Azat Khuzhin
a595877d02
Fix handling INSERTs after async_insert handling from server for fuzzer
...
Rewrite API one more time, since in case of fuzzer that assertion could
not be satisfied, due to it fuzz the query, but leave the insert->data,
and if we will parse query from formatted we will loose insert->data
And also cleanup interface, remove report_error that is unused, and move
echo_query out.
CI: https://s3.amazonaws.com/clickhouse-test-reports/json.html?PR=77578&sha=296cda664bb7288f786ef98e2af9ea1e0e5094fd&name_0=PR&name_1=AST%20fuzzer%20%28amd_debug%29
2025-04-11 14:46:11 +02:00
Azat Khuzhin
296cda664b
Re-introduce query_to_execute for processParsedSingleQuery()
...
The problem is that without this distinction we have a query with
trailing comments (due to applied adjustQueryEnd())
2025-04-11 10:43:49 +02:00
Azat Khuzhin
c6f93e16bb
Pass query for execution as string_view
...
The data that ASTPtr may refer to (in case of INSERT) should point to
the same memory region, since in case of async_insert, we need to
calcualte the end of the query, and we cannot do this without this.
2025-04-10 17:52:28 +02:00
Azat Khuzhin
bafa45a7b1
Fix applying async_insert from server (via apply_settings_from_server)
...
The problem is that async_insert is checked during parsing, to know what
to send to the server, even before server settings available. And this
leads to "Unknown packet 11 from server" errors.
So, first approach will be to simply ignore async_insert for
apply_settings_from_server, but this setting is interesting to apply
from the server (since this may have performance impact).
Another is to detect what part of the query we should after server
settings had been received, this patch follows this approach.
2025-04-10 16:37:26 +02:00
Pedro Ferreira
58ae24324e
More const
2025-04-08 14:32:34 +00:00
Pedro Ferreira
5c6208ae1d
More settings and fix tidy compilation
2025-03-27 12:50:41 +00:00
Pedro Ferreira
cf82890c3e
Fix and better answer
2025-03-26 17:39:15 +00:00
Pedro Ferreira
d17e2d887d
Sometimes use other settings when reading the data when not validating content
2025-03-26 17:18:09 +00:00
Pedro Ferreira
570f73386f
Increase probability and keep lists sorted
2025-03-26 16:15:23 +00:00
Pedro Ferreira
98fa5a4757
Cleanup and rename setting
2025-03-26 15:47:41 +00:00
Pedro Ferreira
68a32ca839
Attempt to dump and read into another table to test formats with bad inputs
2025-03-26 15:08:00 +00:00
Raúl MarÃn
fa763588c1
Fix buzzhouse build
2025-03-20 19:34:30 +01:00
Nikita Mikhailov
f6694387bc
Fix build
2025-03-14 20:15:47 +01:00
Nikita Mikhailov
450c8ba2a8
Remove formatAST completely
2025-03-14 20:04:45 +01:00
Nikita Mikhailov
b600314abb
Rename method and add another one
2025-03-14 19:57:10 +01:00
Nikita Mikhailov
050135d5c9
Remove queryToString
2025-03-14 19:57:10 +01:00
Pedro Ferreira
a7a764f642
Merge remote-tracking branch 'origin/master' into ast-update
2025-03-13 11:59:57 +00:00
Pedro Ferreira
fee5e405c7
Use avaialble Parquet readers when possible
2025-03-10 15:37:46 +00:00
Pedro Ferreira
c789154767
Settings are now propagated from the server to the client
2025-03-10 13:29:54 +00:00
Pedro Ferreira and GitHub
3432fbccbb
Merge branch 'master' into ast-update
2025-03-10 08:56:33 +01:00
Pedro Ferreira
b719f75b65
Fix peer check
2025-03-05 16:28:53 +00:00
Pedro Ferreira
9953bad215
More features added in v25.2
2025-03-03 12:09:02 +00:00
Pedro Ferreira
bd47f63996
Removing duplicated code
2025-03-03 10:40:59 +00:00
Pablo Marcos
cb988ef115
Remove unnecessary include
...
If no destructor is added, an implicit destructor is automatically
generated in the header file for every translation unit including
the header. Since it's the destructor the one that needs to know
the size of what std::unique_ptr points to in order to know how
to destroy it, the best thing we can do is to define a default
destructor in the .cpp to prevent any automatic destructors
being generated via header include.
2025-02-27 08:55:06 +00:00
Pedro Ferreira
77764c1a14
More review comments
2025-02-26 16:29:46 +00:00
Pedro Ferreira
57b1e3f459
Addressing review comments
2025-02-26 16:05:59 +00:00
Pedro Ferreira
feec401376
Remove include
2025-02-25 13:10:13 +00:00
Pedro Ferreira
af808e629a
Remove this branch, is not working properly
2025-02-24 15:17:23 +00:00
Pedro Ferreira
7c28b6a4f6
Don't AST fuzz added INSERT INTO /dev/null for performance testing
2025-02-14 11:30:53 +00:00
Pedro Ferreira
1c58b9ef5f
This step has to be done later
2025-02-14 10:55:59 +00:00
Pedro Ferreira
eb198d8ee9
Dump results to /dev/null with INSERT INTO FUNCTION
2025-02-14 09:28:04 +00:00
Pedro Ferreira
5f43361cf7
Fixes
2025-02-14 07:49:33 +00:00
Pedro Ferreira
840bbac6ea
Dump fuzzed AST
2025-02-14 07:37:15 +00:00
Pedro Ferreira
7ff51b7d8b
Compare query success and revert change
2025-02-14 07:23:05 +00:00
Pedro Ferreira
1375820369
Fix possible AST fuzzer bug
2025-02-13 17:37:14 +00:00
Pedro Ferreira
b467b070f8
Fixes
2025-02-13 17:13:58 +00:00
Pedro Ferreira
67205cf230
Fix
2025-02-13 16:52:03 +00:00
Pedro Ferreira
78bc77895f
Make sure settings are printed
2025-02-13 16:15:45 +00:00
Pedro Ferreira
19b0202910
Fixes
2025-02-13 15:54:44 +00:00
Pedro Ferreira
42d327aeeb
Fixes
2025-02-13 15:38:54 +00:00
Pedro Ferreira
9de824684a
Fix
2025-02-13 15:20:06 +00:00
Pedro Ferreira
39c3f356f0
Don't keep new settings
2025-02-13 15:13:42 +00:00
Pedro Ferreira
37e8a56e13
Code cleanup
2025-02-13 14:57:27 +00:00
Pedro Ferreira
3af4840d88
Fix
2025-02-13 14:38:06 +00:00
Pedro Ferreira
e58173d016
Dump result into /dev/null
2025-02-13 13:00:33 +00:00
Pedro Ferreira
49eddeac43
Set settings per query
2025-02-12 12:54:23 +00:00
Pedro Ferreira
15dbebee25
Print set query
2025-02-12 12:11:02 +00:00
Pedro Ferreira
8d24227a28
Set random settings while measuring performance
2025-02-11 17:28:15 +00:00
Pedro Ferreira
89c286be63
Fixes
2025-02-10 15:26:22 +00:00
Pedro Ferreira
437e63b6e7
Attempting multiple fixes
2025-02-10 14:24:31 +00:00
Azat Khuzhin