrot) to that set. (check-in: 06c106a7 user: stephan tags: fiddle-opfs) | |
| 09:51 | Remove the .timer script entry from fiddle app because emscripten has removed getrusage(), which breaks the timer. (check-in: 9034f19a user: stephan tags: fiddle-opfs) | |
| 09:49 | wasm util: remove superfluous function.bind() calls to eliminate a level of call() indirection. (check-in: 1aab9627 user: stephan tags: fiddle-opfs) | |
| 00:11 | Any function call can abort. Take this into account when deciding if a DML statement needs a statement journal. See forum thread 9b9e4716c0d7bbd1 for more information. (check-in: 5c95ae6c user: drh tags: trunk) | |
|
2022-08-29
| ||
| 18:58 | Lots of tweaking in batch-runner.js. Minor internal API update in OO #1 API. (check-in: 24b82b95 user: stephan tags: fiddle-opfs) | |
| 17:41 | Add test descriptions to the speedtest1 '-- begin' markers for potential display by the downstream JS code which uses those markers. (check-in: e5b7006f user: stephan tags: fiddle-opfs) | |
| 12:39 | Add batch-runner.js for running batch SQL scripts with timing info. (check-in: 11f3ed61 user: stephan tags: fiddle-opfs) | |
| 12:31 | Add get/setPtrValue() to the common wasm utils. (check-in: 24d70fc4 user: stephan tags: fiddle-opfs) | |
| 09:51 | Add begin/end comment markers to the test blocks in speedtest1 --script output to facilitate chopping up the output into chunks for JS-side processing. (check-in: ef0b7ef2 user: stephan tags: fiddle-opfs) | |
| 08:04 | oo #1 api: correct a case where a null callback is called. Rename some vars for clarity. Increase wasm-side memory in order to be able to load the speedtest1 output. (check-in: b5058f14 user: stephan tags: fiddle-opfs) | |
|
2022-08-27
| ||
| 16:11 | More assert() statements in the page cache. (check-in: 7c96a56c user: drh tags: trunk) | |
| 14:43 | Ensure that every pcache page always has either the PGHDR_DIRTY or the PGHDR_CLEAN bit set, even during transient page reshuffling during btree balancing. (check-in: e67f03c1 user: drh tags: trunk) | |
|
2022-08-25
| ||
| 19:38 | Merge changes from trunk into the fiddle-opfs branch. (check-in: 34279797 user: stephan tags: fiddle-opfs) | |
| 19:29 | Ensure that sqlite3_prepare() inputs always have a semicolon terminator in speedtest1. (check-in: c72756b0 user: drh tags: trunk) | |
| 19:19 | Add the --script option to speedtest1. (check-in: 1c87d7c5 user: drh tags: trunk) | |
| 18:29 | Fix the bytecode engine so that it is able to handle an incoming SQLITE_TOOBIG error, which is now possible using the statement cache. (check-in: 7763b98b user: drh tags: stmt-cache) | |
| 17:12 | Do not attempt to cache DDL statement. (check-in: 8f6a1f77 user: drh tags: stmt-cache) | |
| 15:09 | Improved documentation of new features. Minor speed enhancements. (check-in: da9aa547 user: drh tags: stmt-cache) | |
| 14:31 | Merge changes from trunk into the stmt-cache branch. (check-in: bc988ce7 user: drh tags: stmt-cache) | |
| 13:32 | Remove a NEVER() macro for a condition that is actually reachable following an OOM. dbsqlfuzz crash-6ef3cd3b18ccc5de86120950a0498641acd90a33.txt. (check-in: b573e2cf user: dan tags: trunk) | |
| 13:27 | Consolidate oo1.DB.exec() and oo1.DB.execMulti() into oo1.DB.exec(). This is a bit less efficient but certainly easier for a client to deal with and lightens the maintenance burden. (check-in: 7eff7213 user: stephan tags: fiddle-opfs) | |
| 13:09 | If a perpared statement runs the Expire opcode, then disqualify it from being cached. (check-in: f27d919f user: drh tags: stmt-cache) | |
| 11:39 | Refactor and expand the worker1 docs, consolidating them into the top of their file instead of scattered around the internals. Accommodate an API change from yesterday in demo-oo1.js. (check-in: 0a657470 user: stephan tags: fiddle-opfs) | |
|
2022-08-24
| ||
| 23:50 | Change sqlite3_prepare_v3() to require SQLITE_PREPARE_CACHE in order for a statement to be a cache candidate. But any statement can pull from the cache. Also add statement cache control to the CLI. (check-in: c9dc536d user: drh tags: stmt-cache) | |
| 20:57 | Change DB.exec() rowMode default from 'stmt' to 'array', per /chat discussion. Add DB.exec() rowMode option for fetching a specific column by name. Add result column names to worker1 exec() callback interface, as there's otherwise no way to get that info from a worker. (check-in: 1bb37e5c user: stephan tags: fiddle-opfs) | |
| 18:39 | Expand the worker1 'exec' op handling for per-row callbacks for API-level consistency and smooth some edges between worker1 core and worker1-promiser. Add worker1 'config-get' message to fetch the serializable parts of the sqlite3.config state. Improve the 'open' op's handling of the 'persistent' option (noting that we cannot yet test that case from a worker). (check-in: 509f8839 user: stephan tags: fiddle-opfs) | |
| 17:59 | Add a mutex to the sqlite3_db_config() interface so that it is threadsafe when two or more threads call it on the same database connection at the same time. (check-in: 459ad884 user: drh tags: trunk) | |
| 17:55 | First attempt at adding a built-in prepared statement cache. This is mostly working, but still has a few obscure faults. (check-in: c217b763 user: drh tags: stmt-cache) | |
| 14:50 | js: resolve the mysterious "extra" unhandled exception notification, caused by inadvertently forking one promise into two separate ones (failing to properly reassign a then() result). Fix a typo in new Worker 1 code which caused the DB(filename) name to be incorrect. (check-in: 7467ac88 user: stephan tags: fiddle-opfs) | |
| 11:51 | Fix an inefficiency in the ".dump" command when outputing large blob literals. (check-in: c055d05d user: dan tags: trunk) | |
| 05:59 | More work on how to configure the sqlite3 JS API bootstrapping process from higher-level code. Initial version of sqlite3-worker1-promiser, a Promise-based proxy for the Worker API #1. (check-in: b030f321 user: stephan tags: fiddle-opfs) | |
| 00:51 | The very basics of a Promise-based proxy for the Worker #1 API. Still requires considerable cleanup, testing, and a solution for the exec-callback-via-event-type-name problem. (check-in: 1e447849 user: stephan tags: fiddle-opfs) | |
| 00:10 | Significant restructuring of the Worker #1 request/response object structures to improve readability and clarity. (check-in: 03b9db9b user: stephan tags: fiddle-opfs) | |
|
2022-08-23
| ||
| 20:11 | Change the names of the Vdbe.pVNext and Vdbe.ppVPrev fields to make them unique. Change to ppVPrev to save a few bytes and a few CPU cycles. (check-in: 34b8ea31 user: drh tags: trunk) | |
| 17:51 | Enhance assert() statements to enforce tighter bounds on the P5 operand of the OP_Next and OP_Prev opcodes. (check-in: 4e0a07fc user: drh tags: trunk) | |
| 17:02 | Wasm build flag tweaks and documentation. (check-in: c8eb3aa8 user: stephan tags: fiddle-opfs) | |
| 16:55 | Add mksourceid binary to the clean rules and $(TESTPROGS) to distclean. (check-in: ae3e82f5 user: stephan tags: trunk) | |
|
2022-08-22
| ||
| 21:37 | More experimentation with how client-side config state can be passed on to initialize the JS-side sqlite3 API. (check-in: ea2acc45 user: stephan tags: fiddle-opfs) | |
| 16:30 | Merge recent trunk enhancements into the reuse-schema branch. (check-in: 9c05a48e user: drh tags: reuse-schema) | |
| 15:57 | Merge all recent trunk enhancements into the wal2 branch. (check-in: 0b7578bf user: drh tags: wal2) | |
| 15:18 | Performance enhancement for sqlite3DbFree(). (check-in: c296a9d9 user: drh tags: trunk) | |
| 14:03 | Merge recent trunk changes into the fiddle-opfs branch. (check-in: e215d55a user: stephan tags: fiddle-opfs) | |
| 13:34 | Refactor JS API amalgamation such that the bootstrapping/configuration is deferred until the whole amalgamation is available, to facilitate providing clients with a way to initialize the API with their own config (noting that we're still one small level of refactoring away from being able to actually do that). (check-in: 9dbe9a6a user: stephan tags: fiddle-opfs) | |
| 08:55 | wasm: accommodated a JS API rename. (check-in: 00991335 user: stephan tags: fiddle-opfs) | |
