Tested with python 2.7 and python 3
Merged

Thanks a lot

Sorry, but where are the commits? ;)

Ho i forgot again to push

I just come across a problem when I try to replicate from mysql binlog to postgresql. Postgresql seperate smallint from boolean, so this conversion broke the replication, since the column type and column value changed.

The problem here is, the mapping changed the original value, and I don't think it's correct. The bool() process could be easily done in the application layer, while the replication should expose the original value stored in database.

Any thoughts on this?

How do you know that the value was a TINYINT(1)?

TINYINT(1):

a very small integer. The signed range is -128 to 127. The unsigned range is 0 to 255.

BOOL:

A value of zero is considered false. Nonzero values are considered true

So it's actually a number, sometimes we store a number in TINYINT(1), and convert it to BOOL will lost this.

Check this example:

mysql> create table b (i tinyint(1));

mysql> insert into b values (42);

mysql> select * from b;
+------+
| i    |
+------+
|   42 |
+------+

mysql> create table c (i boolean);

mysql> insert into c values (20);

mysql> select * from c;
+------+
| i    |
+------+
|   20 |
+------+

There's no BOOLEAN in mysql, it's only number.

It's always easy to translate a number to bool in apps. In this data layer, let's expose the original data.