How to handle special chars when importing issue-data from a trac into jira installation?

[Jan Philipp]

My steps are:

1. Copy the trac.env respective make a archive containing the trac.env
2. Upload the archive with the "External system import", using "Trac"

Except the timestamp issue (1.1.1970), all data were imported correctly. But the texts seem to be missconverted (we have german texts with umlauts).

How I am able to instrument the importer respecting the charset conversion? Or should it be handled by sql directly (only)?

Enviroment

• Trac 0.11.7 on a RHL
• JIRA 4.3 on a Mac OSX (Demo-Setup, with HSQLDB)

[John Garcia · Accepted solution]

I think if you were going to do anything, it would look like:

https://confluence.atlassian.com/display/JIRA/Removing+invalid+characters+from+XML+backups