Technical Mailinglisten Archive


Re: Multistage imports best practices and examples.

- 24.09.2014 10:52:16
Though I was asking for multi-stage import wizards done *inside* 
OpenERP, this may help with my actual case: A one-time-only importing of 
QuickBooks exported account entries into our accounting system in OpenERP.

We have experience with Pentaho Spoon for doing ETL from our extended 
spreadsheets to a legacy system our clients still uses. Do you know if 
Spoon is equally a good fit?  Kettle is currenty a 600+ MB download that 
with my connection would be prohibitive (200kbps ~ 20KB/s).

Thanks and best regards,

Side note:

We can't do the usual "Opening ..." cause we need to have our entries.  
The main goal is to validate the reports before going into production 
with accounting.  I say this just avoid the answers that point that "you 
could just do an opening of accounts and the go...".  After validation 
we might just do the opening with the current state of accounts.

Le 24/09/14 04:39, Daniel Reis a écrit :
> > Is there's any best practice example about this kind of import that
> > takes several "wizards" passing data between them.  Data that would
> > ultimately be deleted and only remain the effect the last wizard window.
> Looks like the perfect job for an ETL tool:
> Your data is read by an "input" step, from whatever source you need, and
> flows through a series of "transformation" steps, until it's in the
> proper form to "load" into the destination.
> You might want to try Pentaho Kettle: it's quite user friendly and
> includes OpenERP input and output steps out of the box: see
> For version 8.0 there is a currently a regression bug with the load()
> method, but you can find a fix here:
> Only disadvantage is that it isn't able to handle ExternalIDs, so I
> wrote a "polyfill" to overcome that. I should write a blog post
> explaining with a friendly explanation on it, but you can find the
> module here:
> Regards
> Daniel Reis
> _______________________________________________
> Mailing-List:
> Post to:
> Unsubscribe: