z is the new j

First, there was j: http://github.com/rupa/j/.

Then, there was j2: http://github.com/rupa/j2/, with some python added. Made
it easier to experiment with some stuff I'd been thinking about.

Now there's this, where I rewrote j2 back in bash. I like it. Sorry about all
the repos , but it kind of grew this way.

The biggest change from j is the use of 'frecency' for weighting. Directories
that have low ranking but were accesed recently, will quickly have higher rank.

The -r switch preserves the old behavior.

# maintains a jump-list of the directories you actually use
#
# INSTALL:
#   * put something like this in your .zshrc:
#         . /path/to/z.sh
#         function precmd () {
#             z --add "$(pwd -P)"
#         }
#   * cd around for a while to build up the db
#   * PROFIT!!
#
# USE:
#   * z foo     # goes to most frecent dir matching foo
#   * z foo bar # goes to most frecent dir matching foo and bar
#   * z -r foo  # goes to highest ranked dir matching foo
#   * z -t foo  # goes to most recently accessed dir matching foo
#   * z -l foo  # list all dirs matching foo (by frecency)