A livereload server for wsgi

[lepture]

Now, livereload will be a library rather than an application. in version: 2.0.0

from livereload import Server

server = Server(wsgi_app)
# run a shell command
server.watch('static/*.stylus', 'make static')
# run a function
server.watch('foo.txt', lambda: print('foo'))
server.serve()

With this snippet it will serve an wsgi application, and it will insert the livereload.js script into the response body.

[lepture]

@miracle2k any suggestions?

[miracle2k]

A significant challenge for me was integrating the tornado livereload server that currently exists with werkzeug's "runserver". Two thoughts:

1. Because both want to run a main loop, I have to put one in a thread.

2. werkzeug's runserver has a bunch of nice functionality, most significantly auto-reload on source change - which was an additional challenge because it starts a new process thus leading to a second attempt to run the live reload server.

The new approach would solve (1) nicely.

For (2), the API I would imagine to API to look something like this:

 wsgi_app = livereload(my_wsgiapp)
 werkzeug.runserver(wsgi_app)

But this presumably cannot be done without getting rid of tornado and thus a huge rewrite.

So I would be happy if instead of letting livereload serve I could still optionally do:

 server = Server()
 server.watch()
 Thread(run=server.serve).start()
 werkzeug.runserver(my_wsgiapp)

[lepture]

@miracle2k you will not need werkzeug's runserver. This server will auto reload on source change too.

[blueyed]

Works good for me, and can easily replace Django's runserver by using this as wsgi_app:

import os
os.environ.setdefault("DJANGO_SETTINGS_MODULE", "project.settings.local")
from django.core.wsgi import get_wsgi_application
application = get_wsgi_application()

The change from current master would be to not have a global Task object, but add them to the server directly?
Makes sense.

[lepture]

The latest code is no wsgi branch.

[lepture]

released.