Sugus » code snipplets

Websockets tutorial/example with pywebsocket

Pedro Assuncao — Fri, 18 Dec 2009 12:37:24 +0000

As everyone already knows, Google Chrome now supports websockets. In essence, this allows you to keep a connection open with a webserver indefinitely (analogous to typical sockets) and send data bi-directionally. Unfortunately Chrome is the only browser currently supporting this, but I’m pretty sure this will change.

So I decided to give this a try and experiment a bit with it. This is my step by step process on getting a web page opening a websocket to a server and receiving the server’s date and time every second. It is based on an article by Joe Armstrong, though he uses Erlang for the server, while I decided to follow the easy road and use Google’s pywebsocket – an apache module (uses mod_python) that allows you to create handlers for websocket connections in a easy-to-use fashion. The code also contains a way to start a standalone server (i.e. not requiring apache) for testing purposes.

So here are the steps I took to get this working:

1. Create a web page

This is the code I borrowed from Joe, though slightly modified to fit my purposes (it requires jquery, by the way):












Interaction experiment

Debug



Clock
I am a clock

2. Download and install pywebsocket

Checkout the code with

svn checkout http://pywebsocket.googlecode.com/svn/trunk/ pywebsocket-read-only

Then do python setup.py build and sudo python setup.py install inside the src folder. This will install it into your python environment.

3. Being lazy, means we will change an example handler

The way pywebsocket works is delegating the connections to something they call handlers. In the pywebsocket-read-only/src/example folder you will find a file named echo_wsh.py. They have this convention where handlers are named _wsh.py. This means that when you later call (from your web page) the url http://localhost:9998/echo the server will delegate the processing of that connection to that file.

I modified that file to something like this:

# Copyright 2009, Google Inc.
# All rights reserved.
#
# Redistribution and use in source and binary forms, with or without
# modification, are permitted provided that the following conditions are
# met:
#
#     * Redistributions of source code must retain the above copyright
# notice, this list of conditions and the following disclaimer.
#     * Redistributions in binary form must reproduce the above
# copyright notice, this list of conditions and the following disclaimer
# in the documentation and/or other materials provided with the
# distribution.
#     * Neither the name of Google Inc. nor the names of its
# contributors may be used to endorse or promote products derived from
# this software without specific prior written permission.
#
# THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
# "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
# LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
# A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
# OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
# SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
# LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
# DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
# THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
# (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
# OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.

from mod_pywebsocket import msgutil
from datetime import datetime
import time

_GOODBYE_MESSAGE = 'Goodbye'

def web_socket_do_extra_handshake(request):
 print 'Connected.'
 pass  # Always accept.

def web_socket_transfer_data(request):
 while True:
 time.sleep(1)
 date = datetime.now()
 #try:
 #    line = msgutil.receive_message(request)
 #except Exception, e:
 #    print 'Foi com os porcos'
 #    raise e
 #print 'Got something: %s' % line
 #msgutil.send_message(request, line)
 msgutil.send_message(request, 'clock!%s' % date)
 #if line == _GOODBYE_MESSAGE:
 #    return

So basically whenever a new connection is made to this entry point, a call to web_socket_do_extra_handshake is made. After that, web_socket_transfer_data is called and it’s your responsibility to create the loop that receives messages and handles the flow (as you can see in the commented lines). I don’t care about that right now, since I only want to push the date and time to the client every second.

3. Start the standalone server

Go to the pywebsocket-read-only/src/mod_pywebsocket folder and run the following command:

sudo python standalone.py -p 9998 -w ../example/

This will start the server in port 9998 and use the handlers directory specified by the -w option. That is where our echo_wsh.py lives.

4. Test it

So now open your browser (remember, only chrome supports websockets currently) and open the html file your created in the beginning. Voilá, server’s date and time every second in the clock div.

Android location provider mock

Pedro Assuncao — Thu, 12 Nov 2009 15:50:42 +0000

So, yeah, I resumed playing around with android, this time version 2.0.

I’m really tempted to buy the new Motorola Milestone that should come out in Europe sometime between… now… and early next year, so I wanna be ready to create all the crazy stuff I have in mind for it

One of the things I noticed was that it’s not that straightforward to provide sample GPS data. Somehow the emulator’s GPX and KMZ functionality is not working fine for me and, even if it was, I don’t want to load the file everytime (i.e. I’m lazy).

So, after some googling around, here’s my solution to read points from a file and feed them 1 per second to the location manager, so I can finally work with them in my app:

My main activity implements LocationListener, so it can be passed to the LocationManager to receive GPS events. Here’s what I do when creating my activity:

public void onCreate(Bundle savedInstanceState) {

		super.onCreate(savedInstanceState);
		setContentView(R.layout.main);

		((TextView) this.findViewById(R.id.textView)).setText("Something else");

		// LocationManager locationManager = (LocationManager)
		// getSystemService(Context.LOCATION_SERVICE);
		// locationManager.requestLocationUpdates(LocationManager.GPS_PROVIDER,
		// 0, 0, this);

		LocationManager locationManager = (LocationManager) getSystemService(Context.LOCATION_SERVICE);

		String mocLocationProvider = LocationManager.GPS_PROVIDER;
		locationManager.addTestProvider(mocLocationProvider, false, false,
				false, false, true, true, true, 0, 5);
		locationManager.setTestProviderEnabled(mocLocationProvider, true);
		locationManager.requestLocationUpdates(mocLocationProvider, 0, 0, this);

		try {

			List data = new ArrayList();
			InputStream is = getAssets().open("data.txt");
			BufferedReader reader = new BufferedReader(new InputStreamReader(is));
			String line = null;
			while ((line = reader.readLine()) != null) {

				data.add(line);
			}
			Log.e(LOG_TAG, data.size() + " lines");

			new MockLocationProvider(locationManager, mocLocationProvider, data).start();

