Facebook comments has recently emerged as replaced for comment boxes on regular websites. There are many advantages - such as spam control, posting comments as yahoo, Facebook or other users, ability to post on Facebook profile's wall etc.

We started using the Facebook comments on most of our community based websites. But we always wanted to find a way to get the SEO benefits from comments posted on website.

Recently i came across a report at xssed.com, about ebharatjobs.com being vulnerable to cross site scripting (XSS). That report was filed way back in 2009, and we have already fixed that when we upgraded www.ebharatjobs.com in 2010.

I wanted to share the code we used to check the input from users to possibly prevent XSS vulnerability.


// cross-site-scripting (XSS) vulnerability
$q = $_GET["q"];
//remove the script tag and its contents

Yesterday, we showed how to Bulk update node url aliases. Now we faced similar problems while generating bulk url aliases for taxonomy terms. For some reason the admin interface was not able to generate them. The screen just struct up.


define('DRUPAL_ROOT', getcwd());
require_once DRUPAL_ROOT . '/includes/bootstrap.inc';
drupal_bootstrap(DRUPAL_BOOTSTRAP_FULL);
// menu_execute_active_handler();
set_time_limit(0);
error_reporting(E_ALL);
$tid = (int) variable_get("url_alias_last_tid",0);

eBharatBase.com has more than 2, 60,000 nodes. We changed the pattern of url aliases to be generated for each node. We tried to generate from admin interface of pathauto module, the system hangs everytime.

The reason - query to fetch node was taking much longer to execute, and php/drupal used to show error message since they were not receiving any response.

To tackle this problem, i started writing small but effect piece of code to generate the url alises of all 2,60,000+ nodes.

1) Created script named "generate_url_alias.php" in drupal root folder

SearchPropertyIndia.com is a vertical search engine targeted to the people looking for properties such as Flat / Apartment, Bungalows, Villas or Land, plots. Visitors can also search for properties available for rent only.

Searching property online can be a tiresome process, and at times you will not be able to find the right kind of property you might be looking for. To make this process easier we have come up with this vertical search engine.

SearchPropertyIndia.com crawls real estate websites, and indexes them. And then people can search them based on city, type, etc.

We are very happy to announce the mobile version of eBharatJobs.com - India's most used job search engine.

The mobile version is located at http://www.ebharatjobs.com/m

Redirection from main page is handled through a php script that detects if the referring agent is Desktop or a mobile/handheld device. The mobile devices are shown this new mobile version of ebharatjobs.

The website is live and we are now waiting for feedback from our users.

Thank you
Vinay Yadav
Web Developer

eBharatJobs.com is one of India's most used job search engine. We continuously keep working on improving the user experience.

Recently we have been working heavily on optimizing the load speed of ebharatjobs. Google has announced that they are now considering page speed for ranking websites, so it becomes very important for websites to improve their loading time.

We first introduced APC for caching the files in memory. It helped a lot, since were were using few files and caching them in memory helped much more.

We have just released a new drupal module to index and search attached images of nodes.

An ApacheSolr helper module to index, query and retrieve attached images' file paths and ids of content nodes.

Benefits:
- You need not run resource intensive node_load() to get list of attached images for each items in results.
- Image paths / ids are available in apachesolr results array.
- Image path can be passed to imagecache preset

And this can be easily used in search results and other search screens such as themed "more like this" pages.

Dependencies:

IndCareer.com has been built using Drupal 6.x and presently has more than 31, 000 nodes, large number of taxonomy terms (categories).

With this increase in data, we needed an efficient search system so that users can find information easily and afficiently. To provide precise search results to visitors we decided to use ApacheSolr search system for our website.

With new search system, data sorting and further filtering of search results are possible.

For example, college search results can be further filtered by

• location (city, state)
• Category

MySarkariNaukri.com is a complete website for Indian Government Jobs. Along with jobs, we also list Results and Exam notifications related to Government institutions.

MySarkariNaukri.com was launched in March 2009, and we have successfully completed 1 year. We have been using Drupal's core search initially, then shifted to Google Custom Search - and it was serving our purpose perfectly.