[sloppypoet]

We are running on OTM 4.5 and our Python scripts crash whenever we receive a non-ascii character in an order file. We have been able to replace the non-ascii characters with a similar ascii character but this only works for a one to one replacement. We are also receiving ½ characters (U+00BD), which crash the python scripts, and are unable to replace these with a single character. Has anyone run across this issue and, if so, how did you resolve it?

Thanks!

[bmj_23]

Hi - I seem to remember running in to a similar issue when using python scripts to export data out of the database which contained some non-supported characters (it was "'s in my case I believe).

Unfortunately, I dont seem to remember a solution to that so in your case would be tempted to log that with Oracle support (although I am not sure that either a) 4.5 is still supported and that b) the python scripts that were used in the Glog days are still supported by Oracle.

Would definitely be interested to hear the outcome of that dicsussion with support.

[AlejandroBarba]

Hi,

We had a similar issue but for inbound order transmissions where the character "&" is part of the Shipper contact. The issue was solved by adding the "amp;" suffix so we would have "&" instead of "&".

If you are talking about inbound transmissions, I would refer to the Integration manual to check how to change the encoding in the header to include non 7bit ASCII characters.

Hope this helps,

Alejandro B.

[bmj_23]

Hi, just as follow up, in the python file site.py what is the value of "encoding=" set to ?

I seem to remember having to modify this from encoding='ascii' to encoding='UTF-8' to allow certain characters to be processed correctly.

However, before changing anything, I would get the definitive word from Oracle support. The character support has to be co-ordinated across a number of different layers e.g. database so I dont think it is as simple as just changing site.py.

Post what you find out.