I'm attempting to create a generic python script that will take an EDI document specification and a database of transactions and be able to generate a raw EDI file (ansi X12).
I'm looking for ideas on how best to use the strengths of Python to implement this. I've initially tackled the EDI 812 specifications and narrowed this down to a specific implementation by a company. This narrowed down version is what I want to achieve for output. But I want the script to be generic enough such that I can plug-in another EDI specification + a database transaction set and it will output the proper raw file accordingly. My initial thought was to use: 1. .ini files to declare the EDI configuration 2. Use SQLAlchemy as ORB to simplify access to database objects. INI file configuration: * A "root" INI file indicates other INI files that define each segment of the EDI document. * Each segment INI defines the data elements of each segment and the behavior of the segment (one instance or multiple-instance as in a loop, sequence order, segment code, etc.) * Specify data elements as either constant, system function (like datetime), database field or object method (object being the segment's) * Load all the ini configuration into a "template" object. Each segment ini maps to its own segment object. DB using SQLAlchemy Gather a Trading Partner data and Credit Transaction (EDI 812 remember?) into one dictionary object Gather Credit Transaction details into another dictionary object where it can generate the multiple instance segments The heart of the matter is how to fuse together the template and the data from the dictionary objects efficiently. It should be generic enough to take another set of data dictionary and another template to generate a completely new EDI document. I'm stuck at this fusing together thing. Is this a good approach? Is there an easier to implement approach? Comments, suggestions, questions please. Allan -- http://mail.python.org/mailman/listinfo/python-list
