DataFormatFileNameConventions.README.txt

This document describes a filename convention for describing data which is 
ready to be part of the Prices Paid database.

The Prices Paid database at present consists of a directory of files
following the following pattern:

[Whateveryouwant]-pppifver-[FormatName]-[FormatNumber]-[FormatVersion].csv

where 
1) "Whateveryouwant" should not contain the string "pppifver"
2) FormatName should not contain a hyphen (e.g. FedBid or OS2)
3) FormatNumber is a number (no dashes, no decimals.)
4) FormatVersion is a four-digit version number separated by hyphens
 like :0-0-0-1 or 13-2-3-14

The idea here is that the FormatName is human readable, the FormatNumber 
truly identifies a growing set of formats, the Format Version allows us to 
revise the versions.

The purpose of this system is to mark the format of a file.  At the 
time of this writing, there will be two formats: RevAuc (1) and OS2 (2).

We will probably write python code that will decode the filename into 
the python file that knows how to read the data into something resembling
a standard format.  Unbeknownst to the user, fields which don't match
our standard format will still exist, although they won't be rendered in 
grid except as a uniform field.

Their are only 5 fields you have to have at present:
1) Number of Units
2) Unit Cost
3) Purchasing Agency
4) PSC code 
5) Vendor

Both the FedBid and the OS2 data sets have some fields which don't fit 
into these 5 fields, but will still be searchable.