Commit Graph

10455 Commits

Author SHA1 Message Date
Pekka Enberg
9cd869f660 Merge 'Fix various ALTER TABLE bugs' from Jussi Saurio
Closes #3392 , Closes #3395
1. Prevent `ALTER TABLE t DROP COLUMN a` when exists e.g. `CREATE INDEX
ta ON t(a)`
2. Prevent `ALTER TABLE t DROP COLUMN a` when exists e.g. `CREATE INDEX
tba ON t(b) WHERE t.a > 100`
3. Prevent `ALTER TABLE t DROP COLUMN a` when exists e.g. `CREATE VIEW
ta AS SELECT a FROM t`;
and
4. Prevent `ALTER TABLE t RENAME a TO b` when exists e.g. `CREATE VIEW
ta AS SELECT a FROM t`;
Number 4 is incompatible with SQLite because we should be rewriting the
view to include the renamed column. I left that as a FIXME

Closes #3394
2025-09-27 11:10:46 +03:00
Jussi Saurio
283fba2e0d use normalized table name 2025-09-27 09:53:11 +03:00
Jussi Saurio
b43a89e423 Add regression tests for ALTER TABLE stuff 2025-09-27 09:45:15 +03:00
Jussi Saurio
67d320960d ALTER TABLE: prevent dropping/renaming column referenced in VIEW 2025-09-27 09:45:15 +03:00
Jussi Saurio
3137357092 ALTER TABLE: prevent dropping indexed column in VDBE layer 2025-09-27 09:45:15 +03:00
Jussi Saurio
085b92dc4e ALTER TABLE: prevent dropping indexed columns in translate layer 2025-09-27 09:45:15 +03:00
Jussi Saurio
a2d833c073 ALTER TABLE: add comment about things preventing drop column 2025-09-27 09:45:15 +03:00
Pekka Enberg
d796964a1e Merge 'support mixed integer and float expressions in the expr_compiler' from Glauber Costa
Fixes #3373

Closes #3387
2025-09-27 08:22:14 +03:00
Pekka Enberg
e9155610ed Merge 'github: Add 30 minute timeout to all jobs' from Pekka Enberg
We're getting hit by macOS runner concurrency limits whenever some jobs
get stuck (for example, because of a deadlock).

Closes #3377
2025-09-27 08:21:41 +03:00
Pekka Enberg
2f1ac8eb4a Merge 'bench/tpch: remove "cast('yyyy-mm-dd' as datetime)"' from Jussi Saurio
`CAST('yyyy-mm-dd' as datetime)` causes sqlite and tursodb to interpret
the value as just 'yyyy', e.g. '1995-01-01' becomes '1995', causing a
lot of the queries not to return any results, which is not what we want.

Reviewed-by: Preston Thorpe <preston@turso.tech>

Closes #3374
2025-09-27 08:21:30 +03:00
Pekka Enberg
65b382b9e1 Merge 'Make MVCC code Send and Sync' from Pekka Enberg
Closes #3361
2025-09-27 08:21:01 +03:00
Pekka Enberg
92291ed736 Merge 'Fix offset variable handling' from Nikita Sivukhin
Before, db generated incorrect plan in case when offset parameter were
introduced as variable:
```
turso> EXPLAIN SELECT * FROM users LIMIT ? OFFSET ?;
addr  opcode             p1    p2    p3    p4             p5  comment
----  -----------------  ----  ----  ----  -------------  --  -------
0     Init               0     16    0                    0   Start at 16
1     Variable           1     1     0                    0   r[1]=parameter(1); OFFSET expr
2     MustBeInt          1     0     0                    0
3     Variable           2     2     0                    0   r[2]=parameter(2); OFFSET expr
4     MustBeInt          2     0     0                    0
5     OffsetLimit        1     3     2                    0   if r[1]>0 then r[3]=r[1]+max(0,r[2]) else r[3]=(-1)
6     OpenRead           0     2     0                    0   table=users, root=2, iDb=0
7     Rewind             0     15    0                    0   Rewind table users
8       Variable         2     2     0                    0   r[2]=parameter(2); OFFSET expr
9       MustBeInt        2     0     0                    0
10      IfPos            2     14    1                    0   r[2]>0 -> r[2]-=1, goto 14
11      Column           0     0     4                    0   r[4]=users.x
12      ResultRow        4     1     0                    0   output=r[4]
13      DecrJumpZero     1     15    0                    0   if (--r[1]==0) goto 15
14    Next               0     8     0                    0
15    Halt               0     0     0                    0
16    Transaction        0     1     1                    0   iDb=0 tx_mode=Read
17    Goto               0     1     0                    0
```

