Manage is an example project that demonstrates how simply and elegantly we can
reuse jquery and jquery-claypool code that helps us manage our database backended
projects from the browser and the command-line. Much of the same code is used
to drive the server-side application as well.

If you would like to develop locally you can get the google apps engine development
environment (is really just jetty with a mini bigtable backend built in.) The
appengine dev is here:

Google AppEngine SDK

All you need to do is unzip it (we’ll assume unix-like paths) and run:

$> /path/to/appengine/bin/dev_appserver.sh /path/to/manage-js/

At this point you should be able to open up a browser see

http://localhost:8080/

The html interface provides a small group of commands that let you round trip
data from an inital_data.json file in you local development environment as well
as browse and edit items in the current database. They two interfaces are not really
dependent on each other, but are aptly named ‘manage’ and ‘items’.

Django developers would be very familiar with the specific command names. Here’s
the gist of the ‘manage’ interface:

• help — displays this information
• load — generates and saves items to database
• dumpdata — provides a json dump of the current database
• reset — removes all data from the database
• syncdb — loads a previous state of the database from /data/inital_data.json

The items interface will simply provide a large table of item links. Each item
link will bring you to a form, populated with the current state of the item, and
allow you to modify and save it.

The command line interface is a developer friendly way of making the html interface
available in a shell. This is a very concise reuse of code which demonstrates
the elegance of jquery-claypool as a ultra light framework. At this point only the
‘manage’ interface is available via the commandline, though the ‘items’ interface
is also very possible.

Manage is very simply used at the command line. The general command line
pattern is.

$> /path/to/manage-js/manage.sh <command> [args]

Most commands dont require any args but when you use ‘dumpdata’ at the html
interface you wil be prompted to select a local file location to download it as
(/path/to/manage-js/data/initial_data.json is the convention to support syncdb). The
same choice is true for ‘dumpdata’ at the command line so the dump file must be
specified as an absolute url (include the file:// protocol).

These should all work:

$> /path/to/manage-js/manage.sh help
$> /path/to/manage-js/manage.sh load
$> /path/to/manage-js/manage.sh dumpdata file:///path/to/manage-js/data/initial_data.json
$> /path/to/manage-js/manage.sh reset
$> /path/to/manage-js/manage.sh syncdb

The shell is an explorative arena that loads all the prerequisites for you to
be able to play with the same environment available in the browser html interface.
Also if you have to execute several cammands in succession it will be faster
because, at the command line, the browser page is loaded for each command, while
the shell can execute multiple commands without reloading the page.

In order to use it, you only need to execute 2 commands:

$> /path/to/manage-js/manage.sh shell

which will provide the following shell environment:

Entering interactive shell, please js> load('shell.js');
Rhino 1.7 release 2 2009 03 22
js>

Then just load the shell as suggested:

js> load('shell.js');
[ Envjs/1.6 (Rhino; U; Mac OS X i386 10.5.8; en-US; rv:1.7.0.rc2) Resig/20070309 PilotFish/1.2.0.11 ]
js>

The commands available in the shell require that you interact with the html
interface using javascript in order to produce the events that trigger the same
event.

[ Envjs/1.6 (Rhino; U; Mac OS X i386 10.5.8; en-US; rv:1.7.0.rc2) Resig/20070309 PilotFish/1.2.0.11 ]
js> var _ = MyProject.Commands;
js> _.help();
@ INFO: [Thu, 25 Mar 2010 03:01:26 GMT] {MyProject.Commands} help@
@ INFO: [Thu, 25 Mar 2010 03:01:26 GMT] {MyProject.Views.Manage}@
@ @help -- displays this information
load -- generates and saves items to database
reset -- removes all data from the database
dumpdata -- provides a json dump of the current database
syncdb -- loads a previous state of the database from /data/inital_data.json
@ @ INFO: [Thu, 25 Mar 2010 03:01:26 GMT] {MyProject.Commands} complete

You can also try the other commands…

js>_.dumpdata('file:///path/to/manage-js/data/initial_data.json');

Of course the HTML Interface can be explored while in the browser with Firebug,
but you can also use any of the commandline commands in the rhino debugger by
adding the debug parameter.

These should all work:

$> /path/to/manage-js/manage.sh debug help
$> /path/to/manage-js/manage.sh debug load
$> /path/to/manage-js/manage.sh debug dumpdata file:///path/to/manage-js/data/initial_data.json
$> /path/to/manage-js/manage.sh debug reset
$> /path/to/manage-js/manage.sh debug syncdb

At some point soon we will be integrating the the manage-js project with jquery-claypool
directly, but believe its an interesting, and general enough idea to merit its
own project space for awhile.

Enjoy!

Thatcher