The 3-Minute Must Watch Video


This video, narrated by astronomer Carl Sagan, definitely puts into perspective how tiny we are in relation to the whole universe. However, it also reminds me that despite the enormity and vastness of space, Earth is the only home we have; as a result, we’d better take care of it and the people in it.

Too often, the human race gets so consumed with power and being “right” and dirty politics that we forget about our ultimate obligation – to take care of each other by acting wisely, conservatively and responsibly. It’s an obligation that if neglected will be our downfall and if fulfilled will be our legacy.

So it is must watch video. Spend your 3 minutes and that will make any problem you have seem astonishingly trivial :)

A summer full of code


Four months of constant coding and excitement comes to an end – I still can’t believe I am saying this. While I am really really happy with how these times went – how fast, how satisfying, how perfectly awesome it went. It is just still hard to believe that it has ended – time flies, literally!

Here goes my wrap-up blogpost for all these months of tireless programming. Here’s to a summer full of code! :-)

Background

As I have mentioned in my last few blog posts on several updates, I have been working on the Language Coverage Matrix Dashboard project for the Language Engineering Team of Wikimedia Foundation as my Google Summer of Code project. This web-based application automates all the information that are available on the Language related topics that Wikipedia and its Sister project covers – tools they support, languages they have support for, statistics that are available for public use etc. The target users are obviously the researchers, linguists, i18 developers, and even MediaWiki developers concerned about the Language technology, willing to implement somewhere else, you know, for good!

That said, I am going to post here a detailed description of the final product that we have achieved so far.

Demo Instance: http://tools.wmflabs.org/lcm-dashboard/lcmd/

The User Interface

We have actually done a lot of thinking around the user interface throughout the period of GSoC. The final one turned out be quite satisfying. The whole page is now divided into two major parts – a) where all the queries that are available would be displayed. b) the results to show. So, now it has become same for the Language based queries, as well as the tools based ones. Filter is there for the tools based queries.

The diagram of the UI is uploaded on commons. Check it out!

Relevant Code: On GitHub!

LCMD – Landing Page

Basic Functionalities

Now that you’ve seen the UI that we have aimed to achieve. Let’s go deeper into the functionalities.

  • Search

    The whole searching concept has been divided into two parts, two ways, to be specific.

    • Language Based Search: Like you can see from the design and the interface that has been implemented, a search bar for the Language search is there. Essentially you get to search for all the Languages that are on the database of ours, i.e. the ones that are supported by Wikimedia Projects. jQuery.ime – a development work by the Language Engineering developers has been implemented on the search. Language search have features like : Suggestions, Auto-Completion etc, to provide users with more efficient search. Output of the search for a particular language would obviously give all the relevant details that we have for that particular language.

Language Search with autocomplete facility

Language Search with autocomplete facility

Language Search Result

  • Tools Based Search: This is important. This basically a feature to play around all the language data we have. I have got a filter for that – a JS application that filters out all the language data with respect to tools based queries. So, essentially this outputs the languages that support particular tools. Tools based queries work around three basic queries -
    • Input Method
    • Webfonts
    • Internationalization Library
  • Filter: 

    As mentioned above the tools based query part does have this feature where you can filter out all the language data with respect to the tools based query. A JavaScript code runs behind to have this running. Outputs the language names that support the particular tools that may be selected on the radio button

  • API Console

    This is probably my favourite feature from LCMD.  If a third party user wants to fetch these language related data then they can easily get data by calling these APIs. For better understanding of API I have developed API console.
    This is similar to MediaWiki API sandbox. In which you just have to select from dropdown and then press on request button and then “BINGO”. Yes you will find

Language API Console

That is almost what it is. Now a quick overview on how I have developed stuff over time, things I have just described above, and the processes behind. This would also include a few stuff that we changed afterwards. Some of the updates you might have followed on my previous blogposts, this one would work as a summary of those.

June – Community Bonding Period:

  • In the community bonding period I first thought/planned about new technologies which I would be using in my project.
  • Started with the database – created the database schema.
  • All the data was in a spreadsheet –  created one python script so that it could transfer all the data from spreadsheet to database automatically.
  • The week after, I started working on new language entry system, database connection and completed it.
  • Implemented language search system with suggestion for the language names.
  • Implemented the feature to get the language details.

