* fix another issue
* use shard and replica name from Replicated database
* fix
Co-authored-by: mergify[bot] <37929162+mergify[bot]@users.noreply.github.com>
In case of multiple writes File() engine will set doNotWritePrefix(),
and this will avoid serializations initilization, move this to do this
always.
Fixes: #31004
If the column name contains invalid UTF-8 sequences
and the output data types are all considered safe,
then the output will not be sanitized and the generated
JSON will be invalid.
A minimal reproduction case is :
`SELECT length('\x80') FORMAT JSONCompact`
where we auto-generate a non-UTF-8 column name with only
integer outputs, whereas :
`SELECT '\x80' FORMAT JSONCompact`
would be sanitized because the column type is String and
will trigger UTF-8 sanitization over the entire document.
* clarify that cannot show create table of system's tables
* clarify that cannot show create table of system's tables in the document
* bypass test style-check for required `database=currentDatabase()` because `show create table` do not have condition
* rename supportsShowCreateTable to isSystemStorage
* build fake create query for show create table of system's tables
* fix tests for show create table on system's table:
1. only build fake create query when cannot get real query
2. only test rocksdb when it's enabled.
* fix test for show create table system.tables
* fix test for ASTCreateQuery set table and database
* change to LOGICAL_ERROR when cannot get in_memory metadata
Co-authored-by: tavplubix <tavplubix@gmail.com>
* implement common getCreateQueryFromStorage
Co-authored-by: tavplubix <tavplubix@gmail.com>
1. Dropped support for DatabaseOrdinary for MaterializeMySQL. It
is marked as experimental, and dropping support makes the code
more maintaible, and speeds up integration tests by 50%.
2. Get rid of thread name logic for StorageMaterializeMySQL wrapping,
use setInternalQuery instead (similar to MaterializedPostgreSQL).