| 02:00 | Add the new internal interface sqlite3DbNNFreeNN(db,ptr) where both the db and ptr parameters are guaranteed to be non-NULL. Use this where appropriate to save more than 2 million CPU cycles on the standard performance test. (check-in: e5eaa80e user: drh tags: trunk) | |
|
2022-08-20
| ||
| 19:45 | Merge the test case for dbsqlfuzz 18fe4e257be7fa3ecfb0424ab7427e41e97ef9e3 assertion fault, which is fixed with the previous check-in. (check-in: 2d13ec08 user: drh tags: trunk) | |
| 19:33 | Enhance the enforcement of SQLITE_VTAB_DIRECTONLY so that it applies to DML statements within triggers. Do not allow DML stratements against virtual tables within triggers unless either the virtual table is SQLITE_VTAB_INNOCUOUS or PRAGMA trusted_schema is ON. (check-in: 9433ea40 user: drh tags: trunk) | |
|
2022-08-19
| ||
| 20:10 | Add test case for fuzzer case crash-18fe4e. (Closed-Leaf check-in: 8372468b user: dan tags: crash-18fe4e) | |
|
2022-08-18
| ||
| 15:53 | Add notes to ext/wasm/README.md explaining how to run tests on a remote machine. (check-in: 7a3c444f user: drh tags: fiddle-opfs) | |
| 12:21 | javascript: rename and simplify DB.callInTransaction() as DB.transaction(). Add DB.savepoint(), which works the same as transaction() but uses savepoints. Correct concatenation of arguments to SQLite3Error to use spaces instead of commas. Test that demo-oo1.js works with persistent OPFS storage (output differs due to persistent rows, but the demo works). (check-in: e8c323f1 user: stephan tags: fiddle-opfs) | |
| 11:16 | Improve an exception check in demo-oo1.js. (check-in: 55e1b775 user: stephan tags: fiddle-opfs) | |
|
2022-08-17
| ||
| 20:18 | In the ".dump" command of the CLI, if a schema statement ends with an unterminated comment, try to terminate that comment prior to appending the ";" at the end. Forum post d7be961c5c. (check-in: 96e112da user: drh tags: trunk) | |
| 18:07 | Enhance the "PRAGMA integrity_check" statement so that it verifies the rows of a WITHOUT ROWID table are in the correct order. (check-in: 62f934bf user: drh tags: trunk) | |
| 17:14 | Minor code simplification in the code generator for PRAGMA integrity check. (check-in: a3b9c7a6 user: drh tags: trunk) | |
| 16:44 | Minor cleanups, reorgs, and doc updates for the JS APIs. Renamed sqlite3(-api)-worker.js to sqlite3(-api)-worker1.js, for symmetry with sqlite3-api-oo1.js. (check-in: f5059ee6 user: stephan tags: fiddle-opfs) | |
|
2022-08-16
| ||
| 17:29 | wasm: minor cleanups in the OO API #1 demo. (check-in: b9cdcc06 user: stephan tags: fiddle-opfs) | |
| 17:18 | Change the legacy RC4-based PRNG to use the RFC-7539 chacha20 algorithm. (check-in: a0f80115 user: drh tags: trunk) | |
| 16:57 | Fix obsolete comments. Add new comments. Fix non-standard spacing. (Closed-Leaf check-in: a0d224c6 user: drh tags: chacha20-prng) | |
| 16:40 | Fix an off-by-one error in the ChaCha20 initialization code. (check-in: 72e220ee user: drh tags: chacha20-prng) | |
| 16:36 | wasm: add a small demo/presentation app for JS OO API #1 and make a few minor additions to that API. (check-in: d6d79b66 user: stephan tags: fiddle-opfs) | |
| 16:16 | wasm: move another file and update testing1/testing2 to account for [e38d00c2b82d]. Disable wasmfs by default as it breaks the worker-based module loader (reason as yet unknown). (check-in: 6dad5e05 user: stephan tags: fiddle-opfs) | |
| 16:11 | wasm: accommodate moving generated sqlite3.js in the previous checkin. (check-in: 41762f95 user: stephan tags: fiddle-opfs) | |
| 16:06 | wasm: disable shared cache mode by default. Experimentally move wasm-build generated files up one dir to rule the extra dir out as a problem for Emscripten-related worker-loading failures. (check-in: e38d00c2 user: stephan tags: fiddle-opfs) | |
| 14:09 | Omit an unnecessary test case from corruptN.test that does not work with the new RFC-7539 PRNG. (check-in: d9e8c65e user: drh tags: chacha20-prng) | |
| 10:52 | Add assert() statements to verify that the SQLITE_OPEN_EXCLUSIVE flag is always passed to the VFS when opening a temporary file. (check-in: e123da49 user: dan tags: trunk) | |
| 00:04 | Replace the RC4-based PRNG with one based on ChaCha20 (RFC 7539). 3x faster. (check-in: 084d8776 user: drh tags: chacha20-prng) | |
|
2022-08-15
| ||
| 19:23 | Extra test for SQLITE_MAX_COMPOUND_SELECT. (check-in: c2710967 user: dan tags: trunk) | |
| 18:36 | The new --script option to dbtotxt.c looks for an SQL script at the head of the file and generates a CLI script that will construct the database then run the SQL. (check-in: 6170e638 user: drh tags: trunk) | |
| 12:26 | In the documentation emphasize that the use of shared cache is discouraged. Fix test cases so that they all work with shared cache disabled. (check-in: 52c3ae06 user: drh tags: trunk) | |
|
2022-08-13
| ||
| 17:13 | Add worker-style variant of the tests added in [ae24ac0f7dd9], but building this with wasmfs causes them to throw inexplicable exceptions from the Emscripten glue (without wasmfs it builds and runs fine, but storage is not persistent). (check-in: 6401595e user: stephan tags: fiddle-opfs) | |
| 16:36 | wasmfs: use unix-none VFS by default to avoid locking errors in non-OPFS mode. (check-in: 75561dea user: stephan tags: fiddle-opfs) | |
| 16:11 | Add scratchpad/test app for WASMFS/OPFS running in the main window thread. Enable WASMFS by default in the library build. (check-in: ae24ac0f user: stephan tags: fiddle-opfs) | |
| 13:56 | Cleanups in the wasmfs/opfs integration but disable it in order to get the build into a known-working state before continuing with experimentation. (check-in: 41045be7 user: stephan tags: fiddle-opfs) | |
| 13:51 | Remove OPFS from the fiddle build for the time being - will re-enable once the breakage is figured out via testing with the core API. (check-in: 3bc510a6 user: stephan tags: fiddle-opfs) | |
| 13:46 | wasm OO API #1: added DB.callInTransaction() and Stmt.stepFinalize(). (check-in: e37dddc1 user: stephan tags: fiddle-opfs) | |
| 13:42 | Corrected TextDecoder.decode() usage to run when its input references a SharedArrayBuffer. (check-in: d4d77340 user: stephan tags: fiddle-opfs) | |
|
2022-08-12
| ||
| 18:54 | Use new -DSQLITE_DEFAULT_UNIX_VFS="unix-none" for fiddle build to bypass OPFS locking errors. (check-in: b3a93ec7 user: stephan tags: fiddle-opfs) | |
| 18:46 | Add the (undocumented) SQLITE_DEFAULT_UNIX_VFS compile-time option. (check-in: 49828bde user: drh tags: trunk) | |
| 18:07 | Updated fiddle deps to ensure that sqlite3.c is built first. (check-in: f0ca0261 user: stephan tags: fiddle-opfs) | |
| 17:57 | Build fiddle with WASMFS OPFS support and attempt to use it if available. It does not work because of an inexplicable exception in Emscripten-generated code and perpetually-locked db, but it's not yet clear why. (check-in: a16f0a46 user: stephan tags: fiddle-opfs) | |
| 17:55 | Minor wasm-related doc clarification and remove an obsolete code comment. (check-in: 1b1f650a user: stephan tags: trunk) | |
| 13:07 | wasm: document the role of sqlite3-wasm.c. Other minor doc updates. (check-in: 4c10b9b1 user: stephan tags: trunk) | |
| 09:32 | wasm: corrected (in as-yet unused code) a flags check. (check-in: 06b61137 user: stephan tags: trunk) | |
|
2022-08-11
| ||
| 15:45 | Wasm-related doc additions and tweaks. (check-in: ec55a3aa user: stephan tags: trunk) | |
| 09:18 | Merge in wasm-cleanups branch, reorganizing and updating the wasm-related components. (check-in: c072594d user: stephan tags: trunk) | |
|
2022-08-10
| ||
| 18:40 | Fix another harmless comment typo that causes a typo in the documentation. (check-in: bb084adb user: drh tags: trunk) | |
| 18:33 | Remove (harmless) duplicate words in comments used to generate parts of the documentation. forum post 54abbe9ccd (check-in: b5e4f0db user: drh tags: trunk) | |
| 15:29 | Fix handling of columns with names that are SQL keywords in the ".expert" command. (check-in: 5e1b8221 user: dan tags: trunk) | |
| 13:22 | wasm opfs: error handling fix for an impossible-to-reach error case. Minor cosmetic tweaks in the wasm JSON enum. (Closed-Leaf check-in: 683a3b93 user: stephan tags: wasm-cleanups) | |
| 11:26 | wasm refactoring part 2 of (apparently) 2: moved ext/fiddle/... into ext/wasm and restructured the core API-related parts of the JS/WASM considerably. (check-in: 27f9da4e user: stephan tags: wasm-cleanups) | |