		} catch (IOException e) {

			e.printStackTrace();
		}
	}

This will basically setup the test location provider and read the points into a list. Then I feed that stuff to my mock location provider (just a normal thread) that will read them 1 per second and trigger the new location back to this activity. Here’s the code for MockLocationProvider:

public class MockLocationProvider extends Thread {

    private List data;

    private LocationManager locationManager;

    private String mocLocationProvider;

    private String LOG_TAG = "faren";

    public MockLocationProvider(LocationManager locationManager,
            String mocLocationProvider, List data) throws IOException {

        this.locationManager = locationManager;
        this.mocLocationProvider = mocLocationProvider;
        this.data = data;
    }

    @Override
    public void run() {

        for (String str : data) {

            try {

                Thread.sleep(1000);

            } catch (InterruptedException e) {

                e.printStackTrace();
            }

            // Set one position
            String[] parts = str.split(",");
            Double latitude = Double.valueOf(parts[0]);
            Double longitude = Double.valueOf(parts[1]);
            Double altitude = Double.valueOf(parts[2]);
            Location location = new Location(mocLocationProvider);
            location.setLatitude(latitude);
            location.setLongitude(longitude);
            location.setAltitude(altitude);

            Log.e(LOG_TAG, location.toString());

            // set the time in the location. If the time on this location
            // matches the time on the one in the previous set call, it will be
            // ignored
            location.setTime(System.currentTimeMillis());

            locationManager.setTestProviderLocation(mocLocationProvider,
                    location);
        }
    }
}

Notice the location.setTime() call. Read the comment why it is necessary. Took me forever to find this one in google

Peace and great Androiding

Grabbing title tag from web page

Pedro Assuncao — Wed, 21 Oct 2009 13:16:24 +0000

At least a couple of options, the first using BeautifulSoup:

import urllib
import BeautifulSoup

soup = BeautifulSoup.BeautifulSoup(urllib.urlopen("https://www.google.com"))
print soup.title.string

And the second one using lxml:

import lxml.html
t = lxml.html.parse(url)
print t.find(".//title").text

Related posts:Several years of email statistics – code included Double click/tap detection on android’s MapView

Django template filter: Show list of objects as table with fixed number of columns

Pedro Assuncao — Wed, 30 Sep 2009 12:14:41 +0000

I recently ran into the following problem: I needed to be able to display a list of users in a table that had a maximum of X columns. Since I could not find the solution on the Internet I decided to give it a try and here is my resulting template filter to do it:

def tablecols(data, cols):
    rows = []
    row = []
    index = 0
    for user in data:
        row.append(user)
        index = index + 1
        if index % cols == 0:
            rows.append(row)
            row = []
    # Still stuff missing?
    if len(row) > 0:
        rows.append(row)
    return rows
register.filter_function(tablecols)

Then you can use it in your templates like so:


{% for row in members|tablecols:5 %}
    
    {% for member in row %}
        
    {% endfor %}
    
{% endfor %}

            {% show_simple_profile member user %}

It will break down the “members” list into a list of lists, each of the first being a group of (in this case) 5 users max.

Have fun

Python: Sort list of tuples by second item

Pedro Assuncao — Wed, 16 Sep 2009 11:38:01 +0000

Straight from the PythonInfo Wiki:

>>> import operator
>>> L = [('c', 2), ('d', 1), ('a', 4), ('b', 3)]
>>> map(operator.itemgetter(0), L)
['c', 'd', 'a', 'b']
>>> map(operator.itemgetter(1), L)
[2, 1, 4, 3]
>>> sorted(L, key=operator.itemgetter(1))
[('d', 1), ('c', 2), ('b', 3), ('a', 4)]

Related posts:Django template filter: Show list of objects as table with fixed number of columns JIRA (...)
[continue reading] Related posts:

Django: Reverse HTTP redirect with parameters

Pedro Assuncao — Fri, 28 Aug 2009 10:08:52 +0000

Here’s how to, from a view, redirect to another URL passing parameters to it. For instance, to redirect the user to a certain page after login:

return HttpResponseRedirect(reverse('dz_details', kwargs={'dz_id':dz.id}))

This will lookup the ‘dz_details’ name in your urls.py file, which could be defined like so:

url(r'^details/(\d+)/$', views.dz_details, name='dz_details'),

There is a simpler way to do it, if you don’t want to use the parameters’ names, but you have to know the order on which they appear in the method:

return HttpResponseRedirect(reverse('dz_details', args=[ dz.id ]))

[Django] Dynamic redirection after login

Pedro Assuncao — Mon, 17 Aug 2009 17:17:24 +0000

Here’s how to redirect to the login page in django, making it go to a certain view (by name) after a successful login:

login_url = '%s?next=%s' % (reverse('acct_login'), reverse('jumplog_index'))
return HttpResponseRedirect(login_url)

Notes:

acct_login is django’s view name for the login page (double check this, as it might change in future django versions);
jumplog_index is the target view where I want the user to go after successfully logging in.

UPDATE: An alternative, much more awesome, is to use the @login_required decorator on the view you want login to be required. Like so:

@login_required
def index(request):
   # Your code here

The decorator can be imported from this module:

from django.contrib.auth.decorators import login_required

I guess the first dirtier method can be used when you want to do something more custom, like sending some data to the login view, or something like that

Thanks for the tip, Steve