Incorrect Results for Int data

[detroitcoder]

A select query of integer data returns incorrect results when I am running pyodbc 3.0.7 with python 2.7 on RHEL6 against a Sybase database. I do not experience this issue when running on 3.0.3

Here is a basic example of incorrect results for int data

import pyodbc
dbconn = pyodbc.connect("DSN=....")
c = dbconn.cursor()
c.execute("select 0, 1, 2")
print c.fetchone()

>> (210453397504, 210453397505, 210453397506)

All of the integers consistently add 2^32 * 2^7 (210453397504) to the correct value. This remains true until the value gets to 2^31 (2147483648), at which point results are correctly converted to type Long.

When I explicitly cast the data to numeric, it returns correctly:

c.execute("select cast(0 as numeric), cast(1 as numeric), cast(2 as numeric)")
print c.fetchone()

>> (Decimal('0'), Decimal('1'), Decimal('2'))

Likewise if I explicitly add a decimal to the values, it will return the correct values as Decimal instances

c.execute("select 0.0, 1.0, 2.0")

>> (Decimal('0'), Decimal('1'), Decimal('2'))

However when I try to cast the data as Ints, I get the same incorrect results as I do by default:

c.execute("select cast(0 as int), cast(1 as int), cast(2 as int)")
print c.fetchone()

>> (210453397504, 210453397505, 210453397506)

There are a couple other users that have experienced the same issue on code.google.com
https://code.google.com/p/pyodbc/issues/detail?id=360

[mkleehammer]

Finally confirmed - I had to rebuild a new test VM.

[mkleehammer]

git bisect points finger at 916c69e