| 09:36 | wasm/fiddle refactoring part 1 of N: move fiddle app from ext/fiddle to ext/wasm/fiddle, which only contains files intended to be pushed to the live site. Disabled build of the non-fiddle wasm parts, pending a later step of the refactoring. (check-in: fb4eb930 user: stephan tags: wasm-cleanups) | |
| 07:58 | Merged in trunk for pending tree refactoring. (check-in: c3a3cb01 user: stephan tags: wasm-cleanups) | |
|
2022-08-09
| ||
| 16:13 | Allow the name of an index to collide with a table in a different schema. (check-in: f963c252 user: drh tags: trunk) | |
| 14:28 | Omit the (undocumented) SQLITE_PCACHE_SEPARATE_HEADER compile-time option. (check-in: 60947da6 user: drh tags: trunk) | |
|
2022-08-08
| ||
| 19:29 | Avoid some cases of signed integer overflow in fts5 by casting to unsigned values. (check-in: 46a78c8c user: dan tags: trunk) | |
| 16:25 | Avoid trying to cast an over-sized floating point value into an integer. (check-in: 3518cd7c user: drh tags: trunk) | |
| 13:04 | Remove an unreachable branch in the numericType() helper routine. (check-in: a99cc008 user: drh tags: trunk) | |
| 12:19 | Do not attempt to convert an oversized floating point value into an integer. (check-in: 6c4fc838 user: drh tags: trunk) | |
|
2022-08-06
| ||
| 15:28 | Fix a rounding error caused by scalar->logarithm->scalar conversion when using stat4 data to estimate some range scans. (check-in: 86e30d7c user: dan tags: trunk) | |
|
2022-08-05
| ||
| 19:37 | Avoid having fts3 read uninitialized values when processing deferred tokens. (check-in: 8e2bfe58 user: drh tags: branch-3.39) | |
| 17:16 | Avoid having fts3 read uninitialized values when processing deferred tokens. (check-in: 98729c5f user: dan tags: trunk) | |
| 05:30 | Fix harmless compiler warning seen with MSVC. (check-in: e36217f2 user: mistachkin tags: trunk) | |
|
2022-08-04
| ||
| 17:17 | Fix a problem with the query optimizer for LIMIT/OFFSET queries when underlying query is a UNION ALL and both arms of the UNION ALL are subqueries with an ORDER BY clause. (check-in: 5e1db87c user: drh tags: branch-3.39) | |
| 17:15 | Fix a problem with the query optimizer for LIMIT/OFFSET queries when underlying query is a UNION ALL and both arms of the UNION ALL are subqueries with an ORDER BY clause. This bug was reported at forum post 6b5e9188f0657616. The problem was introduced in 2015 (SQLite version 3.9.0) by check-in [4b631364354068af]. See also ticket [b65cb2c8d91f6685]. (check-in: 6c806f64 user: drh tags: trunk) | |
| 14:02 | Fix a typo on a comment used to generate bytecode documentation. (check-in: 64348d0b user: drh tags: trunk) | |
|
2022-08-03
| ||
| 19:53 | Performance optimization to the math opcodes of the bytecode engine. (check-in: 597347d9 user: drh tags: trunk) | |
| 19:37 | Small performance increase on the binding interfaces. (check-in: aab24c37 user: drh tags: trunk) | |
| 17:27 | Improvement on the previous check-in. (check-in: 0d0c3111 user: drh tags: branch-3.39) | |
| 17:26 | Improvement on the previous check-in. (check-in: c1ab0ea2 user: drh tags: trunk) | |
| 16:04 | For an IN operator used with a RIGHT JOIN, use the number of terms in the vector, not the number of equality terms, to size the column map. dbsqlfuzz 14cfdad6ca45e607163f54049ddf5065183dc657. (check-in: 003e4eee user: drh tags: branch-3.39) | |
| 15:58 | For an IN operator used with a RIGHT JOIN, use the number of terms in the vector, not the number of equality terms, to size the column map. dbsqlfuzz 14cfdad6ca45e607163f54049ddf5065183dc657, CVE-2022-3195. This problem was injected by [a193749730d6cfba] which was part of the RIGHT JOIN enhancement for version 3.39.0. (check-in: b184c8d9 user: drh tags: trunk) | |
|
2022-08-01
| ||
| 13:17 | Bump the version number up to 3.39.3. (check-in: d0f7f22a user: drh tags: branch-3.39) | |
| 13:15 | In the xUpdate method of the GeoPoly virtual table, make sure that the number of updated columns does not exceed the underlying implementation, even if the virtual table object records an excess number of column in the nAux field due to table constraints in the table definition. Fix for the problem reported by forum post a096ab7d96bb057a. (check-in: a8ae7216 user: drh tags: branch-3.39) | |
| 13:14 | In the xUpdate method of the GeoPoly virtual table, make sure that the number of updated columns does not exceed the underlying implementation, even if the virtual table object records an excess number of column in the nAux field due to table constraints in the table definition. Fix for the problem reported by forum post a096ab7d96bb057a. A test case for this problem is in dbsqlfuzz. (check-in: a04dd475 user: drh tags: trunk) | |
|
2022-07-26
| ||
| 19:10 | Restore a VDBE coverage macro that is mistakenly deleted for check-in [92ac01d41d46ab73]. (check-in: 01de7ec4 user: drh tags: trunk) | |
| 18:43 | Fix another problem related to aggregate query LEFT JOIN flattening. (check-in: 8e12c6f5 user: dan tags: trunk) | |
| 15:41 | Omit a unused variable initialization in order to suppress a harmless compiler warning. (check-in: 2d703c5e user: drh tags: trunk) | |
| 15:39 | Add test cases related to fix [e717e029]. (Closed-Leaf check-in: 2909f36b user: dan tags: flatten-left-join) | |
| 15:32 | Make sure IF_NULL_ROW expressions receive a separate slot in the sorter used to implement GROUP BY. (check-in: 2bda4fca user: drh tags: flatten-left-join) | |
| 14:37 | Improved AggInfo.aCol debugging output, intended to debug a problem with LEFT JOIN flattening into an aggregate query with GROUP BY. (check-in: e717e029 user: drh tags: flatten-left-join) | |
| 10:16 | Add a OOM term to a single assert() statement. (check-in: 217fc3eb user: drh tags: trunk) | |
| 01:20 | Performance optimization: avoid unnecessary calls to computeLimitRegisters(). (check-in: f48bd8f8 user: drh tags: trunk) | |
|
2022-07-25
| ||
| 23:34 | Fix an assert() that was made out-of-date by [b52393ac28debe98]. (check-in: db8230e8 user: drh tags: trunk) | |
| 23:01 | Performance optimization by only invoking sqlite3FkCheck() when it is actually needed. (check-in: 98b0e830 user: drh tags: trunk) | |
| 22:31 | Performance optimization in sqlite3ViewGetColumnNames(). (check-in: 390717e6 user: drh tags: trunk) | |
| 22:02 | Fix a harmless compiler warning. (check-in: 9aba7417 user: drh tags: trunk) | |
| 21:37 | In-line a call to sqlite3ExprCode() in insert.c, for a size reduction and performance increase. (check-in: 35066b14 user: drh tags: trunk) | |
| 20:21 | Performance optimization in computing the Expr.nHeight field. (check-in: 1798ce97 user: drh tags: trunk) | |
| 19:05 | Small performance increase and size reduction by splitting out the sqlite3VdbeGetLastOp() from sqlite3VdbeGetOp(). (check-in: 92ac01d4 user: drh tags: trunk) | |
| 16:06 | Fix an error in the aggregate query LEFT JOIN flattening optimization from [2cf373b10c9bc4cb] and further enhance that optimization so that it works even if there is a GROUP BY clause. (check-in: b52393ac user: drh tags: trunk) | |
| 15:54 | Allow subqueries on the right-hand side of a LEFT JOIN to be flattened even if they contain a GROUP BY clause. (Closed-Leaf check-in: 816da9a8 user: drh tags: flatten-left-join) | |
| 14:05 | TK_IF_NULL_ROW expressions must be accumulated in the same way as TK_COLUMN expressions in an aggregate query. Proposed fix for the problem identifyed by dbsqlfuzz 8e17857db2c5a9294c975123ac807156a6559f13. (check-in: 40d08807 user: drh tags: flatten-left-join) | |
| 11:19 | If the LHS of a LEFT JOIN is flattened into an aggregate parent query, avoid executing OP_IfNullRow on an unopened cursor by disallowing the index-only optimization. (Later:) Does not work if automatic indexes are disabled. (Closed-Leaf check-in: dbe522b0 user: dan tags: nonworking-flatten-left-join) | |
| 11:04 | Reduce a timeout in walsetlk.test from 2000ms to 1100ms so that the test runs a bit faster. (check-in: 836fa097 user: dan tags: trunk) | |
|
2022-07-23
| ||
| 12:51 | Simplifications to sqlite3FinishCoding() for a small size reduction and performance increase. (check-in: a995614b user: drh tags: trunk) | |
| 00:53 | Remove a branch that is no longer reachable due to the previous check-in. (check-in: 8b4d1b93 user: drh tags: trunk) | |
| 00:44 | Use sqlite3ParserAddCleanup() rather than pParse->pConstExpr to implement sqlite3ExprDeferredDelete(). This is a better solution than check-in [c538d07535092722]. (check-in: 2a6f6971 user: drh tags: trunk) | |
|
2022-07-22
| ||
| 20:48 | Updates to the 'vtshim' extension. (Leaf check-in: 163c6cdf user: mistachkin tags: xShadowName2) | |
| 20:24 | Add experimental 'xShadowName2' method for virtual table modules. (check-in: 57beb700 user: mistachkin tags: xShadowName2) | |
| 19:28 | Omit the EP_MemToken flag that was made obsolete by [e1f1cfe7f4387b60], for a size reduction and performance increase. (check-in: 28934a9d user: drh tags: trunk) | |
| 18:25 | In a TK_BLOB Expr node, the Expr.zToken might not be a well-formed BLOB literal if there has been a prior OOM. dbsqlfuzz 23871e5805d6c45b392f9b7aa1e8a2b98f3c27cd. (check-in: c538d075 user: drh tags: trunk) | |
| 14:52 | Update the documentation for SQLITE_OPEN_NOFOLLOW to state more clearly that nothing in the database path is allowed to be a symbolic link. (check-in: de922269 user: drh tags: trunk) | |
|
2022-07-21
| ||
| 18:37 | Fix harmless compiler warning seen with MSVC. (check-in: 648172de user: mistachkin tags: trunk) | |
| 16:07 | Merge the fixes from branch-3.39 into the reuse-schema branch. (check-in: 578538ba user: drh tags: reuse-schema-3.39) | |
| 15:24 | Version 3.39.2 (check-in: 698edb77 user: drh tags: release, branch-3.39, version-3.39.2) | |
| 12:26 | In the query planner, restore the former aggressiveness in reordering of FROM clause terms that existed prior to version 3.39.0 for queries that contain no RIGHT or FULL JOINs. (check-in: 9141e873 user: drh tags: branch-3.39) | |
|
2022-07-20
| ||
| 20:36 | Make use of the sqlite3ExprDeferredDelete() interface in the previous check-in, and in another place where it might be helpful. (check-in: 22f90e96 user: drh tags: trunk) | |
| 17:01 | Simplify the logic that converts the "1" expression in "ORDER BY 1" into a copy of the expression that defines the first output column. (check-in: 44993591 user: drh tags: branch-3.39) | |
| 16:42 | Simplify the logic that converts the "1" expression in "ORDER BY 1" into a copy of the expression that defines the first output column. Fix for CVE-2022-3034. Test case in TH3. (check-in: e1f1cfe7 user: drh tags: trunk) | |
| 10:09 | Merged in trunk. (check-in: d662796c user: stephan tags: wasm-cleanups) | |
|
2022-07-19
| ||
| 21:12 | Improve accuracy of julian day milliseconds calculation. (check-in: e5e93118 user: larrybr tags: trunk) | |
|
2022-07-18
| ||
| 19:32 | Remove a few unsuitable scripts from the "veryquick" test suite. Also have every second testrunner.tcl process favour running test scripts that contain text like "testrunner: slow" before any others. (check-in: 22d280a5 user: dan tags: trunk) | |
| 18:13 | Add the "testrunner" makefile target. (check-in: 954c6593 user: drh tags: trunk) | |
| 15:27 | Increase the size of loop variables in the printf() implementation to avoid integer overflow on multi-gigabyte string arguments. CVE-2022-35737. (check-in: 26db4fc2 user: drh tags: branch-3.39) | |
| 15:02 | Increase the size of loop variables in the printf() implementation to avoid integer overflow on multi-gigabyte string arguments. CVE-2022-35737. (check-in: aab790a1 user: drh tags: trunk) | |
| 13:55 | Enhance the REGEXP extension so that it will accept the start-of-input mark ("^") in the middle of parentheses. Forum post 0d6a9160f81ef1a8. (check-in: ed8a8ebd user: drh tags: trunk) | |
| 13:10 | Fix a problem in the REGEXP extension for the {M,N} construct where M is zero. See forum post 8694e55a2c29963c for more information. (check-in: af15bb75 user: drh tags: trunk) | |
| 11:44 | Enhance the ext/misc/regexp.c code so that when it is compiled with SQLITE_DEBUG, a new function named regexp_bytecode() is available that prints out the compiled NFA as human-readable text, for debugging purposes. (check-in: cb5c0897 user: drh tags: trunk) | |
|
2022-07-16
| ||
| 18:08 | Fixes for the generated "mallocs.tcl" and "leaks.tcl" scripts generated by running tcl tests with the --malloctrace=1 option. (check-in: 449799e2 user: dan tags: trunk) | |
|
2022-07-15
| ||
| 20:39 | In the query planner, restore the former aggressiveness in reordering of FROM clause terms that existed prior to version 3.39.0 for queries that contain no RIGHT or FULL JOINs. (check-in: 92d60b64 user: drh tags: trunk) | |
| 15:24 | Fix harmless compiler warnings seen with MSVC. (check-in: 2be8fa15 user: mistachkin tags: branch-3.39) | |
| 15:11 | Fix a memory leak in fts3 that could occur when processing a corrupt database. (check-in: 48f323b9 user: dan tags: branch-3.39) | |
| 15:08 | Fix a memory leak in fts3 that could occur when processing a corrupt database. (check-in: d74f6f6d user: dan tags: trunk) | |
| 12:34 | Fix the whereKeyStats() routine (part of STAT4 processing only) so that it is able to cope with row-value comparisons against the primary key index of a WITHOUT ROWID table. Forum post 3607259d3c. (check-in: 2a6f7618 user: drh tags: branch-3.39) | |
| 12:16 | Fix the whereKeyStats() routine (part of STAT4 processing only) so that it is able to cope with row-value comparisons against the primary key index of a WITHOUT ROWID table. Forum post 3607259d3c. (check-in: 0620e419 user: drh tags: trunk) | |
| 12:07 | Update some faulty assert() statements in fts3. (check-in: 958d104b user: dan tags: branch-3.39) | |
| 11:34 | Update some faulty assert() statements in fts3. (check-in: b072851b user: dan tags: trunk) | |
|
2022-07-14
| ||
| 21:17 | Ensure all testrunner.tcl processes use the same pending-byte value when accessing testrunner.db. Otherwise locking doesn't work and the db is corrupted. (check-in: b6522565 user: dan tags: trunk) | |
| 18:09 | Update testrunner.tcl so that it can run the test suites defined in permutation.test. (check-in: 15ce937e user: dan tags: trunk) | |
| 01:54 | Bump the version number up to 3.39.2. (check-in: c4dbcb28 user: drh tags: branch-3.39) | |
| 01:49 | When applying the omit-ORDER-BY optimization, defer deleting the AST of the deleted ORDER BY clause until after code generation ends. (check-in: b88d6c4b user: drh tags: branch-3.39) | |
| 01:48 | When applying the omit-ORDER-BY optimization [85ddaf1b59a19cbd], defer deleting the AST of the deleted ORDER BY clause until after code generation ends. Fix for CVE-2022-3039. Test case in TH3. (check-in: f22f95b8 user: drh tags: trunk) | |
|
2022-07-13
| ||
| 21:28 | Fix testrunner.tcl so that it can detect the number of logical cores on osx. (check-in: 14918f28 user: dan tags: trunk) | |
| 21:10 | Minor change to help message in testrunner.tcl. (check-in: e4f9cb01 user: dan tags: trunk) | |
| 21:02 | Add new script test/testrunner.tcl. For running a set of test scripts using multiple processes. (check-in: 0122e93d user: dan tags: trunk) | |
| 20:26 | Fix a problem preventing "testrunnter.tcl all" from working. (Closed-Leaf check-in: 0ed1e83c user: dan tags: testrunner) | |
| 19:57 | Update testrunner.tcl to allow the user to specify which tests to run on the command line. (check-in: 900febcf user: dan tags: testrunner) | |
| 19:41 | Version 3.39.1 (check-in: 7c16541a user: drh tags: release, branch-3.39, version-3.39.1) | |
| 17:46 | Fix testrunner.tcl so that it checks for memory leaks. (check-in: 106f6724 user: dan tags: testrunner) | |
| 16:06 | The query flattener should not run if the subquery is a compound that contains a RIGHT JOIN in any arm and the subquery is not the first element of the outer query. Otherwise, prior elements of the outer query will not have the JT_LTORJ flag set. Fix for the problem reported in forum post 174afeae5734d42d. (check-in: cf9ed7f2 user: drh tags: branch-3.39) | |
| 15:52 | The query flattener should not run if the subquery is a compound that contains a RIGHT JOIN in any arm and the subquery is not the first element of the outer query. Otherwise, prior elements of the outer query will not have the JT_LTORJ flag set. Fix for the problem reported in forum post 174afeae5734d42d. (check-in: 274e244c user: drh tags: trunk) | |
| 11:27 | Fix test case error messages so that they work with after the removal of the unnecessary zErrMsg clearing from [44d77a7f807f5dc3]. (check-in: 7b167b87 user: drh tags: branch-3.39) | |
