If I tell it to delete everything, and everything it references, I expect it to do it, not complain about the fact that that could happen.
One way or another, all of those records are getting deleted, and I hate it when I have to run circles around a "this is for your own protection". I don't want to write 20 more lines of code to get around the fact one function that is supposed to delete everything, won't because it thinks I am not smart enough to realize that everything will be deleted. If a table has cascading deletes, it should well... cascade. If I don't want a .delete() to cascade, then I wouldn't declare a table as cascading. I want it to do what I mean. -Thadeus On Wed, Dec 9, 2009 at 3:03 PM, mdipierro <[email protected]> wrote: > and that is? > > On Dec 9, 2:09 pm, Thadeus Burgess <[email protected]> wrote: >> we're all consenting programmers here are we not? >> >> I expect it to do what I mean it to do :) >> >> -Thadeus >> >> On Wed, Dec 9, 2009 at 4:50 AM, Maciek Sykulski <[email protected]> wrote: >> >> > On Dec 8, 10:09 pm, mdipierro <[email protected]> wrote: >> >> You are right. I think this line was introduced to fix a problem and >> >> created another. The "SET FOREIGN_KEY_CHECKS=0" should be executed >> >> only when dropping a table in MySQL. Do you agree? >> >> > Hi Massimo, >> >> > I'm not sure. With SET FOREIGN_KEY_CHECKS=1 MySQL won't allow DROP >> > TABLE to be executed when there are other tables referencing it, which >> > might be a very good thing because database remains consistent. >> > In an ideal world user should alter referencing tables and remove all >> > referencing columns before delete/truncate is performed. Oracle has >> > CASCADE CONSTRAINTS which causes all relevant constraints to be >> > dropped. MySQL drop table CASCADE is not doing anything in later >> > versions. >> > On the other hand, TRUNCATE TABLE might be an opposite problem - it >> > does delete all rows in a table and also all others from other tables >> > referencing it. This means that a whole database may be purged by >> > db.table.truncate() not well thought through (ON DELETE CASCADE is the >> > default setting for all foreign key references in web2py, isn't it?). >> >> > The FOREIGN_KEY_CHECKS should be 1 for most of operations. >> > When comes to delete, truncate - I'm not sure how to weight on that. >> > It depends on general policy of web2py - how user-friendly, dummy- >> > proof, do-what-i-want-no-matter-what it is. >> >> > Maciek >> >> >> On Dec 8, 2:47 pm, Maciek Sykulski <[email protected]> wrote: >> >> >> > Hi, >> >> >> > We noticed that in our MySQL database ON DELETE CASCADE is not working >> >> > when a row is deleted by web2py controller. >> >> > It is working ok when I run SQL delete from mysql console. >> >> > Because of that, it is possible to get database into inconsistent >> >> > state with web2py >> >> >> > When looking into this problem I noticed self._execute('SET >> >> > FOREIGN_KEY_CHECKS=0;') in web2py source: >> >> >> > web2py$ grep -n -A 2 -B 10 -e "FOREIGN_KEY_CHECKS" -R * >> >> > gluon/sql.py-870- self._pool_connection(lambda : >> >> > MySQLdb.Connection( >> >> > gluon/sql.py-871- db=db, >> >> > gluon/sql.py-872- user=user, >> >> > gluon/sql.py-873- passwd=passwd, >> >> > gluon/sql.py-874- host=host, >> >> > gluon/sql.py-875- port=int(port), >> >> > gluon/sql.py-876- charset=charset, >> >> > gluon/sql.py-877- )) >> >> > gluon/sql.py-878- self._cursor = self._connection.cursor() >> >> > gluon/sql.py-879- self._execute = lambda *a, **b: >> >> > self._cursor.execute(*a, **b) >> >> > gluon/sql.py:880: self._execute('SET >> >> > FOREIGN_KEY_CHECKS=0;') >> >> > gluon/sql.py-881- self._execute("SET >> >> > sql_mode='NO_BACKSLASH_ESCAPES';") >> >> > gluon/sql.py-882- elif not is_jdbc and self._uri[:11] == >> >> > 'postgres://': >> >> >> > What is the rationale for self._execute('SET FOREIGN_KEY_CHECKS=0;') >> >> > there? >> >> > I'd like to have my database consistent and ON DELETE CASCADE working >> >> > - >> >> > can I/ should I change it to =1 ? >> >> >> > I've found a discussion about FOREIGN_KEY_CHECKS in web2py >> >> > here ...but not sure if it's >> >> > related.http://markmail.org/message/6472owgwupttlblq >> >> >> > Regards, >> >> > Maciek >> >> > -- >> >> > You received this message because you are subscribed to the Google Groups >> > "web2py-users" group. >> > To post to this group, send email to [email protected]. >> > To unsubscribe from this group, send email to >> > [email protected]. >> > For more options, visit this group >> > athttp://groups.google.com/group/web2py?hl=en. >> >> > > -- > > You received this message because you are subscribed to the Google Groups > "web2py-users" group. > To post to this group, send email to [email protected]. > To unsubscribe from this group, send email to > [email protected]. > For more options, visit this group at > http://groups.google.com/group/web2py?hl=en. > > > -- You received this message because you are subscribed to the Google Groups "web2py-users" group. To post to this group, send email to [email protected]. To unsubscribe from this group, send email to [email protected]. For more options, visit this group at http://groups.google.com/group/web2py?hl=en.
