KB Article #179111

EDIFACT messages with LF or CR LF characters between each segments

Problem

-- sending EDIFACT messages with ' as a segment separator but also with a new line after each segment.
This is good for readability, but this implies that the segment separator is '/n.

Ex :

UNA:+.? '
UNB+UNOC:3+xxxx:14+xxxxx:14+170914:1647+17'
UNH+1+DESADV:D:96A:UN:EAN005'
...
...

-- such kind of message is detected as an InHouse message, not a EDIFACT message.
when removing the LF in the file, it is detected as a EDIFACT message.

-- is there a way to handle EDIFACT messages with <LF> ?

Resolution

* according to the standard, a message with 2 characters as segment separator is not a correct EDIFACT message

* there is no native way to remove such LF or CR LFcharacters

* a pre-processing has to be done (using an inline processor or an MBC) to remove the characters, so that it can be detected as a valid EDIFACT message