Character changing (Somebody need it?)
Is there any plans to include this feature to the EDI system?
In my case it is quite routine to get EDI documents with characters inside like 0x92 or 0x96 or any from restricted XML chars set. Is it because the partners have the old systems to generate these EDI or they work on different languages... Anyway this is ordinary.
In Covast Accelerator this feature is very useful (to decode these characters in what we want, say '`' --> '"')
Thanks,
Not quite sure if I understand your issue here. Are you suggesting having a filter expression in UI where you can suggest the conversion you mentioned above, and everytime you see any of those character in the payload, you just apply the conversion based on the filter?
Yeah, something like this.
Now seems all my EDI applications have to include a custom preprocessor to convert all "bad" symbols to something more applicable.
For instance:
N3*RYCHMOND.B.C.V6X CTC:MR.MC`RONALD *TANG TEL;1-604-279- V6X
- here we have a prohibited for EDI pipeline symbol ` (not a ' or ", but `)
Or it could be "bad-xml" symbol as 0x92 etc. We have in "good" and "bad" symbols as it described in
| [2] | Char | ::= | [#x1-#xD7FF] | [#xE000-#xFFFD] | [#x10000-#x10FFFF] | /* any Unicode character, excluding the surrogate blocks, FFFE, and FFFF. */ |
| [2a] | RestrictedChar | ::= | [#x1-#x8] | [#xB-#xC] | [#xE-#x1F] | [#x7F-#x84] | [#x86-#x9F] |
If those restricted symbols are passed to the xml we've got errors.
Then this xml can go to SQL where we have different symbol-set requirements...
I think this preprocessor should be a part of EDI system because it is in everyday case. I inderstand that it is not beautifull feature, but it is very usefull.
BTW it can be implemented as a separate pipeline component and we could use it for other applications, not only EDI.
It can be implemented without UI, .config file is enough.
Have the same problem now with EDIFACT.
Partner uses a lot of restricted symbols in EDI fields.
It is the standard situation and I need the preprocessor to replace or/and cut these symbols.
This feature is not a part of R2 release, and given the stage we are in the product cycle, it will be not included in the final release. It will be considered for future release or probably SP1 at best.
As you previously mentioned, a workaround would be to include a custom component in pipeline that parses the incoming interchange and replaces all the "bad" values according to the mapping system.
Thanks
Mohsin
