an one to many relationship may also be established without an inverse foreign key, just set the foreign key to the 'one' side as primary key also. this would do for a parent- children relationship, less for a 'many' side table that may have multiple relationships with the same master, i.e. for lookup fields. the last case would require a foreign key for every such relationship.
alex On Thu, Oct 18, 2018, 19:03 Alex Bodnaru <alexb...@gmail.com> wrote: > > hello dave, friends, > i'm afraid i had a quite incomplete and optimistic understanding of the > foreign keys handling in xsd. > fortunately, there is a straightforward solution to implement primary and > foreign keys altogether. > see the following link for an example: > https://stackoverflow.com/a/15789233/4444742 > however, as my xsd inputs have no such keys, i'd also have to implement > them at the custom annotation level, as i did with the models renaming and > suppressing in the last version. > but, i will cleanup the interpretation of the minoccurs and maxoccurs > attributes to consistently represent 1:1, 1:m and m:m relationships. > > On Fri, Oct 12, 2018 at 5:39 PM Alex Bodnaru <alexb...@gmail.com> wrote: > >> the sax approach seems very simple and reliable. >> i hope i will soon be able to load data. >> >> thanks again, >> alex >> >> On Fri, Oct 12, 2018 at 12:22 AM Alex Bodnaru <alexb...@gmail.com> wrote: >> >>> thank you dave. >>> >>> i understand i should have zipped the code. >>> or somehow upload it. >>> yes sqlalchemy is a huge piece of code, that should be a target for any >>> data related project. >>> my next aim is doing the same dml in a bunch of sax parsing events. >>> please step in and give your opinions. >>> >>> best regards, >>> alex >>> >>> On Fri, Oct 12, 2018, 00:05 Dave Kuhlman <dkuhl...@davekuhlman.org> >>> wrote: >>> >>>> Attached is Alex's message about his progress with support for >>>> SQLAlchemy. I've deleted his attachments (the python code) because >>>> it made the message too big to send to the list. If anyone wants >>>> this, please let me know and I'll send those attachments to you >>>> separately. >>>> >>>> Alex, thank you for this work. I've been experimenting with this >>>> capability, myself. But that's mostly in an attempt to learn more >>>> about SQLAlchemy. So, I'll defer to you on this. Maybe if I do >>>> learn more about SQLAlchemy, I'll be able to offer help for your >>>> work. I'll take a look at your code. >>>> >>>> Dave >>>> >>>> -- >>>> >>>> Dave Kuhlman >>>> http://www.davekuhlman.org >>>> >>>> >>>> >>>> ---------- Forwarded message ---------- >>>> From: Alex Bodnaru <alexb...@gmail.com> >>>> To: generateds-users@lists.sourceforge.net >>>> Cc: >>>> Bcc: >>>> Date: Thu, 11 Oct 2018 22:00:51 +0300 >>>> Subject: generate sqlalchemy model >>>> hello dave, all friends here, >>>> >>>> i'm attaching my second cut at a builder and loader for sqlalchemy. >>>> the changes are: >>>> support for minidom, dynamic editing of the models list to generate and >>>> the primary keys, support for multiple model files in sqlalchemy, better >>>> checking of the files that may be overwritten etc. >>>> what i'm still about to do, hopefully with your help, is a sax iterator >>>> for the data loading, since both etree and minidom lead to memory overflow. >>>> also, class aliases, won't be hard to implement. >>>> -- >>>> -- >>>> alex >>>> >>> >> >> -- >> -- >> alex >> > > > -- > -- > alex > > > -- > -- > alex >
_______________________________________________ generateds-users mailing list generateds-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/generateds-users