Reviewed-by: Jussi Saurio <jussi.saurio@gmail.com>

Closes #3360
2025-09-27 08:20:54 +03:00
Pekka Enberg
a932ac5450 Merge 'core: recover logical log on Database::connect ' from Pere Diaz Bou
If we open database and logical log is not empty we need to recover from
it. We also make sure a single recover executes concurrently and other
connections just wait for it to finish.
I also changed the fuzz tester to use `restart` instead of calling
`load_logical_log` manually to test this behaviour.

Reviewed-by: Jussi Saurio <jussi.saurio@gmail.com>

Closes #3359
2025-09-27 08:20:41 +03:00
Danawan Bimantoro
a58c6a3c0b Merge branch 'main' into main 2025-09-27 11:05:44 +07:00
Glauber Costa
d28022b491 support mixed integer and float expressions in the expr_compiler
Fixes #3373
2025-09-26 21:11:38 -03:00
Preston Thorpe
5dc30b5a68 Merge 'core/storage: Display page category for rowid integrity check failure' from Pekka Enberg
Let's add more hints to hunt down the reason for #2896.

Reviewed-by: Preston Thorpe <preston@turso.tech>

Closes #3378
2025-09-26 19:07:45 -04:00
Preston Thorpe
ddbedecace Merge 'fix encryption config in the sync-client' from Nikita Sivukhin
Reviewed-by: Jussi Saurio <jussi.saurio@gmail.com>

Closes #3379
2025-09-26 19:01:27 -04:00
PThorpe92
5d8a735aaf fix clippy error 2025-09-26 18:06:09 -04:00
PThorpe92
abab04dac9 Add regression test for col named rowid 2025-09-26 17:33:53 -04:00
PThorpe92
d4dc458328 Evaluate table column refs before checking rowid to allow using it as col name 2025-09-26 17:33:38 -04:00
PThorpe92
e52aa1538e Remove unused BTreeTable method for checking single field on Column in schema 2025-09-26 17:32:51 -04:00
PThorpe92
af215c2906 Check cols first before falling back to explicit rowid in UPDATE translation 2025-09-26 17:32:16 -04:00
PThorpe92
9bd852297a Allow in parser using rowid explicitly for a col when creating table 2025-09-26 17:31:02 -04:00
Preston Thorpe
e43184fb98 Merge 'translate: refactor arguments and centralize parameter context' from Preston Thorpe
This PR contains _no semantic changes_.
I made this cleanup on another branch when I was working on subqueries,
as the inconsistency with passing around `Schema` and `SymbolTable` had
been kinda bothering me. By adding `param_ctx` to the program, it
prevents accidentally resetting it to zero.

Reviewed-by: Jussi Saurio <jussi.saurio@gmail.com>

Closes #3382
2025-09-26 13:05:29 -04:00
PThorpe92
d9658070a9 Fix clippy warnings 2025-09-26 12:17:34 -04:00
PThorpe92
5fcc187434 translate: refactor arguments and centralize parameter context 2025-09-26 12:06:44 -04:00
Pekka Enberg
222ab125c1 Turso 0.2.0-pre.9 2025-09-26 19:00:14 +03:00
Nikita Sivukhin
fe4bfb7c88 fix encryption config in the sync-client 2025-09-26 19:29:37 +04:00
Pekka Enberg
931cf2658e core/storage: Display page category for rowid integrity check failure
Let's add more hints to hunt down the reason for #2896.
2025-09-26 18:25:49 +03:00
Pere Diaz Bou
99adf73168 core/mvcc/logical-log: rename to needs_recovery 2025-09-26 16:59:57 +02:00
Pekka Enberg
f7bf60e856 github: Add 30 minute timeout to all jobs
We're getting hit by macOS runner concurrency limits whenever some jobs
get stuck (for example, because of a deadlock).
2025-09-26 17:45:42 +03:00
Nikita Sivukhin
f80650586a remove misleading comment 2025-09-26 18:06:20 +04:00
Nikita Sivukhin
5bf69350b3 add simple tests for offset/limit binding 2025-09-26 18:06:20 +04:00
Nikita Sivukhin
63a9fa8c28 fix handling of offset parameter set through variable
- before the fix db generated following plan:

