A parser for S.W.I.F.T. MT940 Files (for banking transactions)

andrew cooke andrew at acooke.org
Mon Mar 9 17:23:04 CET 2009

one other thing - LEPL *only* works with Python 2.6 and 3.0.  it WILL NOT
work with Python 2.5.  if this is an issue then it is not a suitable
solution (it is not possible to back-port the library).


andrew cooke wrote:
> a month is more than enough - i would expect to have something in a week.
> if possible, a data size of 1GB or more would be useful, but otherwise the
> largest file you have.
> i will use http://srd.tcg-net.com/Documentation/MT940/MT9402GB.htm as an
> initial guide.
> please remember that this is a "best efforts" attempt only.  i have no
> previous experience with this file format, or with banking data, and the
> result will come with no warranty.  you may want to continue to
> investigate other solutions.  i have no idea what your final application
> is, but you should at the very least test any parser (and related
> libraries) thoroughly before using it in a critical application.
> cheers,
> andrew
> ps i'm not sure this conversation should continue on the python list!
> Mustafa Sakalsiz wrote:
>> I can wait for a month. I can also provide test data, but I am not sure
>> if
>> their sizes are big enough for you. If you mention your expectation for
>> test
>> data sizes, I will try to find them.
>> Saki
>> On Mon, Mar 9, 2009 at 5:06 PM, andrew cooke <andrew at acooke.org> wrote:
>>> Saki wrote:
>>> > Hello,
>>> >
>>> > I need an MT940 file parses. It can be either a pure python library
>>> or
>>> > a binding/wrapper, no matter. Almost all european banks provide
>>> > transactions extract in MT940 format. There are parsers around, but
>>> > none of them are for python. Any help is greatly appreciated.
>>> How urgent is this?  I need some "real life" applications to test my
>>> (newish) parser library (http://www.acooke.org/lepl), but (1) it will
>>> take
>>> me some time to write a parser (this is work done in my free time) and
>>> (2)
>>> the motivation is to test large data sets.  I think the second point
>>> may
>>> be an issue as you may not be able to make public test data for
>>> confidentiality/security reasons?
>>> Andrew
> --
> http://mail.python.org/mailman/listinfo/python-list

More information about the Python-list mailing list