Thread: Determining correct table order for insert or drop statements to satisfy foreign keys

Hello,

First, sorry if this has already been covered, though I didn't find anything 
similar in the archives. Also sorry if this is the wrong list for this type 
of question, though it seemed to be the most relevant list for this question.

On to the question:

I'm writing a simple import script that reads in an excel file that a customer 
fills out. Each worksheet represents a different table and the tables' fields 
are columns in each worksheet. 

The tables represented in the file are variable, I have no idea what table 
combinations will be in the excel file when it's sent to the script. 

What I need is a way for the script to determine what order it should process 
the worksheets in the excel file.

For example, you have this table structure (each table would be a worksheet):

PERSON (1) <--> (many) ORDER (1) <--> (many) ORDER_ITEM

Obviously, the people would have to be inserted first, then the orders, and 
finally the items, to keep the foreign keys between the tables happy.

Is there a mechanism in PostgreSQL that would tell me this order?  Or is it 
something that would have to be assembled from analyzing the schema?  Anybody 
needed to perform an operation like this before? Any help/direction would be 
greatly appreciated.

Thanks for your time and help,
Pete Hanson

On 11/2/06, Peter Hanson <lists@pkhanson.com> wrote:
What programming language are you using?  Many languages provide mechanisms to ask the database for metadata about tables including primary keys, indexes, data types and foreign keys.  That is one option for you.

--
==================================================================
   Aaron Bono
   Aranya Software Technologies, Inc.
   http://www.aranya.com
   http://codeelixir.com
==================================================================

Hi Aaron,

I'm using PHP5 as the programming language.

Thanks,
Pete


On Thursday 02 November 2006 16:58, Aaron Bono wrote:
> On 11/2/06, Peter Hanson <lists@pkhanson.com> wrote:
> > Hello,
> >
> > First, sorry if this has already been covered, though I didn't find
> > anything
> > similar in the archives. Also sorry if this is the wrong list for this
> > type
> > of question, though it seemed to be the most relevant list for this
> > question.
> >
> > On to the question:
> >
> > I'm writing a simple import script that reads in an excel file that a
> > customer
> > fills out. Each worksheet represents a different table and the tables'
> > fields
> > are columns in each worksheet.
> >
> > The tables represented in the file are variable, I have no idea what
> > table combinations will be in the excel file when it's sent to the
> > script.
> >
> > What I need is a way for the script to determine what order it should
> > process
> > the worksheets in the excel file.
> >
> > For example, you have this table structure (each table would be a
> > worksheet):
> >
> > PERSON (1) <--> (many) ORDER (1) <--> (many) ORDER_ITEM
> >
> > Obviously, the people would have to be inserted first, then the orders,
> > and
> > finally the items, to keep the foreign keys between the tables happy.
> >
> > Is there a mechanism in PostgreSQL that would tell me this order?  Or is
> > it
> > something that would have to be assembled from analyzing the
> > schema?  Anybody
> > needed to perform an operation like this before? Any help/direction would
> > be
> > greatly appreciated.
>
> What programming language are you using?  Many languages provide mechanisms
> to ask the database for metadata about tables including primary keys,
> indexes, data types and foreign keys.  That is one option for you.

On 11/2/06, Peter Hanson <lists@pkhanson.com> wrote:


--
==================================================================
   Aaron Bono
   Aranya Software Technologies, Inc.
   http://www.aranya.com
   http://codeelixir.com
==================================================================