turso> EXPLAIN SELECT * FROM users LIMIT ? OFFSET ?;
addr  opcode             p1    p2    p3    p4             p5  comment
----  -----------------  ----  ----  ----  -------------  --  -------
0     Init               0     16    0                    0   Start at 16
1     Variable           1     1     0                    0   r[1]=parameter(1); OFFSET expr
2     MustBeInt          1     0     0                    0
3     Variable           2     2     0                    0   r[2]=parameter(2); OFFSET expr
4     MustBeInt          2     0     0                    0
5     OffsetLimit        1     3     2                    0   if r[1]>0 then r[3]=r[1]+max(0,r[2]) else r[3]=(-1)
6     OpenRead           0     2     0                    0   table=users, root=2, iDb=0
7     Rewind             0     15    0                    0   Rewind table users
8       Variable         2     2     0                    0   r[2]=parameter(2); OFFSET expr
9       MustBeInt        2     0     0                    0
10      IfPos            2     14    1                    0   r[2]>0 -> r[2]-=1, goto 14
11      Column           0     0     4                    0   r[4]=users.x
12      ResultRow        4     1     0                    0   output=r[4]
13      DecrJumpZero     1     15    0                    0   if (--r[1]==0) goto 15
14    Next               0     8     0                    0
15    Halt               0     0     0                    0
16    Transaction        0     1     1                    0   iDb=0 tx_mode=Read
17    Goto               0     1     0                    0

- the problem here is that Variable value is re-read at step 8 - which is wrong
2025-09-26 18:05:36 +04:00
Preston Thorpe
1b2d7ea534 Merge 'fix avg aggregation' from Nikita Sivukhin
- ignore NULL rows as SQLite do
- emit NULL instead of NaN when no rows were aggregated
- adjust agg column alias name

Reviewed-by: Jussi Saurio <jussi.saurio@gmail.com>

Closes #3376
2025-09-26 09:59:50 -04:00
Nikita Sivukhin
a0c47b98b8 fix test 2025-09-26 17:42:11 +04:00
Nikita Sivukhin
5b5379d078 propagate context to stringifier to properly derive column names 2025-09-26 17:40:41 +04:00
Nikita Sivukhin
52f3216211 fix avg aggregation
- ignore NULL rows as SQLite do
- emit NULL instead of NaN when no rows were aggregated
- adjust agg column alias name
2025-09-26 17:11:06 +04:00
Jussi Saurio
045b11b255 bench/tpc-h: don't fail build if query 1 has output difference (known floating point precision issue) 2025-09-26 16:02:37 +03:00
Jussi Saurio
a783f82470 bench/tpch: remove "cast('yyyy-mm-dd' as datetime)"
this causes sqlite and tursodb to interpret the value as just 'yyyy',
e.g. '1995-01-01' becomes '1995', causing a lot of the queries not to
return any results, which is not what we want.
2025-09-26 15:55:38 +03:00
Pere Diaz Bou
9e47cc3700 clippy 2025-09-26 14:16:11 +02:00
Pekka Enberg
1402e9841e core/mvcc: Wrap StreamingLogicalLogReader::buffer with RwLock 2025-09-26 14:19:16 +03:00
Pekka Enberg
96accef06c core/mvcc: Wrap header with RwLock 2025-09-26 14:10:18 +03:00
Pere Diaz Bou
59d3e37b9f fmt 2025-09-26 13:01:13 +02:00
Pere Diaz Bou
9c1d94a355 core/mvcc/logical-log: assert we don't call begin_load_tx twice 2025-09-26 12:59:13 +02:00
Pere Diaz Bou
4cdf293a2b core/mvcc/logical-log: fuzz test recover use db.restart 2025-09-26 12:56:58 +02:00
Pere Diaz Bou
ae994146af core/mvcc/logical-log: on mvcc restart clear DATABASE_MANAGER 2025-09-26 12:56:43 +02:00
Pere Diaz Bou
2a7abd82f7 core/lib: recover mvcc logical log if needed on connect 2025-09-26 12:47:52 +02:00
Pere Diaz Bou
83d8a7c775 core/mvcc/logical-log: lock recover logical log process 2025-09-26 12:47:31 +02:00
Pavan-Nambi
fdabbed539 length shall not count when it sees nullc 2025-09-26 15:07:33 +05:30