|
From: SourceForge.net <no...@so...> - 2007-02-07 14:54:18
|
Patches item #1653898, was opened at 2007-02-07 09:57 Message generated for change (Comment added) made by mfi You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=540674&aid=1653898&group_id=74338 Please note that this message will contain a full copy of the comment thread, including the initial issue submission, for this request, not just the latest update. Category: None Group: None Status: Open Resolution: None Priority: 5 Private: No Submitted By: Maxim F. Ischenko (mfi) Assigned to: Oleg Broytmann (phd) Summary: proper init mysql conn encoding Initial Comment: Current implementation of the mysqlconnection.py accepts 'sqlobject_encoding' and 'charset' parameters but doesn't quite properly use them. In particular, MySQLdb does query.encode(charset) call in execute() where charset is taken from conn.get_character_name(). The problem is - SQLObject never initializes it properly. This one-line patch fixes it by adding a call to conn.set_character_set() in makeConnection() helper. Index: sqlobject/mysql/mysqlconnection.py =================================================================== --- sqlobject/mysql/mysqlconnection.py (revision 2269) +++ sqlobject/mysql/mysqlconnection.py (working copy) @@ -62,6 +62,8 @@ if hasattr(conn, 'autocommit'): conn.autocommit(bool(self.autoCommit)) + conn.set_character_set(self.encoding) + return conn def _setAutoCommit(self, conn, auto): ---------------------------------------------------------------------- >Comment By: Maxim F. Ischenko (mfi) Date: 2007-02-07 16:54 Message: Logged In: YES user_id=113894 Originator: YES Note that I also had to tweak UnicodeStringLikeConvertor so that unicode parameters are encoded into utf8. Without this tweak I got errors like this one when unicode parameters are passed into queries built with sqlbuilder: File "/home/max/oss-projects/SQLObject-0.7/sqlobject/sqlbuilder.py", line 237, in __sqlrepr__ return "(%s %s %s)" % (s1, self.op, s2) UnicodeDecodeError: 'ascii' codec can't decode byte 0xc3 in position 19: ordinal not in range(128) The converter trick is controversial but I haven't found a better solution. ;( ---------------------------------------------------------------------- Comment By: Maxim F. Ischenko (mfi) Date: 2007-02-07 16:46 Message: Logged In: YES user_id=113894 Originator: YES It was tested against MySQL 4.1 (5.x should work while 4.0 shouldn't) and MySQLdb driver versions 1.2.0 and 1.2.1. I only tested it with utf8 encoding, the other ones like latin1 should work though I don't know why would you want to use them. ---------------------------------------------------------------------- Comment By: Maxim F. Ischenko (mfi) Date: 2007-02-07 16:44 Message: Logged In: YES user_id=113894 Originator: YES Please ignore that one-line fix; it won't work with MySQLdb 1.2.0 and there are other issues as I have discovered during testing. I attached new patch which is quite more involved. ;) ---------------------------------------------------------------------- Comment By: Maxim F. Ischenko (mfi) Date: 2007-02-07 16:42 Message: Logged In: YES user_id=113894 Originator: YES File Added: mysql_charset.patch ---------------------------------------------------------------------- Comment By: Oleg Broytmann (phd) Date: 2007-02-07 10:34 Message: Logged In: YES user_id=4799 Originator: NO Thank you. Does all MySQLdb versions have the call or have I test "if hasattr(conn, 'set_character_set')"? ---------------------------------------------------------------------- You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=540674&aid=1653898&group_id=74338 |
