Turned out a test wasn't really testing what it should have been testing
and thus did hide a bug. Still puzzles me why it still worked some
times.
This patch also sets the default iteration count for bmd5 and pmd5 to 8.

no problems found

This patch applies the correct pagename cleaning to namespaces as well.
Namespaces should follow the same rules as pagenames but due to a bug it
was possible to have namespaces ending in a underscore.

If you used such a namespace it will be inaccessible after applying this
patch. You should rename the namespace directory on the filesystem
(removing the trailing underscore). You old links will then work
automatically again. Rebuilding the search index is recommended.

This fixes a problem with running the minified jQuery through the
compressor.

Note, any filenames encoded with the previous SafeFN scheme will need to
be converted to the new scheme.  Users of the old SafeFN scheme should
not use this new scheme until after converting their filenames.

All linebreaks before p_close have been removed.

this splits the long auth_cryptPassword() function into many member
functions of a new class PassHash which should make it more
maintainable and reusable for other projects.

This also adds two new methods djangomd5 and djangosha1 as used by the
popular python framework Django.

Maybe the auth_cryptPassword() and auth_verifyPassword() functions
should be deprecated in favor of using the class directly?

This function abstracts checking a given user and her groups against a
given member list (as used in the superuser and manager options).

It is also used in auth_isManager() and auth_isAdmin(), unlike the
previous function, this one skips the nameencode step as it should be
unnessary here (all input is given decoded).

The test cases where extended by some non-ID user and group names.

People with non-plain auth backends should check that their
administrator and manager setups still work as expected

As of VIM 7.3 it is no longer possible to specify the encoding in the
modeline. This gives an error message whenever such a file is opened,
thus this commit removes the enc setting from the modeline.

SafeFN encoding now always end the encoding block by appending a dot at the end of the file name. This is necessary since the file name may get an extension which is not encoded.

Neither I nor Andi has any idea what this file is for and why it explicitly //grants// permission and why its documentation is wrong.