MariaDB Connector/ODBC 3.0.1 Beta Changelog

You are viewing an old version of this article. View the current version here.

The most recent Stable (GA) release of MariaDB Connector/ODBC is:
MariaDB Connector/ODBC 3.1.20

Note: This page describes features in the source repository for MariaDB Connector/ODBC. There are currently no official packages or binaries available for download which contain the version of the driver described here. If you want to try out any of the new features described here you will need to get and compile the code yourself.

For the highlights of this release, see the release notes.

The revision number links will take you to the revision's page on GitHub. On GitHub you can view more details of the revision and view diffs of the code modified in that revision.

  • Revision #e5dd744 2017-07-31 02:30:34 +0200
    • [ODBC-105] Connector now constructs query for SQLColumns considering server version.
  • Revision #ee0a63d 2017-07-31 01:19:28 +0200
    • Fix of errom in CMakeLists.txt - addition to previous commit. FIx of possible memleak. Could occur if cursor was closed after "emulated", i.e. query not producing resultset and not containing parameters, query. Pretty extravagant, but possible. Expecially with unixodbc in some cases.
  • Revision #b269d37 2017-07-30 19:47:57 +0200
    • [ODBC-106] Support of MariaDB prepared statements bulk operations
  • Revision #0f68677 2017-05-19 02:27:26 +0200
  • Revision #6a1619b 2017-05-16 04:44:37 +0200
    • [ODBC-97] Fix and addition to old testcase to care about this issue. The problem was that parser would think of \ that it escapes closing quote, even if it was escaped in its turn(i.e. if it represented the character in the string. Also the case of NO_BACKSLASH_ESCAPES is now respected.
  • Revision #4b805ec 2017-05-15 09:36:34 +0200
    • [ODBC-95] If any of statements in a bacth could not be prepared, application would crash on the attempt to drom that stmt handle. Fix and testcase
  • Revision #ac3747b 2017-05-10 08:57:32 +0200
  • Revision #e7276e0 2017-05-10 05:19:33 +0200
    • Removed internal function for SQLFetch, using internal function for SQLFetchScroll with SQL_FETCH_NEXT instead. Added testcase for SQLFetch with array size > 1. I felt like there could be problems. And just won't hurt to have. Removed FetchType from Stmt handle structure as it wasn't used anywhere.
  • Revision #daaf929 2017-05-03 22:35:16 +0200
    • [ODBC-94] The driver would crash with ODBCv2 with exotic exotic sqlstate, e.g. OP000, as a result of bad GRANT syntax. The reason was that that sqlstate fields in MADB_ERROR struct are arrays, but connector waited for NULL sqlstate to stop traversing the MADB_ErrorList array. Testcase has been added. Patch also adds convenience function for tests to check if the server meets some minimum version criteria. We seemingly had problem SQL_DBMS_VER info type. The patch fixes as well.
  • Revision #45a1418 2017-04-14 14:53:21 +0200
    • Added to cmake config parameter to specify system name(SYSTEM_NAME), to be used in the package name.
  • Revision #66caaec 2017-05-07 05:44:58 +0200
  • Revision #fd9139d 2017-04-07 13:20:13 +0200
  • Revision #8dfaded 2017-04-07 12:35:49 +0200
    • Version bump -> 2.0.15
  • Revision #2c5a9a0 2017-04-05 15:40:38 +0200
    • Fix of the error in computation of returned ParameterValuePtr in SQLParamData on 64bit linux.
  • Revision #0d1fdac 2017-04-04 23:25:13 +0200
    • Cherry-picking from 3.0 - fixed (not) closing of MYSQL_STMT handles, and optimized some exessive close/init of them. Fix of possible memory leak in case of preparing of multistatement query an not executing it.
  • Revision #817ec5e 2017-04-03 21:14:05 +0200
    • To mitigate ODBC limitations causing ODBC-83, Fixed, or rather added aupport of hour_to_minute/second interval types. It allows to insert/fetch time values > 23:59:59(or < -23:59:59)
  • Revision #d94541b 2017-03-17 21:28:22 +0100