|
2022-07-12
| ||
| 20:31 | Add test/testrunner.tcl, an experimental script for distributing the work of veryquick.test between multiple processes. (check-in: ef229cbb user: dan tags: testrunner) | |
| 15:53 | Renamed the SQLITE_SHELL_WASM_WEB_MODE to SQLITE_SHELL_FIDDLE, which seems to be more in line with project convensions and indicates that that flag is only intended for /fiddle mode, as opposed to arbitrary wasm-on-the-web use. (check-in: d1d019bf user: stephan tags: trunk) | |
| 15:17 | Fix another test case error message similar to those fixed in [b3d6b3c3]. (check-in: 6d0f6772 user: dan tags: trunk) | |
| 15:12 | Update makefiles to fix building the non-amalgamation testfixture with SQLITE_DEBUG. (check-in: 1ca70564 user: dan tags: branch-3.39) | |
| 15:10 | Update makefiles to fix building the non-amalgamation testfixture with SQLITE_DEBUG. (check-in: d9c4a9d0 user: dan tags: trunk) | |
| 10:46 | Fix a bug in wapptest.tcl introduced by [51255bad4c1fb607]. (check-in: b26d097e user: drh tags: trunk) | |
| 09:40 | Renamed SQLITE_SHELL_WASM_MODE to SQLITE_SHELL_WASM_WEB_MODE and no longer automatically enable it if __EMSCRIPTEN__ is defined, in order to facilitate using Emscripten to build the shell for CLI-based WASM runtimes (which cannot make use of the web-specific user input changes). The fiddle build now explicitly passes the new flag on at compile-time. (check-in: ee059ad5 user: stephan tags: trunk) | |
| 07:13 | Back out the pager performance enhancement at [a1c090e08139f99d3], because it turns out we should never allow a zero key into the pcache interface according to the design specs, even if that page is immediately released without ever being used. (check-in: ec96293e user: drh tags: trunk) | |
|
2022-07-11
| ||
| 22:20 | Add the --config option to wapptest.tcl. The argument is a glob pattern. All configurations must match the glob pattern in order to run. The default value is "*". (check-in: 51255bad user: drh tags: trunk) | |
| 21:40 | Update the expected error messages in some OOM test cases to account for [44d77a7f807]. (check-in: b3d6b3c3 user: dan tags: trunk) | |
| 19:48 | Use 64-bit memory allocation APIs in the sqlite_stmt virtual table, to avoid harmless compiler warnings. (check-in: 4c40b38c user: drh tags: branch-3.39) | |
| 19:47 | Use 64-bit memory allocation APIs in the sqlite_stmt virtual table, to avoid harmless compiler warnings. (check-in: afb9e60e user: drh tags: trunk) | |
| 19:29 | Fix harmless compiler warnings. (check-in: 9931bb20 user: drh tags: branch-3.39) | |
| 19:12 | In the sqlite_stmt extension, store the result of strlen() in a 64-bit integer to avoid a compiler warning, even though we know that the length will always fit comfortably in 32 bits. (check-in: 3fe19452 user: drh tags: trunk) | |
| 18:26 | Fix harmless compiler warnings about unused debugging functions in treeview.c. (check-in: 4d6f9077 user: drh tags: trunk) | |
| 18:11 | Back out the optimization at [1a8c2e54375ee2cf7] because there are some cases where it does not work. (check-in: fe39c8d5 user: drh tags: trunk) | |
| 14:44 | Increase the version number to 3.39.1. (check-in: 88b1b032 user: drh tags: branch-3.39) | |
| 14:43 | Ensure that the Parse.nErr flag is set following an SQLITE_TOOBIG error on a nested parse. Fix for the problem identified by forum post d5a82ba9eedee30c. Also, remove unnecessary clearing of the Parse.zErrMsg field following a nested parse. (check-in: eaef7b71 user: drh tags: branch-3.39) | |
| 14:39 | Fix a problem in fts3 to do with deferred tokens and OR expressions. (check-in: bbe999e3 user: drh tags: branch-3.39) | |
| 14:36 | Ensure that the Parse.nErr flag is set following an SQLITE_TOOBIG error on a nested parse. Fix for the problem identified by forum post d5a82ba9eedee30c. Also, remove unnecessary clearing of the Parse.zErrMsg field following a nested parse. (check-in: 44d77a7f user: drh tags: trunk) | |
| 14:26 | Fix a problem in fts3 to do with deferred tokens and OR expressions. (check-in: d0bfe5c5 user: dan tags: trunk) | |
|
2022-07-10
| ||
| 21:12 | When an OOM occurs and sets the Parse.nErr value, also set the Parse.nErr value for all outer Parse objects. dbsqlfuzz d33f60aaa67733aa700cd69dacf8e0e23a327a29 (check-in: 9a494d25 user: drh tags: trunk) | |
|
2022-07-08
| ||
| 20:03 | Performance optimizations in the WHERE clause processing of the query planner. (check-in: 50c8e8de user: drh tags: trunk) | |
| 18:23 | Omit an unnecessary structure initialization in whereLoopAddAll(), replacing it with an assert() to show that the initialization has already occurred, for a small performance increase and size reduction. (check-in: 27be9e96 user: drh tags: trunk) | |
| 17:57 | Size reduction and performance optimization in whereLoopAddBtreeIndex(). (check-in: 64d2312c user: drh tags: trunk) | |
| 16:56 | In wherePathSolver(), defer initializing variables until they are actually needed, in case they are not needed. This gives a small performance increase. (check-in: 6f28a965 user: drh tags: trunk) | |
| 15:06 | Upgrade the TEA build system used to build the Tcl package. (check-in: ace65da8 user: dan tags: trunk) | |
| 13:57 | Fix typo in autoconf/tea/win/makefile.vc. (Closed-Leaf check-in: 094b2aad user: dan tags: tea-upgrade) | |
| 11:55 | Remove an obsolete statement from a comment. No changes to code. (check-in: 99afb3f4 user: drh tags: trunk) | |
| 11:21 | Update options in the TEA package msvc makefile to match other builds. (check-in: d72e12b9 user: dan tags: tea-upgrade) | |
| 11:14 | Update the build options used for the TEA package to match those used to build the binaries at download.html. (check-in: 9d6d9dba user: dan tags: tea-upgrade) | |
|
2022-07-07
| ||
| 22:59 | Very small size reduction and performance increase in btree.c. (check-in: 9a7c031a user: drh tags: trunk) | |
| 21:04 | Size reduction and performance increase in defragementPage() of btree.c. (check-in: 1b03f197 user: drh tags: trunk) | |
| 20:49 | Upgrade the TEA build system in autoconf/tea/. To match tclconfig commit 20fe9e6f59 and Tcl Sample Extension be47fb0446. (check-in: 1531f739 user: dan tags: tea-upgrade) | |
| 20:29 | Size reduction and performance increase in the pageFindSlot() routine of btree.c. (check-in: 5d247e38 user: drh tags: trunk) | |
| 20:11 | Size reduction and performance increase in the freeSpace() routine of btree.c. (check-in: 7d7aed05 user: drh tags: trunk) | |
|
2022-07-06
| ||
| 23:50 | Fix harmless compiler warnings seen with MSVC. (check-in: 61e2094a user: mistachkin tags: trunk) | |
| 15:44 | Avoid ignoring the last line of a csv file if the final field is empty and there is no trailing CFLS. Also have the csv extension treat the last line of a file in the same way as any other line if it is short fields. (check-in: 587795d4 user: dan tags: trunk) | |
| 13:59 | Avoid dropping error codes in the xBegin() method of virtual table sqlite_dbpage. (check-in: 570e2bce user: dan tags: trunk) | |
|
2022-07-05
| ||
| 19:56 | Fix for builds with both SQLITE_DEBUG and SQLITE_OMIT_WINDOWFUNC defined. (check-in: dfe46de2 user: dan tags: branch-3.39) | |
| 19:53 | Fix for builds with both SQLITE_DEBUG and SQLITE_OMIT_WINDOWFUNC defined. (check-in: 77916947 user: dan tags: trunk) | |
| 17:49 | Fix a problem in the csv extension that was triggered when the very first field in the csv file is zero bytes in size. (check-in: b12ddabf user: dan tags: trunk) | |
| 10:40 | Fix an assert() in btree to be more precise, as the previous form of that asser might fail due to recent performance optimizations. (check-in: 47749381 user: drh tags: trunk) | |
|
2022-07-04
| ||
| 15:14 | Size reduction and performance optimization in sqlite3WalFindFrame(). (check-in: 1a8c2e54 user: drh tags: trunk) | |
| 09:41 | In the getNormalPage() routine of pager.c, consolidate pgno error checking into a single spot for small size reduction and performance increase. (check-in: a1c090e0 user: drh tags: trunk) | |
|
2022-07-03
| ||
| 18:12 | Enhance the REGEXP extension so that the end-of-input indicate ("$") is allowed to occur on one branch of an OR ("|"). Forum post 0107d5d40dd273e2, second issue. (check-in: 3c04d21e user: drh tags: trunk) | |
