Christian Weiske [2010-10-26 06:18 UTC]- Please follow the PEAR directory structure
- use "File_Edifact_" as class prefix
- put the classes in the correct directories (split by _)
- put the base edifact classes into File_Edifact_Base_
David Jean Louis [2010-10-26 11:54 UTC]Hi Reinhard,
Back in 2006, at work and on my paid time, I implemented a full featured EDI php library, it currently supports UN/EDIFACT only, but the goal was also to support X12 and EANCOM standards.
*All* UN/EDIFACT messages and revisions are supported, from 1994 (D94B) to 2007 (D07A).
What I did was implementing a generic parser that parsed edifact message structure stored in xml files generated by a php script (it is not included in the source, I think I've lost it :().
It's a huge project, pear packing is not terminated, there's still many things to do (for example, the name should be File_EDI and not EDI as Christian noted).
If you're interested on what needs to be done, I'll try to write a todo list, but as far as I remember, the goal was to provide a core package (the Parser and Element classes), and subpackages for each revision of the standard xml data files (I think this part needs a custom role for PEAR packaging).
I've never had time to finish it and I do not plan to work on it again either, to be honest it was not a very fun experience to write a parser for this prehistoric and cryptic format :/
It's ok to open-source it, so you may want to have a look at the sources...:
Daniel O'Connor [2011-07-03 02:20 UTC]Hi Reinhard,
This package appears to be missing the source links (404). Also, it's been some time since activity has taken place.
Can you let us know if you still want to go ahead with this package?
Reinhard Mayr [2012-07-20 21:01 UTC](1) Christian, thank you for your comments. I tried to implement them to the best of my abilities. I am afraid I do not understand what kind of directories you are proposing (3rd comment) and what you consider to be a "base class" and why it should be renamed. -- Please let me know what you think that still needs to be done in the code in order to be ready for accepting. Thanks!
(2) David, your work is very impressing! However, my resources are also limited, therefor I am afraid I have to stick with the rather simple aproach I have chosen. Thanks anyway for letting us know, perhaps some day someone will make use of it!
(3) Daniel, I apologize for the bad links! They are now fixed and updated and point to a more reliable server now.