> El dijous 25 d’octubre de 2012 18:32:59 UTC+2, Bertrand Chenal va escriure: > Le Thu, 25 Oct 2012 18:14:05 +0200, > Mathias Behrle <[email protected] <javascript:>> a écrit : > > > * Betr.: " Re: [tryton-dev] Statement Import" (Thu, 25 Oct 2012 > > 17:37:12 +0200): > > > > I know what I'm talking about. I worked 2 years in a bank. > > > First of all, in belgium asking for a statement you already > > > received is charged by the bank. And a bank should not send you > > > twice the same statement by default, this is very basic logici if > > > they don't want customer complaints. > > > Second, detect duplicates could not be generic, again it will > > > depend on each format. I'm pretty that each standard format that > > > exists will have a unique reference for the statement and that this > > > statement could not contains duplicates lines. > > > > PLease re-read. I talked about exports from a banking software. And > > sorry, no, there is no unique reference despite all assumptions one > > would make (specifically talking about format MT940, which is a very > > current exchange format in Germany). > > Are you sure? from martin.hinner.info/bankconvert/swift_mt940_942.pdf, > I see a field "Statement Number/Sequence Number" which identify the > statement inside the year. > > As Albert said earlier, there is a lot of different format. I fail to > see how to find common concepts except the statement itself.
It may not be possible to have a common identifier but maybe we could agree on a way to manage/store those duplicates. -- [email protected] mailing list