| 14:32 | Fix the initial-prefix optimization for the REGEXP extension such that it works even if the prefix contains characters that require a 3-byte UTF8 encoding. This should fix the problem reported by forum post 96692f8ba5. (check-in: 7a32cccc user: drh tags: branch-3.39) | |
| 14:25 | Fix the initial-prefix optimization for the REGEXP extension such that it works even if the prefix contains characters that require a 3-byte UTF8 encoding. This should fix the problem reported by forum post 96692f8ba5. (check-in: c94595a6 user: drh tags: trunk) | |
| 11:16 | Improved comment on sqlite3VdbeSwap(). No changes to code. (check-in: 6a8e4fb7 user: drh tags: trunk) | |
| 11:12 | Remove debugging code that has always been commented out. Cosmetic change. (check-in: b5406417 user: drh tags: trunk) | |
|
2022-07-01
| ||
| 21:03 | Do not attempt the OP_Count optimization on queries with HAVING clauses. This fixes a problem exposed by [2cf373b10c9bc4cb]. (check-in: 566b7842 user: dan tags: trunk) | |
| 19:42 | Performance improvement in resolveP2Values(). Save over 1 million CPU cycles by omitting the loop termination condition and exiting when the OP_Init (always the first opcode in any VDBE program) is encountered. (check-in: bb179140 user: drh tags: trunk) | |
| 12:09 | In the sqlite_stmt virtual table, add two ALWAYS() macros on branches that have been come always true due to the [84a91c255e3d7772] optimization. REVISED: Dbsqlfuzz found a case where the ALWAYS() is false. (Closed-Leaf check-in: bf34eddb user: drh tags: mistake) | |
|
2022-06-30
| ||
| 22:46 | Small performance improvement to sqlite3_finalize(). (check-in: 8a6913b6 user: drh tags: trunk) | |
| 14:19 | Fix documentation and test-script typos and a dependency problem on a Makefile. Fix the sqlite_stmt extension virtual table so that it shows the state of all prepared statements for a single instant in time. (check-in: 0a9e08be user: drh tags: branch-3.39) | |
| 11:01 | Add missing "finish_test" command to the end of test script merge1.test. (check-in: 13cb3f1e user: dan tags: trunk) | |
|
2022-06-29
| ||
| 15:16 | Have the sqlite_stmt table buffer all data for the current scan within the xFilter method. (check-in: 84a91c25 user: dan tags: trunk) | |
|
2022-06-27
| ||
| 21:42 | Fix a problem causing false corruption reports following recovery of a *-wal file that contains one or more transactions and a *-wal2 file that contains a valid header but no valid transactions. (check-in: f6eafb65 user: dan tags: wal2) | |
| 11:28 | Fix a missing dependency for json.lo in Makefile.in (check-in: 65930a5c user: drh tags: trunk) | |
|
2022-06-25
| ||
| 21:41 | wasm: corrected the isInt32() check to account for negative values. (check-in: 7223f4bb user: stephan tags: wasm-cleanups) | |
| 20:32 | Put an ALWAYS() on an unreachable branch. (check-in: 58caa50a user: drh tags: trunk) | |
| 19:43 | Allow flattening of a subquery that is the right operand of a LEFT JOIN in an aggregate query as long as there is no GROUP BY clause. (The GROUP BY clause will interfere with the operation of the TK_IF_NULL_ROW expression nodes.) (check-in: 2cf373b1 user: drh tags: trunk) | |
| 19:30 | Increase the version number to 3.40.0 so as to begin the next development cycle. (check-in: 4542e323 user: drh tags: trunk) | |
| 19:07 | wasm: added bindings for sqlite3_extended_result_codes(), sqlite3_open_v2(), and the SQLITE_OPEN_... flags. (check-in: ac876ab9 user: stephan tags: wasm-cleanups) | |
| 18:55 | Fix documentation type. forum post 8d900996ed. (check-in: 869061f1 user: drh tags: trunk) | |
| 18:18 | wasm: lots of doc additions and refactoring. Refactored the WASM memory heap usage to hopefully eventually account for a runtime-growable heap. Differentiate between supported TypedArray types for input SQL strings vs binding/fetching blobs. Might (untested) have implemented the ability to bind UtfNNArray values as blobs, where NN is one of 16 or 32. (check-in: e10d57df user: stephan tags: wasm-cleanups) | |
| 17:20 | Merge version 3.39.0 into the reuse-schema branch. (check-in: bee9903d user: drh tags: reuse-schema) | |
| 16:53 | Merge version 3.39.0 into the wal2 branch. (check-in: ad3a7005 user: drh tags: wal2) | |
| 14:57 | Version 3.39.0 (check-in: 14e166f4 user: drh tags: trunk, release, major-release, version-3.39.0) | |
| 10:30 | wasm: added utility C code to generate a JSON-format "enum" of the numerous SQLITE_xyz constants so that we do not risk those getting out of sync in the JS code. Renamed initSqlite3Module to sqlite3InitModule. Cleanups in the TypedArray handling. (check-in: 778062e3 user: stephan tags: wasm-cleanups) | |
| 07:42 | wasm: eliminated the need for Stmt objects to keep ahold of memory allocated for bound strings and blobs. Added alternate string/blob bind impls which are hypothetically more efficient but not yet proven to be so. (check-in: da1d3151 user: stephan tags: wasm-cleanups) | |
| 06:46 | wasm: eliminated the dependency on the deprecated emcc-provided allocate() function. Adjacent cleanups in blob binding. (check-in: 140618b2 user: stephan tags: wasm-cleanups) | |
| 03:53 | wasm: corrected the propagation of text/blob values via UDFs. DB.exec()'s sql may now be an array of strings which get concatenated together before passing it on to sqlite3_prepare_v2(). DB.exec()'s callback now applies to the first statement which has result columns instead of only the first statement. Fixed a precedence but which caused isInt32() to report false positives. (check-in: 37a8fecb user: stephan tags: wasm-cleanups) | |
| 02:54 | Minor wasm doc tweaks. (check-in: 42dc5008 user: stephan tags: wasm-cleanups) | |
| 02:39 | Change a harmless assert() into a testcase(). The actually test case code is in TH3. Fix for the problem described at forum post ed29e196d5c4f3d5. (check-in: cd6254fc user: drh tags: trunk) | |
| 02:37 | wasm binding: consolidated the two sqlite3_prepare_v2() bindings behind a single dispathcer. Various internal cleanups and refactoring. Branched because trunk is in pencils-down mode for pending 3.39 release. (check-in: ab3e50da user: stephan tags: wasm-cleanups) | |
|
2022-06-24
| ||
| 16:15 | Merge the 3.39.0 release candidate 1 changes into the reuse-schema branch. (check-in: 506a3e6f user: drh tags: reuse-schema) | |
| 15:15 | Merge 3.39.0 release candidate 1 into the wal2 branch. (check-in: 5a5e6e36 user: drh tags: wal2) | |
| 12:56 | Fix a harmless UBSAN warning associated with PRAGMA schema_version found by OSSFuzz. (check-in: e93fd170 user: drh tags: trunk) | |
| 11:05 | In sqlite3WhereBegin, do not proceed with coding the loop if an error is detected as part of WHERE expression analysis. For for the assertion-fault described by forum post c3496cf6b1. (check-in: db5266de user: dan tags: trunk) | |
| 11:02 | A minor fix to test/fuzzinvariants.c so that it works even with column names that originally contain a ':' and that are disambiguated. (check-in: 8d9b1fff user: drh tags: trunk) | |
|
2022-06-23
| ||
| 22:43 | New test cases for query flattening with LEFT JOIN. (Closed-Leaf check-in: 27f68e47 user: drh tags: flatten-left-join) | |
| 20:56 | Reintroduce flattener constraint (3c), but this time make it apply only if the outer query holds a GROUP BY, not if the outer query is an aggregate. (check-in: 641dfb91 user: drh tags: flatten-left-join) | |
| 15:15 | Add back the ability to flatten a LEFT JOIN subquery - previously removed due to ticket [cad1ab4cb7b0fc344]. (check-in: f8fe936a user: drh tags: flatten-left-join) | |
| 12:36 | Minor correction to a comment. No code changes. (check-in: 5fa00959 user: drh tags: trunk) | |
|
2022-06-22
| ||
| 18:51 | Enhance the TCL scrip that generates sqlite3-all.c so that it outputs all text in its original order. (check-in: 83ff1a28 user: drh tags: trunk) | |
| 18:33 | Fix requirements marks that changed due to typo fixes in the documentation. (check-in: 5247df05 user: drh tags: trunk) | |
| 15:55 | More comment fixes. No changes to code. (check-in: 5c9133ef user: drh tags: trunk) | |
| 14:43 | Allow more line with on the debugging output for SrcItem elements in the parse tree. This affects debugging builds only and is a no-op for production builds. (check-in: edbe24e7 user: drh tags: trunk) | |
| 14:25 | Additional enhancements to comments. No changes to code. (check-in: d9a32044 user: drh tags: trunk) | |
