Why is an incident created by email in the Oracle B2C Service console displaying the wrong characters when compared to the original email?

Environment:

Foreign language, ISO, Email header

Resolution:

The language display, or transcoding, for an incident created by email is based on the actual language encoding from the email header (Content-Type: text/html; charset="ISO-8859-1", for example). This charset is used to display the message. The customer may not know what charset was used when the email was created.

How the message displays may be different than what is set on the incident itself, as the charset does not impact the language detection on an incoming email. Language detection occurs on every message, and a proper lang_id is set on the incident if:

• The site has a language interface matching the detected language of the message.
• The content and subject have sufficient content to evaluate against the language-detection software.

Setting the language ID on an incident based on email content is determined as follows:

1.  Language-detection software reviews the content of the body (and subject if necessary) in order to determine the language in the message. If the site has an interface of the matching language installed, then the code will set the language based on what is detected.

2.  If the language cannot be determined from the content or subject, or the detected language does not have a matching interface, then the language ID is set to the language of the interface for the mailbox to which it was sent.

Incidents generated for techmail processing errors will have the language ID set to match that of the current interface.

For additional information, refer to Answer 2663: Language setting for incidents received via email.