July – Coding out the Functionalities, start!

  • Created and set up primary thing on Wikimedia Labs.
  • Done minor changes in Language Search System.
  • Created on the spot editing facility for any language detail under admin privileges
  • Created Filter Facility
  • Database schema changed
  • jquery based new alternative REST architecture created
  • Language to Font mapping
  • Language to Input method mapping
  • Redefining search implementation

August – Coding, coding, coding!

  • Created login system for admin + session management
  • Designed new UI on langfilter.php page
  • Developed PI visualization chart
  • Developed API for language detail
  • Developed example for API usage
  • Created preview system for new language entry system as well as direct editing feature

September – the last touch.

  • Developed API console for 3rd party user to use these data for their website.
  • Redesigned entire NEW UI as per proposed scheme
  • Small correction in functionality
  • Solved a few bugs

Documentations:

All the things that I have documented so far in these four months, here goes!

A note to thank you all, who were there

I loved living this four months of my life. But, Google Summer of Code is not why I love Wikimedia Foundation. Rather, it is just another event that happened between the Foundation and myself. I have been an active volunteer to Wikimedia projects for a long time now. And, yes, I have been thankful for each and every day I have spent with MediaWiki, Wikipedia and the people around all those projects. GSoC has been a different kind of excitement – recognition feels good. Honestly, it does. But what feels even better than that is to get a merged status on some hundred lines of code that I have written. GSoC has given me many merges – I am thankful for each of them. Runa and Alolita have been amazing mentors to me. For all the things I have done good, for all the pushes I needed to do them that way – they have been there always. Special thanks to Quim, Sumana and Sucheta to help me throughout. I am sure, I am motivated enough to live for another four decades just to code. JUST for the sake of coding. Yes. GSoC definitely played a great part on achieving that motivation.


Here’s to Google Summer of Code!  :) :)

Re – Designing LCMD – Introducing the New User Interface


I have been spending the last few days re-designing the User Interface of Language Coverage Matrix Dashboard. Although the last one was working smoothly, there was a need of more polished and efficient one than the previous. I met up with my mentors and the designer of our Language Engineering team a number of times to decide on the new UI.

So, let’s walk through the newly added User Interface of LCMD!

Now to begin with, I would like to mention what exactly demanded a new UI here. The issue with the previous one was as simple as: It demanded to show more details and also, the informations needed to be grouped.

About the new UI:

We have a few new features and a few useful changes here in the new UI. Which are basically listed as,

  • Data grouped in Universal Language Selector, Project Milkshake, Legacy Extension and More Info of that language.

Screen Shot 2013-07-28 at 3.56.33 AM

The new options

  • This grouped data is shown in responsive and smooth drop down.

Screen Shot 2013-07-28 at 3.57.06 AM

Dropped Down version of the options

  • New switch button is introduced for more efficient editing by the admin.

Screen Shot 2013-07-28 at 3.57.47 AM

Switch buttons

  • New modal design created to update the information of languages.
  • Icon cross and tick are added for showing language support.

Screen Shot 2013-07-28 at 3.56.46 AM

The cross and tick!

This is all about the new UI that we have right now! :)

How to implement system command using python


In this post I will show you that how you can exactly implemented system command using python.Here we use for example ls -l command. When I use this command in terminal it gives me following output

Harshs-MacBook-Pro:scripts harshkothari$ ls -l
total 32
-rw-r--r--  1 harshkothari  staff    13 Jul 23 02:33 demo.ini
-rw-r--r--@ 1 harshkothari  staff  1864 Jul 23 04:30 filesread.py
-rw-r--r--  1 harshkothari  staff  4121 Jul 23 04:30 langtofontmap.json

Now this command we can implement by following methods using python :)

  • os.system("command with args") passes the command and arguments to your system’s shell. This is nice because you can actually run multiple commands at once in this manner and set up pipes and input/output redirection.
    os.system("command < input_file | anothercommand > output_file")
    However, while this is convenient, you have to manually handle the escaping of shell characters such as spaces, etc. On the other hand, this also lets you run commands which are simply shell commands and not actually external programs. http://docs.python.org/lib/os-process.html