| 14:00 | Give dbtotxt an option to prepend .open --hexdb (check-in: afeee612 user: larrybr tags: trunk) | |
| 12:54 | Fix a typo in a comment. No changes to code. (check-in: ef9de201 user: drh tags: trunk) | |
|
2022-06-21
| ||
| 18:38 | Disable the short-cut query planner if the NOT INDEXED modifier is used. Forum post 454d706296. (check-in: bd87d107 user: drh tags: trunk) | |
| 13:41 | Allow a HAVING clause on any aggregate query, even if there is no GROUP BY clause. This brings SQLite into closer agreement with PostgreSQL and fixes the concern raised by forum post 1a7fea4651. (check-in: 9322a7c2 user: drh tags: trunk) | |
| 12:54 |
| |
|
2022-06-20
| ||
| 19:12 | Add an ALWAYS() macro to an always-true branch. (check-in: 364645d8 user: drh tags: trunk) | |
| 18:26 | Do not allow FROM-clause terms on the left side of a RIGHT or FULL JOIN to be reordered. forum post 6650cd40b5634f35. This is probably more strict that necessary to get correct behavior, but for the first release that supports RIGHT/FULL JOIN it is perhaps better to be correct than fast. A less strict constraint might be to prohibit FROM-clause terms that originate on the left side of a RIGHT JOIN from crossing from the right side to the left side of a LEFT JOIN. Revisit this later. (check-in: 238453ff user: drh tags: trunk) | |
| 17:04 | Do not allow an ON clause to references tables to its right if there is a RIGHT or FULL join anywhere in the query. Other RDBMSes prohibit this always, but SQLite must allow ON clauses to reference tables to their right for legacy compatibility, unless there is a RIGHT or FULL join someplace in the query, in which case there is no legacy to support. (check-in: e615dbe0 user: drh tags: trunk) | |
| 12:42 | The fix at [cab9b4cccd13bf0a] was incomplete, as demonstrated by forum post 57bdf2217d. This check-in should complete the fix. (check-in: fb0a23b6 user: drh tags: trunk) | |
|
2022-06-19
| ||
| 16:55 | Follow-up to check-in [0057bbb508e7662b] - ensure that the database page has been initialized prior to continuing with the optimization. If the page is not initialized, that indicates that the database is corrupt. dbsqlfuzz 09ee46becd5e6d1b2a55c9f8ad767335a90aadb0. (check-in: 11162446 user: drh tags: trunk) | |
|
2022-06-18
| ||
| 20:20 | Enable query invariant checking in fuzzcheck by default. There is no way to turn it off. Update the invariant checking logic to be consistant with dbsqlfuzz. (check-in: 66ca729b user: drh tags: trunk) | |
| 14:50 | Abandon a query-invariant check in fuzzcheck if the call to sqlite3_bind_value() returns anything other than SQLITE_OK or SQLITE_RANGE. (check-in: d31e1cd2 user: drh tags: trunk) | |
| 10:26 | In the --query-invariants option of fuzzcheck, correctly deal with OOMs causing the return value of sqlite3_column_name() to be NULL. (check-in: eabbee4a user: drh tags: trunk) | |
|
2022-06-17
| ||
| 21:31 | Fix the OP_Concat operator such that when concatenating a BLOB with an odd number of bytes on a database that is UTF16, the size of the resulting string is reduced to a multiple of two. (check-in: 5eb2c236 user: drh tags: trunk) | |
| 17:11 | Omit the --query-invariants processing in fuzzcheck for queries that contain the implies_nonnull_row() test function. (check-in: 0602a084 user: drh tags: trunk) | |
| 16:52 | Do not run --query-invariants on statements that make use of sqlite_offset(). (check-in: c5ac4cbf user: drh tags: trunk) | |
| 16:39 | Various fixes and improvements to the --query-invariants option of fuzzcheck. (check-in: 09aca89b user: drh tags: trunk) | |
| 16:32 | Allow integers and floating point numbers to compare equal. (Closed-Leaf check-in: 0fe2e465 user: drh tags: query-invariants) | |
| 16:09 | Improve query-invariants to the point that it would have discovered the [f23a429d4153518d] bug if that bug had not already been fixed. (check-in: 2a7251ed user: drh tags: query-invariants) | |
| 15:52 | Only run query-invariant checks after the original query has run to completion, so that we know it does not error-out on a subsequent row. (check-in: d1fc857b user: drh tags: query-invariants) | |
| 15:11 | Fix the virtual table detection mechanism to avoid false-positives that were blocking all failures. Then fix a few of the additional problems that are revealed by that fix. More fixes are needed. (check-in: 42b2e667 user: drh tags: query-invariants) | |
| 12:25 | Fix the new --query-invariants option on fuzzcheck so that it does not use an unprotected sqlite3_value object as an argument to sqlite3_value_int64(). (check-in: d9f82015 user: drh tags: trunk) | |
| 11:39 | Avoid omitting the rhs of FULL JOINs in cases where it is only correct to omit the rhs of a LEFT JOIN. Fix for the problem reported by forum post 5610c17c3d. (check-in: f23a429d user: dan tags: trunk) | |
| 00:24 | Tiny fiddle cleanups. (Closed-Leaf check-in: 01969ba5 user: stephan tags: fiddle-indexeddb) | |
|
2022-06-16
| ||
| 23:55 | fiddle: initial proof of concept of using Emscripten's IndexedDB virtual filesystem for a persistent db. There's still much work to do to integrate this into something useful for clients but the concept is now proven. (check-in: 65c152d3 user: stephan tags: fiddle-indexeddb) | |
| 20:29 | Minor fix to the query invariant testing logic of fuzzcheck. (check-in: 447e62a0 user: drh tags: trunk) | |
| 13:57 | Merge the latest trunk enhancements into the reuse-schema branch. (check-in: 7f405fe2 user: drh tags: reuse-schema) | |
| 13:44 | Merge the latest trunk enhancements into the wal2 branch. (check-in: c8ad8699 user: drh tags: wal2) | |
|
2022-06-15
| ||
| 20:18 | Improvements to query invariants in fuzzcheck. (check-in: 3a461f61 user: drh tags: trunk) | |
| 16:26 | Improvements to query invariant checking in fuzzcheck. (check-in: 56c60a35 user: drh tags: trunk) | |
| 14:57 | When running an incremental vacuum, detect growth in the size of the database file (which can only occur if the file is corrupt) and fail with SQLITE_CORRUPT. (check-in: cd7a4412 user: drh tags: trunk) | |
| 12:50 | Add the --query-invariants option to fuzzcheck. No changes to the SQLite core. (check-in: fef282f9 user: drh tags: trunk) | |
| 12:32 | Dbsqlfuzz discovered a case where a bytecode branch is in fact taken, so change the designator from VdbeCoverageNeverTaken() to VdbeCoverage(). Test case in TH3. (check-in: 988a2a75 user: drh tags: trunk) | |
| 10:46 | Add the --query-invariants flag to fuzzcheck. The query invariant checks are only run if that flag is enabled. (Closed-Leaf check-in: d13b4621 user: drh tags: query-invariant-tests) | |
| 10:37 | Improvements to query invariant testing. Almost working now. (check-in: e0398204 user: drh tags: query-invariant-tests) | |
|
2022-06-14
| ||
| 22:21 | Fix the OP_NullRow opcode so that it works even if it applied to an ephemeral cursor that has not yet been created. (check-in: 0e925654 user: drh tags: trunk) | |
| 21:34 | Add missing SQLITE_FCNTL_SIZE_HINT call to a path taken during transaction rollback. (check-in: 6c3266c1 user: dan tags: trunk) | |
| 19:12 | Attempt to enhance fuzzcheck to do some simple invariant testing on queries. This is an incremental check-in for a work-in-progress. (check-in: ce2d7801 user: drh tags: query-invariant-tests) | |
|
2022-06-13
| ||
| 12:42 | Do not remove the EP_CanBeNull flag from expressions during a LEFT JOIN strength reduction if the query also contains a RIGHT JOIN. Fix for the problem identified by [forum/forumpost/b40696f50145d21c|forum post b40696f50145d21c]. (check-in: b1be2259 user: drh tags: trunk) | |
|
2022-06-10
| ||
| 16:41 | The same restrictions on the use of WHERE clause terms to drive indexes in the presence of RIGHT JOINs also apply to the use of WHERE clause terms to manufacture automatic indexes. This fixes a problem identified by forum post 51e6959f61. (check-in: 342c501f user: drh tags: trunk) | |
| 15:43 | fiddle: minor style tweaks, including using swapped colors for the input/output fields to help (hopefully) reduce the "which field is which?" dissonance. (check-in: e25dad86 user: stephan tags: trunk) | |
| 11:28 | Do not allow constant propagation between WHERE-clause terms and ON-clause terms as this can confuse RIGHT JOIN. Fix for the problem reported by forum post 8e4c352937e82929. (check-in: cab9b4cc user: drh tags: trunk) | |
