unique conflicts-> then decrementing cause out of bounds error

[stanlay01]

Hello,

First. I do not usually code in Python, so I had problems running it in version 3 of Python.

Second, there is a bug causing PK out of bounds.
If you have two databases: first_database, second_database
By incrementing first_database, you get primary keys like 1000001 and second_database primary keys like 2000001.

Then there is a step : " 7/9 Mapping pk in case of uniques conflict", which finds conflicts. In case of conflict second_database primary key change to 1000001.

In step 9: Decrementing pks: it tries to decrement - 2000000 primary keys in second database. But in those, where was conflict it goes out of bounds because it has changed.

my solution was to change function rollback_pks

                   where = ""
                   if len(table_map['pk_changed_to_resolve_unique_conficts']):
                       where = "WHERE %(pk_col)s NOT IN (%(ids)s)" % {
                           'pk_col': list(table_map['primary'].keys())[0],
                           'ids': ",".join(table_map['pk_changed_to_resolve_unique_conficts'])
                       }
                   
                   
                   self._logger.qs = "UPDATE `%(table)s` SET `%(pk)s` = `%(pk)s` - %(step)d %(where)s" % {"table": table_name,
                                                                                                "pk": col_name,
                                                                                                'step': self._increment_value,
                                                                                                'where': where}
                   where = ""
                   if len(table_map['pk_changed_to_resolve_unique_conficts']):
                       where = "WHERE %(pk_col)s IN (%(ids)s)" % {
                           'pk_col': list(table_map['primary'].keys())[0],
                           'ids': ",".join(table_map['pk_changed_to_resolve_unique_conficts'])
                       }
                   
                   
                   self._logger.qs = "UPDATE `%(table)s` SET `%(pk)s` = `%(pk)s` - %(step)d %(where)s" % {"table": table_name,
                                                                                                "pk": col_name,
                                                                                                'step': self._increment_step,
                                                                                                'where': where}

[adamziel]

Thank you so much for sharing that @stanlay01! Would you like to contribute to the codebase? If you turned this change into a Pull Request, I'd be more than happy to review and merge it.