I can’t really make my point here until you open the image below in a separate tab or something and view it in full size. This plot has been generated using earthquake data for the past hundred years from earthquake.usgs.gov and it was plotted on a world map representation using MATLAB. My question is posed in full glory beneath this image.

Tectonic Plates plotted with data from USGS using MATLAB

Seismic Activity plotted with data from USGS using MATLAB

All the dots on this image are the epicentres of a recorded earthquake as per data from USGS. The more yellow a dot is, the more severe the earthquake that was recorded, so basically higher the Richter magnitude. Blue dots are relatively low magnitude ones.

From the image you can make out the African plate in the middle, the American plates to the left of it[you can see the California fault to the very left of the North American plate], and at the top-right corner of the image is Japan. So now you can make out some serious fault lines represented here as dense yellow dots.

My question is, Why are all the fault lines running from north to south, i.e., vertical, and not more like parallel to the equator?? Has this anything to do with the direction of earth’s rotation??

Please leave your answers in the comments section below.

Walking the DOM in JavaScript has been covered well and good. But I couldn’t find substantial help when it came to walking the DOM tree of HTML files in PHP. It seems all there is to DOM is limited to XML. It would be wonderful if we could keep aside all tags and the related regex mumbo-jumbo to parse tags away, and instead check the values, text contained, IDs of elements, their class names and style rules. The function presented here walks through all the elements presented in a HTML file accessing all the node attributes and node values.

Anyways, being the wonderful thing that DOM is, I cooked up a small little function to walk the tree of an HTML file in PHP and print the output to screen, and boy does it walk. Anyway, less talk and more …
<?php
function walkDom($node, $level = 0)
{
$indent = ”;
for ($i = 0; $i < $level; $i++)
$indent .= ‘&nbsp;&nbsp;’; //prettifying the output
if($node->nodeType != XML_TEXT_NODE)
{
echo $indent.'<b>’.$node->nodeName.'</b>’;
if( $node->nodeType == XML_ELEMENT_NODE )
{
$attributes = $node->attributes; // get all the attributes(eg: id, class …)
foreach($attributes as $attribute)
{
echo ‘, ‘.$attribute->name.’=’.$attribute->value;
// $attribute->name is usually one of these:
// src, type, rel, link, name, value, href, onclick,
// id, class, style, title
// You can add your custom handlers depending on the Attribute.
}
//if( strlen(trim($node->childNodes->item(0)->nodeValue)) > 0 && count($cNodes) == 1 )
//echo ‘<br>’.$indent.'(contains=’.$node->childNodes->item(0)->nodeValue.’)’; // do this to print the contents of a node, which maybe the link text, contents of div and so on.
}
echo ‘<br><br>’;
}
$cNodes = $node->childNodes;
if (count($cNodes) > 0)
{
$level++ ; // go one level deeper
foreach($cNodes as $cNode)
walkDom($cNode, $level); //so this is recursion my professor kept talkin’ about
$level = $level – 1; // come a level up, and had to do it this way or else wordpress would take away one dash. 😦
}
}
?>

Is that good?? Because here is how you use it:
<?php
$doc = new DOMDocument();
@$doc->loadHTMLFile(‘http://www.google.com&#8217;);
walkDom($doc);
?>

And this prints away the entire DOM of the read in file specified by the URL to loadHTMLFile. More information about the used constants and functions can be found here. And believe me, this works.


You have a search page and as the results end you show a navigation bar with links to the other pages of the search. So boring. Seriously. This is the Web, and I am still turning pages.

And I think I am Pied Piper who will save you from this hell. The uber-cool solution is, detect as soon as a user is about to reach the end of the page and get new content[the second page of the search], just in time. Setting out, we need:

The height of the current document. So to this end, we will place a div with a decided upon ID at the end of the page. And all the content in the body will be loaded in another div just above this last div. We are detecting the document height by detecting where this ending div is located, pixel-wise, with this:

pageend = document.getElementById(‘laststop’).offsetTop;

So clearly, the ID of the last div is ‘laststop’. Next we need the screen resolution of the client window to find out how much of the document is visible right now. We do:

winheight = (typeof window.innerHeight != ‘undefined’ ? window.innerHeight : document.body.offsetHeight);

Now having this much, we need to detect when does the user scroll the page to the bottom. The following lines will give exactly how many pixels from the top have you scrolled to. Note that, this is the ‘height’ from the top of the document to the top of the visible area. The real end of the document is obtained by adding this ‘height’ and the window height retrieved earlier.

if(navigator.appName == “Microsoft Internet Explorer”)
alert(document.body.scrollTop);
else
alert(window.pageYOffset);

But these lines need to be run every few seconds to check as soon as as the user reaches the end of the current page. So we will do a setInterval() in the onload event of the body to do so. And keep the interval time to something a little over 2 seconds and nothing less than that, you of course don’t want your user’s browser to hang on them. It is annoying.

Said and done, this is what you end up with, demofile.php, annotated with explanatory comments:

<html>
<head>
<title>Page End Reached Detection Demo</title>
<script type=”text/javascript”>
var oRequest;
var pageend, winheight;
var howoften = 2000;
// this value implies that function howlow() will keep checking every
// 2000 milliseconds(2 seconds) to find out how low is the user on the page

var bufferHeight = 300;
// adjust bufferHeight, to take action as soon as the user is within these
// many pixels of the end of the page, so a larger buffer makes sure the user doesn’t
// see the dynamic page content loading. In this example, the script will fetch content
// as soon as you are within 200 pixels of the end of the page

