While in text protocol the number of rows is resetted in mysql_store/use_result
in binary protocol we need to explicitly reset it when switching to next result set.
- If prepare failed, we need to ignore errormessage from mysql_stmt_execute()
(which is always an error packet, mentioning invalid sttmt_id)
- Added additional check in mysql_stmt_reset for stmt_id == -1.
1) In case a cursor is used in a stored procedure server sends a SERVER_STATUS_CURSOR_EXISTS flag,
even if the client didn't open a cursor via mysql_stmt_attr_set. This ends in a "comands out of sync" error,
since the client is sending a COM_STMT_FETCH command while the result sets were not read before.
We check now in case server sends a SERVER_STATUS_CURSOR_EXIST flag, if the client opened a cursor before.
2) The stmt error codes weren't update for all COM_STMT* commands correctly, e.g. COM_STMT_FETCH didn't show an error even if it failed.
Client application can now register callback functions for either sending or retrieving data:
- typedef void (*ps_result_callback)(MYSQL_STMT *stmt, unsigned int column, unsigned char **row);
- typedef my_bool *(*ps_param_callback)(MYSQL_STMT *stmt, MYSQL_BIND *bind, unsigned int row_nr);
These functions will be registerd via mysql_stmt_attr_set call by specifying options STMT_ATTR_CB_PARAM
or STMT_ATTR_CB_RESULT.
- added travis support
- fixed appveyor settings
- fixed some warnings (gcc 4.8)
- removed sleep commands
- disabled failing tests when running against MySQL server,
mostly related to stored procedures and binary protocol
- reverted fix for MDEV_10361
Still open: TLS/SSL appveyor tests, since .msi installation on appveyor doesn't provide certificates.
- The metadata length value for a column with a zerofill flag was calculated with a fixed length instead of using the reported length.
- fixed a possible stackoverflow, if the display width for a column with zerofill flag is greater then 22. (libmysql ps implementation truncates the result after 21 digits and reports truncation error).
The revision also adds some (mainly VS specific) file/dirs definitions to .gitignore to make 'gid status' usable on Windows, and the typo in bulk1 testsuite
- If no bind variables were bound or the function mysql_stmt_store_result was not called before, the internal bind variables (stmt->bind) was not filled (lengths and null values)
After test fixes.
bulk_null() is disabled. Before the client library took care about MYSQL_TYPE_NULL, now it do not do it so server return an error. and I have no idea what should be chenged test or client library.
it needs to read
* if we're reading resultset row packets - till the EOF (254) packet.
* otherwise - till the *second* EOF packet (the first EOF will separate
resultset metadata packets and resultset row packets)
* or till the OK (0) packet, but not if we're reading resultset row
packets (because they also start from 0).
Also it needs to update server_status for caller to be able to check the
SERVER_MORE_RESULTS_EXIST flag, and the stmt->state to know that it's
not reading the resultset anymore.
Callers were fixed to invoke mthd_stmt_flush_unbuffered() repeatedly,
when they need to flush all result sets, not just one.
This commit fixes unittest/libmariadb/ps.c unit test
(it was failing to clean up mysql.proc table).
If a statement with open (read only) cursor is executed there is no buffered
result set (result set rows will be fetched directly from server), so we need to skip reading unbuffered result sets if a cursor is open.
To prevent unexpected behavior when reusing a statement with mariadb_stmt_execute_direct
a call to mysql_stmt_attr_set with option STMT_ATTR_PREBIND_PARAMS will reset the statement before.
