I have posted a couple of important updates to MongoSession (a PHP MongoDB Session Handler) that I’d like to share with you.
I have removed the secondary indexed id field which should speed up operations as well as reduce the storage requirements. This is most beneficial for larger sites supporting a high rate of concurrent users.
I recently added two boolean constants, FSYNC and SAFE. These two constants are documented on the MongoCollection::update() page and can be turned on to ensure your session data is actually written to the server. Please note that turning these variables on will cause a performance hit to the overall session handling.
Both the garbage collection and session write methods have been updated to perform atomic operations. This was done as a measure to reduce the risk of race conditions as mentioned on the project page.
A bug was fixed in the read method where an extra line of code existed for no reason.
A bug in the garbage collector was fixed. Update previously only updated a single result as opposed to the whole set of matches.
I’ve got a bone to pick with the Kohana development team. I just wasted a couple hours out of my day adding functionality to their core Validate class only to find that the functionality had already been added in a future revision. All I was looking to do was pass parameters to a custom callback function. I had even documented the entire change and was looking to post a feature request with an included patch. I later came to realize that the version I was using, 3.0.3, was outdated to the point of lacking this seemingly trivial functionality that was included in 3.0.7. So bare with me while I rant on why your framework is losing in the popularity contest. more »
The @font-face CSS tag allows web designers and developers to specify custom fonts to display text on their web pages. Allowing websites to utilize custom fonts removes the burden of requiring your website visitors to have the custom fonts installed on their personal computer. The following examples use the DejaVu Sans font, which is freely available for download here. more »
Back in late April, following Facebook’s f8 conference, a few articles began trickling out regarding possible security concerns with the new “Like” button. I had, at that time, unknowingly positioned myself as the potential originator of the term “likejacking.” In the comments section of How to “Like” Anything on the Web (Safely), I coined the term like-jacking; seeing a strong correlation between malicious usage of the button and clickjacking.
Little did I know that my term would be on the forefront of a media frenzy, where a plethora of articles would be posted in a matter of minutes regarding the subject. The frenzy can be attributed to the release of two articles from security experts at Sophos regarding the topic:
An entry on Likejacking has made its way to Wikipedia. I urge security experts and web developers alike to please moderate and update the entry as it is in dire need of an overhaul. This will aid in a speedy addition to Wikipedia as well as a thorough, well-documented resource for web users to gain insight on this new security threat.
There’s not a whole lot of PHP algorithms floating around. Perhaps people just rely upon the SPL. I had a curiosity to find a string sorting algorithm written in PHP and I didn’t find any non array-based solutions. The implementation I came up with this morning implements recursion, passing by reference, and string iteration. It’s worth noting that this is a horribly slow solution. For this reason I’ve included a very concise solution which is an order of magnitude faster as it uses a C implementation of quicksort which generally runs in O(nlogn). To make people aware of quicksort, I have also implemented a solution for string sorting. With that being said, this could be a fun exercise for students. Alternative solutions and runtime analysis are welcome in the comments section.
Version 1 (slow)
Let’s be honest, if I didn’t use a pre-existing algorithm to create a function, odds are it’s going to be slow. This implementation most closely resembles a version of Bubble Sort without comparing apples to apples.
/**
* A (slow) non-array based solution for sorting strings in PHP.
*
* @param &$s The string to be sorted
* @param $len The length of the string
* @param $curPos The current position being sorted (default = 0)
*/
function sortString(&$s, $len = 0, $curPos = 0) {
if ($curPos === $len) return;
$nextPos = $curPos + 1;
while ($nextPos < $len) {
if ($s{$nextPos} < $s{$curPos}) {
$tmp = $s{$curPos};
$s{$curPos} = $s{$nextPos};
$s{$nextPos} = $tmp;
}
++$nextPos;
}
sortString($s, $len, $curPos + 1);
}
// example usage prints:
// 1223344789aaaaadddefffffhhhiillllnnoorrsssuuuuwyy
$string = 'ouhlasfuywernhlasdfoulnarfiuyadf1234987234sdfailh';
sortString($string, strlen($string));
echo $string . "\n";
Version 2 – Quicksort Implementation
Although much faster than version 1, even a user defined quicksort implementation in PHP cannot compare to that of sort.
function sortString($s) {
$left = $right = '';
$l = strlen($s) - 1;
if ($l <= 0) return $s;
$pivot = floor($l/2);
do {
if ($l == $pivot) continue;
if ($s[$l] <= $s[$pivot]) $left .= $s[$l];
else $right .= $s[$l];
} while (--$l);
return sortString($left) . $s[$pivot] . sortString($right);
}
// example usage prints:
// 1223344789aaaaadddefffffhhhiillllnnoorrsssuuuuwyy
$string = 'ouhlasfuywernhlasdfoulnarfiuyadf1234987234sdfailh';
$string = sortString($string);
echo $string . "\n";
Version 3 – Fast and Concise
This version, although utilizing some fairly heavy functions like split and implode, still outperforms both previous solutions. This is due to the fact it utilizes a small subset of PHP core functions. The core PHP functions have heavily optimized implementations which are written directly in C, meaning it’s nearly impossible to get much faster except in small edge cases.
Let’s face it… nobody enjoys spam. A basic PHP contact form is generally susceptible to a massive amount of spam mail. SPF30 is a PHP library which utilizes a number of recommended spambot deterrents in an attempt to reduce form submission spam. SPF30 does not utilize any form of captcha. In addition to spam prevention methods, SPF30 also handles two-way encryption of form data. In other words, your form content cannot be easily sniffed across the wire. This adds a layer of security to your contact forms.