    >>> import os
    >>> os.system('ls -l')
    total 32
    -rw-r--r--  1 harshkothari  staff    13 Jul 23 02:33 demo.ini
    -rw-r--r--@ 1 harshkothari  staff  1864 Jul 23 04:30 filesread.py
    -rw-r--r--  1 harshkothari  staff  4121 Jul 23 04:30 langtofontmap.json
    0
  • stream = os.popen("command with args") will do the same thing as os.systemexcept that it gives you a file-like object that you can use to access standard input/output for that process. There are 3 other variants of popen that all handle the i/o slightly differently. If you pass everything as a string, then your command is passed to the shell; if you pass them as a list then you don’t need to worry about escaping anything. http://docs.python.org/lib/os-newstreams.html
    >>> import os
    >>> print os.popen('ls -l').read()
    total 32
    -rw-r--r--  1 harshkothari  staff    13 Jul 23 02:33 demo.ini
    -rw-r--r--@ 1 harshkothari  staff  1864 Jul 23 04:30 filesread.py
    -rw-r--r--  1 harshkothari  staff  4121 Jul 23 04:30 langtofontmap.json
  • The Popen class of the subprocess module. This is intended as a replacement for os.popenbut has the downside of being slightly more complicated by virtue of being so comprehensive. For example, you’d say
    print Popen("ls -l", stdout=PIPE, shell=True).stdout.read()
    instead of
    print os.popen("ls -l").read()
    but it is nice to have all of the options there in one unified class instead of 4 different popen functions. http://docs.python.org/lib/node528.html

    >>> from subprocess import Popen, PIPE
    >>> print Popen("ls -l", stdout=PIPE, shell=True).stdout.read()
    total 32
    -rw-r--r--  1 harshkothari  staff    13 Jul 23 02:33 demo.ini
    -rw-r--r--@ 1 harshkothari  staff  1864 Jul 23 04:30 filesread.py
    -rw-r--r--  1 harshkothari  staff  4121 Jul 23 04:30 langtofontmap.json
    
  • The call function from the subprocess module. This is basically just like the Popen class and takes all of the same arguments, but it simply wait until the command completes and gives you the return code. For example:
    call(["ls", "-l"], shell=True) http://docs.python.org/lib/node529.html

    >>> from subprocess import call
    >>> call(["ls", "-l"], shell=True)
    demo.ini                filesread.py            langtofontmap.json
    0
  • Using command module http://docs.python.org/2/library/commands.html
    >>> import commands
    >>> commands.getstatusoutput('ls -l')
    (0, 'total 32\n-rw-r--r--  1 harshkothari  staff    13 Jul 23 02:33 demo.ini\n-rw-r--r--@ 1 harshkothari  staff  1864 Jul 23 04:30 filesread.py\n-rw-r--r--  1 harshkothari  staff  4121 Jul 23 04:30 langtofontmap.json')
    >>> commands.getoutput('ls -l')
    'total 32\n-rw-r--r--  1 harshkothari  staff    13 Jul 23 02:33 demo.ini\n-rw-r--r--@ 1 harshkothari  staff  1864 Jul 23 04:30 filesread.py\n-rw-r--r--  1 harshkothari  staff  4121 Jul 23 04:30 langtofontmap.json'
    
  • The os module also has all of the fork/exec/spawn functions that you’d have in a C program

Hope this will help you :)

Pass Random variable to shell script rather then $1 or $2 using php


Here I am showing you how can you pass the random variable to shell script. ( i.e like $uname :) )

You can receive data using $1 or $2 variable but you can’t directly receive data into $uname or any random variable.
Here is my bash file demo.sh

#!/bin/bash
echo $uname

Now I have to pass $uname variable so directly using this php function you cant do this

exec("sh demo.sh harshkothari",$output);

You have to send it via environment variable . This is simple code snippet how you can you pass this. php file name is shell.php

$uname = 'harshkothari';
putenv('uname='. $uname); //put environment variable uname 
exec("sh demo.sh",$output); //execute command
echo($output[0]. "\n"); //print return value

$output will return the last line of output and it will show like this

Harshs-MacBook-Pro:lcmapiuse harshkothari$ php shell.php 
harshkothari
Harshs-MacBook-Pro:lcmapiuse harshkothari$ 

Here is snapshot of terminal output

Output of php script, passing random variable to shell script

Output of php script, passing random variable to shell script

Wikimedia Tool setup for LCM-Dashboard (GSoC Project)


Wikimedia tools login screen using ssh

Wikimedia tools login screen using ssh

wikimedia tool setup is very tedious work but #wikimedia-labs helps a lot.