| 10:10 | Fix a faulty assert() statement identified by forum post 0b91a75039. (check-in: 1f132bb0 user: drh tags: trunk) | |
| 09:31 | fiddle: modernized the UI based on related code in fossil's /pikchrshow. Changed the color scheme to match sqlite.org. (check-in: c4523ffc user: stephan tags: trunk) | |
|
2022-06-09
| ||
| 20:26 | Do not allow the subtype of a value to cross a subquery boundary. This fixes the problem identified by forum post 3d9caa45cbe38c78. (check-in: bbaf1f2e user: drh tags: trunk) | |
| 17:17 | Prevent subtype values from slipping across a subquery boundry when the subquery is implemented as a co-routine. (Closed-Leaf check-in: 9e51a6c0 user: drh tags: subtype-subquery) | |
| 16:19 | The subtype of a value should not propagate across a subquery boundary. Proposed fix for the problem reported by forum post 3d9caa45cbe38c78. Additional works is needed as not all cases are covered. (check-in: 08af1fe2 user: drh tags: subtype-subquery) | |
|
2022-06-08
| ||
| 18:29 | Move an #ifdef in shell.c to avoid a harmless "unused function" compiler warning. (check-in: 5abb5ef5 user: drh tags: trunk) | |
| 18:20 | Avoid zeroing the value returned by sqlite3_changes() when a DML statement is automatically reprepared in sqlite3_step(). (check-in: 09c8f9f1 user: dan tags: trunk) | |
| 17:48 | Improved comments on the new query flattener restriction of the previous check-in. Also a NEVER() macro on an unreachable branch. (check-in: 8c9e2d63 user: drh tags: trunk) | |
| 15:38 | Fix the query flattener so that it refuses a flattening that might leave both an inner-join and outer-join ON-clause constraint (or equivalent) on the same term of the FROM clause. (check-in: f6c4fb48 user: drh tags: trunk) | |
| 15:30 | Add restriction (29) to the query flattener - do not allow flattening that would leave both EP_InnerON and EP_OuterON constraints on the same join term. (Leaf check-in: c585d6a4 user: drh tags: right-join-query-flattener) | |
| 13:13 | New test cases refute check-in [3f45007d544e5f78]. (check-in: 12d3c96c user: drh tags: right-join-query-flattener) | |
| 12:46 | Do not allow a partial index scan on the left table of a RIGHT JOIN, because since the index is partial, some rows will be omitted from the scan, and those rows will subsequently be picked up by the no-match logic in the right-join post-processing loop. forum post c4676c4956. (check-in: 615c0026 user: drh tags: trunk) | |
| 12:35 | Minor fix to test/index9.test so that it can be invoked with other index tests using a wildcard. (check-in: ccbd6e77 user: drh tags: trunk) | |
| 12:20 | Update the routine that determines whether or not a partial index can be used so that it is not specific to LEFT JOIN. (check-in: 5a107fd7 user: drh tags: trunk) | |
|
2022-06-07
| ||
| 13:09 | Fix the query flattener to deal with a RIGHT JOIN corner case described by forum post 323f86cc30. (check-in: 3f45007d user: drh tags: trunk) | |
| 11:02 | Test cases for forum post 323f86cc30 added to join8.test, though commented out so that they do not (yet) run. (check-in: 7f97cb67 user: drh tags: trunk) | |
| 10:14 | Preserve the database encoding on the CAST operator in the sqlite3ValueFromExpr() routine. Forum thread 800eecf5e6cdc3f4. Test case in TH3. (check-in: 3f6a4420 user: drh tags: trunk) | |
|
2022-06-06
| ||
| 23:22 | More precise determination of when a WHERE clause can be used to drive an index on an outer join. (check-in: 318543a9 user: drh tags: trunk) | |
| 22:33 | More precise determination of when a WHERE/ON clause term can be used as an indexed join constraint. Testcase macros added to ensure test coverage. (Closed-Leaf check-in: f419e98c user: drh tags: outer-join-tests) | |
| 21:08 | Extra test cases for outer joins. (check-in: c7505765 user: dan tags: outer-join-tests) | |
| 15:27 | Do not allow a WHERE clause constraint to be used to drive an index for the right operand of a RIGHT JOIN, since this can cause problem if the constraint implies a not-NULL value for one of the columns for the left operand of the same join. See forum post 206d99a16dd9212f. (check-in: 4a31b794 user: drh tags: trunk) | |
| 15:22 | In CLI, drop .dbinfo command when build options prevent it from working. (check-in: 4b50a551 user: larrybr tags: trunk) | |
| 14:00 | Sync w/trunk, pickup fts3corrupt6.test change. (Closed-Leaf check-in: 268072f6 user: larrybr tags: dbinfo-guard) | |
| 06:29 | fiddle: replaced the huge file selection widget with a smaller proxy button. (check-in: f7c17449 user: stephan tags: trunk) | |
| 06:17 | Fix a test case in fts3corrupt6.test. Was failing due to [f1dfbc4f]. (check-in: 04b2e477 user: dan tags: trunk) | |
| 04:09 | shell: in WASM mode, permit ATTACH because the filesystem is a virtual sandbox and ATTACH can be used to provide more import/export options. Minor doc updates in sqlite3-api.js. (check-in: f28de5b7 user: stephan tags: trunk) | |
|
2022-06-05
| ||
| 22:58 | In CLI, drop .dbinfo command when build options prevent it from working. (check-in: 2512d2f6 user: larrybr tags: dbinfo-guard) | |
| 22:40 | Create new branch named "dbinfo-guard" (check-in: 8d942ba7 user: larrybr tags: dbinfo-guard) | |
|
2022-06-03
| ||
| 14:08 | Fix to test case in fts3corrupt.test. No changes to source code. (check-in: d18818af user: dan tags: trunk) | |
| 13:52 | Fix a 1-byte overread in fts3 that could occur when processing corrupt records. (check-in: f1dfbc4f user: dan tags: trunk) | |
|
2022-06-02
| ||
| 16:26 | Fix a problem with flattening and window functions causing an "IS <column>" to be transformed to "IS TRUE" or "IS FALSE" when <column> is a view or sub-select expression that is the literal value "TRUE" or "FALSE". (check-in: 2a952c77 user: dan tags: trunk) | |
| 02:45 | Avoid a direct use of abort() in debug builds to prevent a compilation issue for Windows CE. (check-in: 8eb9a7dd user: mistachkin tags: trunk) | |
|
2022-06-01
| ||
| 20:08 | New test cases for row values and RIGHT JOIN. (check-in: 5f0582b9 user: drh tags: trunk) | |
| 16:05 | Ensure that subqueries associated with row-values are invoked before being used when processing a RIGHT JOIN. Fix for the problem described by forum post 087de2d9ec87305b. (check-in: 5a9465dc user: drh tags: trunk) | |
| 16:04 | fiddle: added another UI element to the list of those which are disabled during long-running activities. Added DB.close() binding to the Worker-based wasm binding. (check-in: 5933163e user: stephan tags: trunk) | |
| 15:53 | Ensure that all necessary row-value subqueries are invoked while running the right-join post processing. (Closed-Leaf check-in: e4e71b4c user: drh tags: row-value-right-join) | |
| 14:52 | Minor tweaks to, and consolidation of, the wasm-related build flags. (check-in: f0ad6b1b user: stephan tags: trunk) | |
| 14:32 | Fix a minor problem in the Tcl "incrblob" command. This does not affect the SQLite core. (check-in: e96feccc user: dan tags: trunk) | |
| 13:32 | Candidate fix for the problem with row-value WHERE clause constraints on a RIGHT JOIN. (check-in: fdd782a7 user: drh tags: row-value-right-join) | |
| 13:32 | Only include the code for sqlite_offset() if compiled with -DSQLITE_ENABLE_OFFSET_SQL_FUNC. (check-in: 382cd7e1 user: drh tags: trunk) | |
| 13:01 | Fix a typo in a comment in a test case. No changes to code. (check-in: 570efa45 user: drh tags: row-value-right-join) | |
| 11:58 | Additional enhancements to the test cases. (check-in: 55cfad74 user: drh tags: row-value-right-join) | |
| 11:46 | Fix test cases in the previous check-in to make them postgres-compatible. (check-in: f223ba72 user: drh tags: row-value-right-join) | |
| 11:43 | Test case for forum post 087de2d9ec showing a problem when a row-value constraint is used with RIGHT JOIN. (check-in: d1e3e0d2 user: drh tags: row-value-right-join) | |
| 11:20 | The Worker-specific variants of the most significant DB-class JS bindings are implemented, most notably various uses of DB.exec(). (check-in: 371d6f74 user: stephan tags: trunk) | |
| 11:05 | Move the sqlite_offset() function implementation to be an in-line function, thereby avoiding special case code and freeing up a bit in the FuncDef.flags field. (check-in: 1c9812c4 user: drh tags: trunk) | |