Revisions for klutz/lib/Klutz.php

Happy New Year

Marcus Ryan has assigned his copyrights to The Horde Project.

Add a few more image types - I had a tiff show up today.

Some clean up, remove a bunch of array_key_exists calls.

More reference fixes.

Add menu accesskeys and a bunch of HTML cleanup.

Remove "@access public".

Remove extra stuff from @var lines in phpdoc.

Clean up phpdoc for constants.

Remove "optional" from phpdoc.

Don't use array_push() when we don't need to.

Remove @version.

use new menu

Don't show an error if the comic isn't listed.  Might be a good idea to
notify the user to reset their preferences, but notify has already been
called by the time we get here.

Apps no longer need to call Help::javascript() - and indeed they shouldn't,
since it no longer exists.

Fix package names.

No more HORDE_LIBS.

Bring RELENG_1 version of klutz back in line with HEAD, most notably convert
from Snoopy to HTML_Request (Snoopy clips images).  Getting ready to finally
start the release cycle on this one.

WARNING: There is a good chance that there are still some bugs in here.  This
was a major update, most of it done by hand.  I've tested what I can, but
no promises of total stability.  That said, this code has been working in HEAD
for months with no known reliability issues.

I haven't touched languages/translations so that needs to be done yet.

Still need install instructions

move Menu.php to a framework package.

Show notifications only from the requested message stack.

PHP 5 fixes.

Last menu.php files.

Nuke constants in favor of strings to avoid some issues of having the
constants defined in javascript files.

Nuke a lot of constants.php files.

Specific Reason: constants should be defined in the application's base
class file UNLESS constants are needed in one of the configuration
files (the case for all those remaining). Only then should you break
things out to a seperate file (and even then, we may want to just
include the app's library file).

General Reason: Including files is *very* expensive. Short files are
pointless if we can avoid them; it's much cheaper to have slightly
longer files than multiple short ones (especially when we're going to
include all the short files anyways).

Oops.

"somewhere before here." :-)

* Need to include horde/lib/String.php here (or somewhere before here)

Rename String::strtolower() to String::lower() and strtoupper() to upper().

Add String:: class and move all string manipulation methods from NLS:: to
String::

More strtolower() replacements.
I'm not sure if I catched all in lib/Comics.php and didn't catch too much.

Not all comics have homepages defined (though they should...but if it doesn't
don't complain - just don't make it a link).

phpdoc...lots of phpdoc.  Please check it if you get a minute...I'm guessing
there are a lot of typos, and it's my first real stab at phpdoc-ing :)

Also cleaned trailing whitespace while I was in there.

Tags: RELENG_1

style - single quotes, etc.

Add a user preference for whether to show unselected comics on the
main browse page.

Clean up (tabs to spaces, extra space at EOL)

Add 'nohistory' option to comic definitions.  If this property is true,
the backend will not try to fetch any comics except today's.

Also, added another ignore to dork tower - should increase fetch time.

Warning: This may be a bit buggy.
Now supports 'days' => 'random' in the comic definition.  This special
setting means the system should try to retrieve the comic every day, but
not display an error if we don't have it (or if the file exists, display
templates/comics/random.inc).

Also added templates/comics/missing.inc to display when an image that SHOULD
exist doesn't.

Somewhat better sorting (and now sorts by author).

Again, these are not necessarily internationalization friendly...

Implement the "days" argument.  Won't try to fetch or display a comic on
days where it's not wanted.  Default is wanted every day. Also added extra
parsing to handle people putting Thursday, Thu, Th, R.

Is there a need to localize this argument, and is it something that could be
reasonably done?  Biggest complication I can see is comics.php.dist would
have to be localized...

Now sorts comics by name (and a stub for sort by author).  Multi-tiered
searches now work (pages requiring more than one search to find).  There
may be other stuff, but I'm too tired to remember it now (12 hours on this
today...am I nuts?! :) )

Okay...klutz can now handle comics that are retrieved directly; that is those
that do a single http get.  It does not have local storage support yet, so
it's PAINFULLY slow and inefficient, but it's enough for a preview if people
want to play.

Hopefully I got them all this time...