Tux-ian

This is Aravindh and Software Development is my profession.I started this blog with the intention to post small snippets, tutorials and other useful things regarding Python, Web, Open source and Linux. Feel free to reach me at : [email protected]

[aravindhramu]

Dusting my blog with Bluetooth API !

Hi Folks !

Its been quite a long time since I wrote a blog entry ( few months perhaps ?! ). So refreshing it again with something I wanted to write for quite some time now.

Not many good tutorials on developing a bluetooth application. And It doesn't get any easier with our Python. I am not gonna brief here about the Bluetooth protocol as such. I believe users are already familiar with it.

If not, nothing to worry. Its quite similar to our traditional Internet programming. If you need to brush on Internet programming basics, Beej's guide is the best place to start !

Quite a few dependencies here. I am using CentOS. You can use your distro's package installer to install the same packages.

yum install bluez bluez-libs bluez-libs-devel

Bluez is the official bluetooth protocol stack of GNU/Linux.

Apart from that you also need to install python wrappers for Bluez. wget http://pybluez.googlecode.com/files/PyBluez-0.18.tar.gz tar -xvzf PyBluez-0.18.tar.gz cd PyBluez-0.18 python setup.py install

Your Bluetooth app can use an underlying RFCOMM/L2CAP protocols. They are similar to TCP/UDP protocols. However L2CAP also has a 'reliable' version that helps to retransmit infinitely. If this infinite retransmit option is suppressed ( for some real-time applications), its analogous to our UDP.

First lets write a simple code to discover devices around you. As usual, you need to ensure all devices are in "discoverable" mode.

https://gist.github.com/5211281

Here, the discover_devices() method returns list of bluetooth addresses. Bluetooth addresses are 48-bit addresses ( similar to MAC ). However, unlike MAC (used only upto data-link), they are used upto the application layer. This address uniquely identifies a device.

We need to perform a "lookup" ( similar to DNS ) to convert this 48-bit address to a human-readable format ( This is the name that we give to our device , say "My Phone" )

So , in my situation above snippet prints :

device address: 08:ED:B9:9F:EF:46 Name: FRIEND-PC device address: 38:16:D1:1F:90:C3 Name: Aravindh phone

Usually a bluetooth application involves both writing a server and client; However I am gonna focus only on writing a client for some well defined services available in most mobiles supporting bluetooth ( say OBEX File transfer service).

Now lets write a script to list out all the services supported by the device.Once we get to know that , we can query the required port.

https://gist.github.com/5211695

The above code snippet returns the port on which the OBEX file transfer service runs.

Once we have obtained the port, we can send a simple file to the device. For this we require the PyOBEX wrapper.

https://gist.github.com/5211721

The above snippet sends a simple text file to the device which will be save in the device in the name - 'date.txt' with contents as 'todays date'.

Its as simple as that. Hope you find it useful !

Cheers !! :-)

Resources :

1) http://people.csail.mit.edu/albert/bluez-intro/ 2) python "help" feature ;-)

[aravindhramu]

Writing a simple in-memory cache with redis

Writing a blog post on a rainy holiday is kindda awesome !! :D

Came across redis a few days back. The API it exposes is simply amazing. Decided to write a simple ( in fact,very simple. redis helps to build extremely complex and scalable architectures with ease) application with it using its python library.

1) Install redis from here.

2) Install python library for redis with easy_install redis

3) Given code-snippet implements an in-memory cache using redis’ string and list data structures.

https://gist.github.com/3999839

4) The above class has a cache class with four methods: add,remove,peek and take. So. why not implement this with just python ? You can bro, if you can guarantee high performance( redis is extremely fast ! Check it here ),atomicity and thread-safety. For the rest of the lazy folks ( like me ! ), use redis ! :D

The expire function is used to invalidate the cache after the specified timeout. Rest is self explanatory. This is just a sample. You can also implement a LRU cache by using redis' sorted set data structure. Or you can use your own invalidation schemes.

Hope this article was helpful !

References: 1) http://redis.io/ 2) http://www.quora.com/Redis-vs-Memcached-which-one-should-I-use-for-a-web-based-application

[kofif-blog]

I’ve finally got my self to do some C++ work.

I haven’t touched the language for over 8 year, which is a real pity, as I always feel so much more stimulated when I work in it.

So what have I done?

I have stared working on a big project I hope to release some time in the distant future (no, it...

[aravindhramu]

Reblogging from Kofif's blog... :)

[aravindhramu]

Writing a crawler in less than 10 minutes

Hola !

Came across Scrapy a few days ago. It was quite simple and easy to get started with a simple crawler.

This post assumes you have read something about scrapy and explains the steps to create a simple application using scrapy.

