Installation guide
mxODBC - Python ODBC Database Interface 
Python Type  SQL Type  Comments 
(signed long)  driver/manager, so expect the usual 
truncations. 
Long Integer  SQL.CHAR  
(char *) 
Conversion from the Python long integer is 
done via the string representation since there 
usually is no C type with enough precision to 
hold the value. 
Float  SQL.DOUBLE 
(double) 
Conversion from the Python float (a C double) 
to the SQL column type is left to the ODBC 
driver/manager, so expect the usual 
truncations. 
decimal.Decimal  SQL.VARCHAR, 
SQL.LONGVARCHAR 
(
char *) 
Conversion from a Python decimal.Decimal 
instance is done via the string representation to 
avoid losing precision. 
The long variant is used for decimal 
representations longer than 256 characters. 
DateTime  SQL.TIMESTAMP 
or 
SQL.DATE 
Converts the DateTime instance into a 
TIMESTAMP or DATE struct defined by the 
ODBC standard.  
The ODBC driver may use the time part of the 
instance or not depending on the SQL column 
type (DATE or TIMESTAMP). 
Seconds are  rounded according to the setting 
of the .timestampresolution setting, which 
defines the resolution of the timestamps in 
nanoseconds. 
mxODBC will round the timestamp's second 
value to the nearest nanosecond fraction 
defined by this setting in order to avoid issues 
with float second values which sometimes 
cannot be represented with full accuracy. If not 
set, the .timestampresolution attribute defaults 
to 1 nanosecond, so rounding usually is done 
to the nearest nanosecond, which is also the 
smallest fraction supported by the ODBC 
standard. 
DateTimeDelta  SQL.TIME  Converts the DateTimeDelta instance into a 
TIME struct defined by the ODBC standard. 
Fractions of a second will be lost in this 
conversion. 
datetime.date  SQL.DATE 
Converts the datetime.date instance into a 
DATE struct defined by the ODBC standard. 
Note that some database backends don't 
support date column types and give an error 
when using datetime.date objects. MS SQL 
Server 2000 and 2005 are examples. MS SQL 
148 