After creating an account and being approved by admin. You will log in to your tool using this command.

ssh -A username@tools-login.wmflabs.org

After logging into your tool you have to use following command to switch to your service

become toolname

Now this have following things

  • public_html – where you can place your html, php, css, js

  • cgi-bin – where you can place your python script etc

  • replica.my.cnf – contains username and password

Now Creation of databse for your project

In the first step you have to find you username and password by using “” file.

After that your database name must be USERNAME__DBNAME (2 underscore)

mysql --defaults-file="${HOME}"/replica.my.cnf -h tools-db 

after that MariaDB console will open and in that you can create databasw

MariaDB [(none)]> CREATE DATABASE USERNAME__DBNAME;

you can access this database by using this command

mysql --defaults-file="${HOME}"/replica.my.cnf -h tools-db USERNAME__DBBAME

If you want to connect the database to your project

host = tools-db
uname = your_username
pass = your_password
dbname = USERNAME__DBNAME

Then you can push the file to remote using git via clone or you can use scp command to push files from remote to local like this

scp <files> USERNAME@tools-login.wmflabs.org:/data/project/<tool-name>

Hope it will help you :)

Link of my tools on wikimedia labs : lcm-dashboard

Being Google Summer of Code Intern


I am very excited (and also a little late :D) to share the news that I have been selected as an intern for Google summer of code in Wikimedia Foundation for the project of developing language coverage matrics dashboard for language engineering team of Wikimedia Foundation.

I would like to thank my mentors Runa Bhattacharjee and Alolita Sharma, for guiding me throughout and very special thanks to Amir Aharohi for his valuable inputs for this proposal. Last but not the least special thanks to Sumana and Quim for polishing up my proposal, and the feedbacks all this while.

Being selected as a GSoC intern was finally a good news after quite while. The result came at 12:30 midnight – I couldn’t sleep. Nevertheless, great night it was and so was the news.

Community bonding period started right after. I tinkered around a few new technologies and did the overall planning for my project.

Until June 17

Until June 24

  • Created new language entry system

    Search + New language button

    Search + New language button

    Language Entry form

    Language Entry form

     

  • Created language search system with suggestion

Language search with suggestion facility

Language search with suggestion facility

Language Search result

Language Search result

 

Until July 1

  • created and set up primary thing on wikimedia labs

  • jquery.ime implemented in project

  • done minor changes in language search system

  • created on the spot editing facility for any language detail under admin privileges

Edit button for on the spot editing

Edit button for on the spot editing

when click on edit button this menu will appear

when click on edit button this menu will appear

After saving the changes it will change

After saving the changes it will change

Until July 8

  • set up basic filtering functionality

  • set up some test cases to find smooth and fast filter facility

  • created simple idea of API for language data

Until July 15

  • Created Filtering functionality : By this time, a filtering functionality seemed to be an important feature to have on the LCMD interface. Filtering simply means that there would be a list of the language tools on one side along with checkboxes and the checking those would output the list of the languages that have those tools enabled. So, if say suppose, we have both jquery.ime and jquery.webfonts checked in, the language output will give a list of the common languages with both the tools enabled. This was kind of important. Screen shot and more detail on another blog post.

  • Set up tools instance for project tool instance

  • Set up new repo for project with lots of commits New repo

  • Created simple php API for language data :  Here I have developed a simple prototype version of Language detail APIs so that any website / user want to use that information that can fetch the details using that API

Return dataType is : JSON and if you are using cross domain data transfer then it should be JSONP.


$.ajax({
    url : 'http://tools.wmflabs.org/lcm-dashboard/lcmd/api/php/dataapi.php?query=jquery_ime',
    dataType : 'JSONP',
    type : 'GET',
    //data : 'query=webfonts jquery_ime' ,
    success : function(data){
        console.log(data);
    },
    error : function(data){
        console.log(data);
    })

It will return the data in JSON format.

Beautiful example of this API usage is here

http://mikipedian.blogspot.in/2013/07/i-see-rather-what-all-i-now-visualize.html