1) Install scrapy

apt-get install scrapy

2) Start a project with scrapy:

scrapy startproject myscraper

3) Define Items to be scraped in items.py:

Inside myscraper/myscraper/items.py , you will find a template to define what items need to be scraped. If you are familiar with Django, this is like models except that there are no data types. All Items are Fields()

https://gist.github.com/3563409

4) Define allowed domains/start urls :

Once you have defined items, open a new file ,say myspider.py in myscraper/myscraper/spiders/. Subclass from BaseSpider ( or from CrawlSpider if you want to specify which links to allow/deny). Provide a crawler name and specify a start_url for the crawler.

https://gist.github.com/3563319

Specify what to extract from the urls in the parse method. Scrapy uses Xpath by default to select contents. You can read more about Xpath from here. Parse method accepts the http response object and returns an Item. In the above example, for bookname , I have assigned contents of title tag and contents of span tag with price attribute to ‘price’

5) Define Pipelines:

Its important to store the scraped data somewhere. For this purpose, scrapy provides pipelines. Define your own pipelines in pipelines.py and enable that class in ITEM_PIPELINE settings in settings.py.

For eg, to write to a json file:

https://gist.github.com/3563350

and in settings.py file :

ITEM_PIPELINES = ['myscraper.pipelines.MyScraperPipeline']

6) Run your crawler

Thats it ! You’re done. Run your crawler using:

scrapy crawl avispider

You can see the scraped items in results.out file. :) Hope this was useful. You can see a full fledged application using scrapy here

[aravindhramu]

Autocompletion with Trie

Autocompletion is one of the key-features that you could observe in any web based product. It enhances the user experience of the product by providing relevant suggestions as the user types in. One good example is Google suggestions.

Could’nt find much tuts on this topic ( except a few good ones that i have mentioned in the resources section at the bottom. )

A trie is a data structure very suitable for applications that involve working with words having common prefixes (ape,apple, app etc) . More precise description about trie can be found here and here.

The root is the start node and is empty . Each entry beginning with distinct alphabet branches out as a child from this root. (See figure, Source: topcoder community). Several implementations of tries exist. I am going to use a map based trie for implementing autocompletion feature.

https://gist.github.com/3393535

In the code, addToTrie is used to add a new string to the trie. It checks if the first character of the string is there in the map. If not , a new entry is created. If it is already present, its called recursively with the updated value of string ( leaving first character )

The most important part is the searchForString method. It takes the prefix as the input and prints matching results. First we traverse the trie till the matching prefix and from that point onwards, we need to print all strings having the same prefix. So we perform depth first traversal from that point. For eg: if the input is “hell” and we have “hella”,”hellb”,”hellc” in the trie, once we reach “hell”, we iterate through all the entries in map of “hell” ( map of “hell” will contain keys [a,b,c] ). Then we call the function recursively for each value of key and print results.

Thanks ! Comments/ Suggestions welcome :)

Resources :

1) http://v1v3kn.tumblr.com/post/18238156967/roll-your-own-autocomplete-solution-using-tries 2) http://igoro.com/archive/efficient-auto-complete-with-a-ternary-search-tree/ 3) http://community.topcoder.com/tc?module=Static&d1=tutorials&d2=usingTries

[aravindhramu]

"The positive thinker sees the invisible, feels the intangible and acheives the impossible ! "

[aravindhramu]

Adding a code snippet to Tumblr - an easy way !

Frankly speaking, writing a first post in a first blog really is a herculean task ! Atleast for me, I didn't know where to begin. Since I had decided that nature of my posts would involve code snippets, needed a way (desperately) to provide that 'syntax highlighting' feature to my posts. After some googling, found a couple of methods including ['prettify'](http://code.google.com/p/google-code-prettify). The other method I found easier ( probably cooler ?!) is from [github gists](https://gist.github.com/). Doesn't take more than 2 minutes. :) So what do you need to go ahead ? - A [Github](https://github.com/) Account ( If you dont have one ,just register for free and get one ) Ok. So here we go. 1) Goto your [Github gists](https://gist.github.com/) profile and select 'New Gist'  2) Select your Programming Language and start typing your code in the box provided.  3) Save your gist by clicking on "Create Public Gist" button and note down its URL . The URL can also be found in 'My Gists' as shown below :  4) Now edit your blog theme ( customize ) by clicking on 'Edit HTML' . Add the following three lines just before </head> tag and save the changes.

https://gist.github.com/3372242

5) Now, in your post where you want the code snippet, add the url (from Step 3) within div tags. for eg:

url here

So my sample Python code will appear as follows :

https://gist.github.com/3372213

Thats it ! Have fun folks ! :) Comments/Suggestions welcome !