[rescue] XML Schema reversing

[William Kirkland]

I am sure I could reformat the XML to something else, using awk ...
though I do not fully understand the details, so am not sure if that  
would help.

On Apr 10, 2007, at 16:31 PDT, rescue-request at sunhelp.org wrote:

Message: 3
Date: Tue, 10 Apr 2007 17:37:41 -0400
From: David Muran-de Assereto <dmuran at tuad.org>
Subject: [rescue] XML Schema reversing
To: Rescue <rescue at sunhelp.org>
Message-ID: <B4A87758-E8A8-4EB1-90FA-5AD33E5D97DB at tuad.org>
Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed

Not exactly a rescue issue, but you guys are smart and of widely
varied experience:

I am in a situation where I need to reliable reverse a 29k line XML
schema, in three parts, back into human readable English. The schema
is generated from a list of database fields, data types, database
field lengths, and a length constraint on each field. Although I can
go back to the source database for the information, the situation is
such that I need to be able to reliably assert that the schema is, in
fact, an accurate representation of the field names, types, and
allowable lengths.

I have googled, and although I have found a lot of information on
validating XML documents, I have found little information on doing
what I need to do. The only examples I can find of doing this are

1) Microsoft-centric
2) Involve comparing the schema with the database, which I cannot do.

What I need to do specifically, is to parse the schema and produce a
tabular output containing the field name, data type, and maximum
allowable length.
The actual schema(s/e) are in three parts: The first file is a list
of user-defined data types. The second file is a description of the
database. The third part contains a subset of the database
description, with only those elements which have data length
constraints which vary from the database.

Any pointers?

Dave

--
bill.kirkland at gmail.com