JRedirector version and change information 1.2 Features added: - Provide data for the Undo tab entries produced by the JRedirector - Change the internal storage of logging data to be more efficient. This change is automatically applied to existing JRedirector instances the first time the log is accessed, either during a redirect or when the administrator looks up the log history in the Zope management interface. This should also fix Tracker issue 197 if it was a real issue (my own testing did not support these conclusions). 1.1 This is the "Josef Meile Appreciation Release". Virtually all suggestions for improvements in this release came from Josef. Features added: - Referrers to faulty URLs are now tracked as well. They are listed on the Log tab. - Mappings and Log views have been reorganized a little to be more obvious about what they show and to avoid having to scroll the browser window sideways if a URL is overly long. - All displayed URLs are now clickable and will open in a new window. This allows for quickly testing the old and new paths on the Mappings tab as well as the referrers listed on the Log tab. - Old paths that are specified with a trailing "/" (slash) characters will have it stripped to prevent matches on directory paths failing because most people just don't use trailing slashes when referring to directories. 1.0 Features added: - Case-insensitive matching of requested paths to redirects (first suggested by Josef Meile). - Ability to clear the current logs from the ZMI (suggested by Josef Meile). - Multiple paths can now be redirected to a single new path by specifying a "Wildcard Mapping". Wildcard mappings are used for those requested URLs that... - have not matched any other non-wildcard rule - are *underneath* the path specified as Old Path when setting up the mapping. - Added more allowed HTTP response codes and an explanation of these codes in README.txt Bugs fixed: - Highlight the correct ZMI tab after submitting anything from the "Mappings" and "Log" tab. - Use REQUEST.PATH_INFO instead of REQUEST.URL because PATH_INFO always contains the full path during traversal, whereas URL only contains the path up to the currently traversed element. This fixes a bug that prevented redirecting to a path that had nothing in common with the requested path. Thanks to Josef Meile for pointing this out. 1.0beta2 Features added: * __call__ is now equivalent to calling the redirect method, making usage even easier. Now you can just activate it like this:: redirector_object(REQUEST) 1.0beta1 First working code version 0.5 Started putting the files together