Parsing out EDI messages

A parser I designed to work with HIPAA EDI files. It reads in files and spits out the individual segments without terminators.

Requires Python 2.3 or greater. (Use can probably use Python 2.2 with from __future__ import generators at the top...)

This parser is currently in use to work with 150-200MB of EDI transactions a day and has proven very reliable. Some considerations I had that not everyone may be worried about:

-Some, but not all, of my data comes from our mainframe and is stored 80 bytes to a line. The line with the final IEA of each message is padded out to 80 bytes. I had to account for extraneous 's and possible linefeeds. However, since the linefeed is a valid segment terminator, I couldn't just ignore them.


-We bundle multiple interchanges within the same physical file. Not all of our trading partners use the same delimiter set so i had to be able to adjust at any point to a new delimiter set.

-At the very high level, I assume that the interchanges aren't complete garbage. (i.e.: All interchanges are wrapped in a valid ISA - IEA envelope.) No sort of syntax or structural checking is done.