    • [ODBC-90] Fix and the testcase. We inserted default values for ignored columns. In case of the timestamp that is a function, and that does not work with prepared statements. The patch does not include ignored columns in the query. Also it fixes SQLCopyDesc - InternalBuffer pointers are not copied, since that can cause them to be freed twice.
  • Revision #939c888 2017-03-11 23:18:25 +0100
    • Fix and tests for ODBC-62. Export from MS Access did not work mainly because we returned wrong values in CREATE_PARAMS column of the SQLGetTypeInfo, and that caused MS Access to generate bad CREATE TABLE queries.
  • Revision #3a77b1a 2017-03-08 00:50:33 +0100
    • Removed last traces of 'def' catalog. Still were in some catalog funcions
  • Revision #b0e0d7c 2017-03-06 02:02:27 +0100
    • [ODBC-73]Fix and testcase. The connector relied on the field's BINARY_FLAG flag to establish the fact whether the field it binary or not. But that flag would be set if any binary collation is selected for any char field. And that is not right. The patch changing that to check only if field's charset is binary charset(63)
  • Revision #168ea7b 2017-03-03 20:34:50 +0100
    • Version bump -> 2.0.14 Added charset conversion processing for WCHAR in SQLGetData
  • Revision #3f3ec84 2017-03-03 00:37:35 +0100
    • [ODBC-72] Fix and testcase. Fixing the case, when while getting WCHAR data in parts, surrogate pair gets divided into different chunks, i.e first SQLWCHAR of the pair falls on the end of buffer.
  • Revision #8c998ae 2017-03-01 01:00:22 +0100
    • Fixes and testcases more around ODBC-70 AND ODBC-83 - connector now checks if date/time are valid ODBC date/time, and returns errors, if they are not. Fixed bug in SQLSetStmtAttr - SQL_ATTR_ROW_OPERATION_PTR and SQL_ATTR_ROW_STATUS_PTR were setting fields in wrong descriptors. Fixed many stuff around row status, return code for rowset fetch.
  • Revision #677f3c9 2017-02-25 21:35:38 +0100
    • Fix building ODBC connector on OSX/macOS (#13)
  • Revision #bf947c0 2017-02-25 18:05:45 +0100
    • [ODBC-71] Fix and testcase for the problems causing issue and around. These include: errors in SQLCopyDesc, that could caue a crash, support of missing (odbc2) info types(SQL_POS_OPERATIONS, SQL_STATIC_SENSITIVITY, SQL_LOCK_TYPES, SQL_SCROLL_CONCURRENCY), support of SQL_DESC_SEARCHABLE and SQL_DESC_SCHEMA_NAME thru SQLColAttribute.
  • Revision #6ccf03c 2017-02-23 12:03:57 +0100
    • Setting Connection->mariadb to NULL, after mysql_close'ing it. Reportedly that helps to avoid certain problems in some applications.
  • Revision #8d85958 2017-02-23 00:46:41 +0100
    • Added locks around all mysql_stmt_init, since they modify list inside MYSQL structure.
  • Revision #06b5e6a 2017-02-22 15:21:06 +0100
    • [ODBC-69] Fix and testcase - charset name is cas-insensitive now
  • Revision #4d2c342 2017-02-22 00:19:12 +0100
    • [ODBC-84] The fix ant the testcase. SQLGetTypeInfo returned SQL_ERROR for SQL_WCHAR and other wchar types.
  • Revision #f185440 2017-02-21 16:49:20 +0100
    • [ODBC-78] Fix and testcase. Besides described in the report case, the fix also take care of fixed length types - specs say also to return SQL_NO_DATA for subsequent calls of SQLGetData for such columns. Also had to introduce some changes, since the fix affected SQLSetPos, that calls same function to get values required to construct queries, and (fix affected) some tests.
  • Revision #4fe4e64 2017-02-02 07:36:02 +0100
    • [ODBC-77] Fix and the testcase. Connector did not recognize ANALYZE TABLE as query returning result. SImilar error fixed for EXPLAIN.
  • Revision #3de37aa 2017-01-31 00:19:25 +0100
    • [ODBC-74] Fix and testcase. The problev was in parsing errors while splitting the batch in single statements. The patch also fixes other parsing errors around comments and quots.

Comments

Comments loading...
Content reproduced on this site is the property of its respective owners, and this content is not reviewed in advance by MariaDB. The views, information and opinions expressed by this content do not necessarily represent those of MariaDB or any other party.