Features
The form submission contains a hashed value of a system defined secret key, the current date, and the user’s user agent.
The form submission is invalidated in the event the submission timestamp exceeds a specific timeout period (default 1 hour).
The form submission is invalidated in the event it was submitted in rapid succession (default 5 seconds).
A hidden input honeypot is utilized in an attempt to trick bots into passing data with the field.
A hidden hash field is validated against the submission time, user agent, and secret key.
A hidden field is sent containing a the array of encrypted fields for decryption to their old field names.
Decrypted form fields are written directly back to the POST array, abstracting the encryption from your backend handling.
User specified form field names can undergo two-way DES encryption to obfuscate form field names.
User submitted form field values can be encrypted on the frontend using a Javascript implementation of DES.
The encryption method goes beyond simple DES encryption for the purposes of transporting UTF-8 characters in POST data.
Version 0.2 of salid, a simple jquery validation plugin, has been officially released. The update includes a number of important bug fixes and optimizations. If you encounter any bugs please contact me as it benefits the community. Below is a list of the changes.
Fix to required validation function
The required validation function was not properly working with radios and checkboxes. The required function now tests whether the form elements are checked as opposed to checking for a non-existant value.
Added phone number validation
Since phone numbers are so commonly used on forms, I went ahead and added it to the list of default validation methods included with the plugin.
Fixed error display for checkboxes and radios
Due to the fact no two browsers handle checkbox and radio styling alike, the default error styling would not suffice. For starters, adding a class to the elements themselves would have no effect on their appearance as the border and background properties often cannot be overriden. For these reasons, you should now be wrapping your checkboxes and radio buttons in a
tag. The error handler will apply error styles to the parent label tag as opposed to the usual application directly on the element itself.
Fixed issue with duplicate form elements
In the event a user had multiple forms on a page that shared at least one common form element name, jQuery would only match the first element found on the page if found by ID selector. Although it is not semantically correct to have two form elements with the same ID, Salid allows for such cases to occur due to the fact numerous frameworks automatically generate ID tags based on element names. The fix involves only allowing selection from children and descendants of the form tag from which Salid was initialized.
An AJAX preview mode is now available in the options for a quick preview of the button.
All META tags are now populating correctly.
Fixed parameter naming issue with META tags (replaced “name” with “property”)
Additional options have been added to the configuration page; height, layout, show faces, and the ability to define your own custom class for the iframe.
Fixed the handling for populating the description meta tag. The code now attempts to get the excerpt first, then the first 255 characters of the post, and lastly the blog description as a fallback.
A few other minor issues were resolved which I found while implementing feature requests.
UPDATED 04.27.10 8:00 am
A number of issues with the admin options were found which would hinder you from modifying certain default settings. This has been resolved and the .tar.gz file has been updated accordingly. A couple of outstanding issues noted in the comments are being addressed. These issues include associated images not displaying on Facebook, no admin option for toggling between display methods, and the og:description meta tag not populating properly.
It is recommended that you add the following namespace parameters to your theme’s header.php HTML tag:
<!-- DEPRACATED. See code reference above. -->
<html xmlns:og="http://opengraphprotocol.org/schema/" xmlns:fb="http://www.facebook.com/2008/fbml">
I have dabbled with the Facebook Open Graph API for the past few days following f8 and eventually arrived at a full featured plugin for displaying Facebook’s new “Like” button functionality on your WordPress blog. The plugin is chock full of goodies, including:
Option of using either the standard or button_count (small) display mode.
Customizable width of the iframe containing the button.
The first included image of the post will be cropped to 50×50 and included in the user’s Facebook profile. (possibly broken)
Customizable placement of the button, which can be either at the top of your post, bottom of your post, or included using the php function me_likey_button(); in your theme files.
Ability to include the button within a post using the shortcode (BBCode) tag [like_button].
Ability to change the button’s wording to either “like” or “recommend”.
Customizable font family, including arial, lucida grande, segoe ui, tahoma, trebuchet ms, and verdana.
Ability to change the color scheme of the button to either “light” or “dark”.
Ability to enable/disable the button on a per post (or page) basis directly from the add/edit post screen.
Automatic inclusion of Open Graph meta tags to ensure your site is fully compliant with Facebook’s new handling.
The plugin will have an official page shortly. Suggestions and feature requests are appreciated as I want to ensure the plugin is useable by all. For more details regarding Open Graph and meta data requirements, please click here.
The getMatchingClass plugin is useful for determing whether a singular matching element contains a particular partial class name. The idea behind the function is to allow for partial matching using the following jQuery selectors:
$= – “ends with” selector
^= – “starts with” selector
*= – “contains” selector
= – “matches” selector
The plugin is useful in the event you need to determine if an element contains a portion of a class name utilizing already known jQuery selectors. Unfortunately this functionality did not exist within the Sizzle selector engine. Improvements and bugfixes are always welcome. more »
There are numerous sites for finding available .COM domain names. Generally, these sites will not turn up the kind of results you are actually looking for when trying to find the perfect domain for your startup, personal site, or business. With the following simple snippet of code you will be able to generate a list of all available two word combinations for a supplied list of keywords. It can be ran from the command line and can output a txt file of all matching domains. I hope you find this as useful as I have. more »
