argh! we don't want to NULL the stmt here; freeing the results is not the same
as destroying the prepared statement handler, so we're leaking and breaking, as
can be seen by running the test

argh! we don't want to NULL the stmt here; freeing the results is not the same
as destroying the prepared statement handler, so we're leaking and breaking, as
can be seen by running the test suite.

brown paper bag for PDO in PHP 5.1RC4?

show more ...

MFH: free && set to NULL S->stmt

MFH: Fixed bug #34623 (Crash in pdo_mysql on longtext fields).

- TSRM fix

Fixup LOB handling for inserts (refs #34630).

Also tripped over the return of PECL #5200; looks like mysql doesn't return an
accurate length for the columns. The PDO driver will sanity

Fixup LOB handling for inserts (refs #34630).

Also tripped over the return of PECL #5200; looks like mysql doesn't return an
accurate length for the columns. The PDO driver will sanity check the real
length against the buffer size it allocated (based on the info provided by
mysql), so that we won't overrun the buffer. In addition, if a varchar field
is reported as having a length of less than 128, we'll allocate 128 just in
case.

If the data is truncated, report it via the appropriate sqlstate code.

There must be a better way to do this stuff.

show more ...

MFH: Fixed bug #34001 (pdo_mysql truncates numberic fields at 4 chars)

Fixed memory corruption (wrong order of operations of stored prep. stmt).
Optimize the max length calculation process.

Proper handling for databases that need to pre-calculate length of large
columns, which is not normally done for performance reasons.

Only set attribute if result set contains some columns.

clean up

this hack can and does segv; let's remove it.

gah!

it is not an error for a statement to return 0 rows.

Andrey: please run the test suite before each commit!
This broke beta 3 and was clearly visible as a result of running the tests.

it is not an error for a statement to return 0 rows.

Andrey: please run the test suite before each commit!
This broke beta 3 and was clearly visible as a result of running the tests.

Running further tests now.

show more ...

strictly check the result of mysql_affected_rows()

stmt->column_count is set only if the result has been already bound, don't
rely on that and use the real count which is kept in st_mysql_stmt
#of course using internal values is bad idea but

stmt->column_count is set only if the result has been already bound, don't
rely on that and use the real count which is kept in st_mysql_stmt
#of course using internal values is bad idea but the comment already says
#that this should be fixed

show more ...

ok, calculate max_length only in case of a blob (normal, medium, long).
in case of a varchar lets allocate 255 bytes and be quick without asking
libmysql to update max_length

if the result set is buffered ask libmysql to compute the lengths, so
later allocate buffer as big as the largest value of the column in the
result set (max_legth positive).

Get precise lengths from the server, as suggested by Andrey.

Added safety checks.

Make cursor closer work with emulation as well.

Added missing header check.

Implement cursor_closer for MySQL driver.

Fixed memory leak

Fix two bugs:

- execute() would not re-fetch meta data after nextRowset() had been called.

- buffered mode would only be enabled on the first execute on a statement handle.

Make prepared statements obey buffering flag in MySQL.