Tuckey URLrewrite adding hostname twice for /browse/ links?

[Josiah_Sansone]

We recently migrated our JIRA and Confluence instances to new servers with new URLs.  All our in-context links to JIRA in Confluence updated to the new base url correctly, but unfortunately there were a lot of hard links created by users which did not update.  We have redirects for the old DNS addresses to the new servers, but it doesn't change the address; if the user uses these as is, they get SSL errors and base URL mismatch errors. So we set up a Tuckey urlrewrite filter to fix the base URL.  This works great with the exception of links that have the "/browse/" sub-path in the URL.  For these links, for some reason they get the base URL inputted twice in the corrected URL, which results in an error.  An example would be something like this:

Like

https://oldjira.company.net/browse/INST-1331

Will resolve to
https://newjira.company.com:8443/browse/https:/newjira.company.com:8443/browse/INST-1331

This is an example of what our urlrewrite looks like:

<rule>
<name>Domain Name Check</name>
<condition name="host" operator="notequal">newjira.company.com:8443$</condition>
<from>^(.*)</from>
<to type="redirect">https://newjira.company.com:8443$1</to>
</rule>

I've tried this a few different ways, including use multiple rules with host equals instead of host notequal, and they all result in the same thing.  ALL other links rewrite correctly, but links with /browse/ get a "double" redirect.  

Any suggestions?

Josiah

[Jack Nolddor _Sweet Bananas_]

Hi

Could you try this way?

    <rule>
        <name>Domain Name Check</name>
        <condition name="host" operator="notequal">newjira.company.com</condition>
        <from>(.*)</from>
        <to type="redirect">https://newjira.company.com:8443/$1</to>
    </rule>

Regards