[
https://issues.apache.org/jira/browse/MIME4J-218?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14191660#comment-14191660
]
Oleg Kalnichevski commented on MIME4J-218:
------------------------------------------
Wolfgang
To make things a little to manage I committed your patch with my changes to SVN
trunk. But please do consider contributing a follow-up patch to replace boolean
(lenient / strict) flag with a Charset to be used as a fallback. See my
comments in https://github.com/apache/james-mime4j/pull/1
Oleg
> Content-Type Fallback Character Set
> -----------------------------------
>
> Key: MIME4J-218
> URL: https://issues.apache.org/jira/browse/MIME4J-218
> Project: James Mime4j
> Issue Type: Bug
> Affects Versions: 0.7.2
> Reporter: Rickard Ekeroth
> Fix For: 0.8.0
>
>
> Would it be possible to add a feature that would allow for specifying a
> fallback character set to use when the character set in a 'Content-Type'
> header is not recognized by Java? In the old 0.6.2 version, that we used
> before, the character set 'ISO-8859-1' was used as a fallback but in the
> 0.7.2 version an UnsupportedEncodingException is thrown when the parser
> encounters an unknown character set in a Content-Type header.
> Here is the relevant part of the exception stack trace:
> Caused by: java.io.UnsupportedEncodingException: x-user-defined
> at sun.nio.cs.StreamDecoder.forInputStreamReader(StreamDecoder.java:52)
> at java.io.InputStreamReader.<init>(InputStreamReader.java:83)
> at
> org.apache.james.mime4j.message.BasicTextBody.getReader(BasicTextBody.java:49)
> We receive, parse and archive a vast number of confidential e-mail messages
> (for which we use Mime4J) and every now and then we get an e-mail message
> that contains a non-standard character encoding name (in this case
> 'x-user-defined'). With the old (0.6) Mime4J version we were still able to
> parse and read most of those e-mail messages because of the fallback
> character set in the parser.
> I can unfortunately not post the entire message here but the content-type
> header that caused the above exception looks like this:
> Content-Type: text/plain; charset="x-user-defined"
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