// ooooo, ajax. ooooooo …
if(window.XMLHttpRequest)
oRequest = new XMLHttpRequest();
else if(window.ActiveXObject)
oRequest = new ActiveXObject(“Microsoft.XMLHTTP”);

function sendAReq()
// a generic function to send away any data to the server
// specifically ‘morecontent.php’ in this case, to get more content from the server
// all content fetched will be handled by showcontent()
{
oRequest.open(“POST”, “morecontent.php”, true);
oRequest.setRequestHeader(“Content-Type”, “application/x-www-form-urlencoded”);
oRequest.onreadystatechange = showcontent;
oRequest.send(null);
}

function showcontent()
{
if(oRequest.readyState == 4)
{
// now all the data that comes adds to the end of the page, specifically
// to the container div referenced with the ID ‘bodycontent’
if(oRequest.status == 200)
document.getElementById(‘bodycontent’).innerHTML = document.getElementById(‘bodycontent’).innerHTML + oRequest.responseText;
else
alert(‘No Response from Server. Try Again Later.’);
}
}

function calcParams()
{
// calculate the end coordinates of the page and the visibile height of a page
// at any time on the client’s desktop
pageend = document.getElementById(‘laststop’).offsetTop;
winheight = (typeof window.innerHeight != ‘undefined’ ? window.innerHeight : document.body.offsetHeight);
}

function isitend(curpos)
{
calcParams();
if((curpos + winheight + bufferHeight) > pageend)
{
alert(‘you are reaching the end of the page;’);
sendAReq();
}
else
{
alert(‘keep going’);
}
}

function howlow()
{
if(navigator.appName == “Microsoft Internet Explorer”)
isitend(document.body.scrollTop);
else
isitend(window.pageYOffset);
}
</script>
</head>

<body onload=”javascript:setInterval(‘howlow()’,howoften);”>
<div id=’bodycontent’>
<?php
for($index=0; $index<200; $index++)
echo ‘Line:’.$index.'<br>’;
?>
</div>
<div id=’laststop’/>
</body>
</html>

Oh and yeah, this sends a request using the XMLHttp object so we need another php file at the server to catch the request and send in additional content. This additional content will be appended to the div with ID ‘bodycontent’. So here is morecontent.php

<?php
for($index = 0; $index < 100; $index++)
echo ‘We Are New Lines:’.$index.'<br>’;
?>

That’s about it. Please do throw in your experiences with this in the comments section below and I will surely respond. I am jobless. 🙂


So … here’s the real deal into measuring how much time a user spends on each individual page by url, and measured in milliseconds.

[1] As soon as page loads, set the current time in a variable in javascript with the help of the onload event. Let this variable be called tstart.

[2] On the unload event, get the current timestamp, and subtract from this the starting timestamp, the first one. So tTotal = tend – tstart.

[3] Now send this time information alongwith location.href to your server, which will record this in a log file, or database to use later, maybe to serve relevant content by keeping in mind a users viewing and browsing patterns.You get the picture right.

So here are the files. Just store monitorme.html & logtimefile.php somewhere on your server and create a writeable file in the same directory named as timelog.txt. Now get monitorme.html and now refresh your page, navigate away to some other page or just plain close the window to find your timelog.txt file piling up with times you spent on the page.

monitorme.html

<html>
<head>
<title>Duration Logging Demo</title>
<script type=”text/javascript”>
var oRequest;
var tstart = new Date();

// ooooo, ajax. ooooooo …
if(window.XMLHttpRequest)
oRequest = new XMLHttpRequest();
else if(window.ActiveXObject)
oRequest = new ActiveXObject(“Microsoft.XMLHTTP”);

function sendAReq(sendStr)
// a generic function to send away any data to the server
// specifically ‘logtimefile.php’ in this case
{
oRequest.open(“POST”, “logtimefile.php”, true); //this is where the stuff is going
oRequest.setRequestHeader(“Content-Type”, “application/x-www-form-urlencoded”);
oRequest.send(sendStr);
}

function calcTime()
{
var tend = new Date();
var totTime = (tend.getTime() – tstart.getTime())/1000;
msg = “[URL:” location.href “] Time Spent: ” totTime ” seconds”;
sendAReq(‘tmsg=’ msg);
}
</script>
</head>

<body onbeforeunload=”javascript:calcTime();”>
Hi, navigate away from this page or Refresh this page to find the time you spent seeing
this page in a log file in the server.
</body>
</html>

logtimefile.php

<?php
function logtimemsg($timemsg)
{
//write your own handling code here, store it in a file or store it in a DB, whatever
$logfilename = ‘timelog.txt’;
if (is_writable($logfilename))
{
if (!$handle = fopen($logfilename, ‘a’))
{
exit;
}
if (fwrite($handle, $timemsg.”\r\n”) === FALSE)
{
exit;
}
fclose($handle);
}
}

logtimemsg($_REQUEST[‘tmsg’]);
?>

UPDATE: People, I love if anyone of you writes back, even if to tell that this doesn’t work. Thanks to JayVee who pointed that <body onunload=”> sucks compared to <body onbeforeunload=”> for this post. Change accomodated.


I spend a considerable time debugging errors, and being a Web developer spans debugging CSS, JS and server end files as well. Often I find myself rolling up and down scrollbars on MySQL manual pages.

It annoys me no end to find that a user is seeing Javascript errors. And since Javascript being something that browsers have been trusted with, they had a field day unsupporting it. So it is very common to find that what you think is OK on browser1 is NOT_OK on browser2. And worse still, this incompatibility is reported by someone who is seeing a webpage on a live setup, a user or someone who least expects a nasty bug or something.

Now try this, save this as something.html on your server:

<html>
<head>
<title>Auto JS Bug Reporting Demo</title>
<script type=”text/javascript”>
var msg = null;
var opdiv;
var oRequest;

// ooooo, ajax. ooooooo …
if(window.XMLHttpRequest)
oRequest = new XMLHttpRequest();
else if(window.ActiveXObject)
oRequest = new ActiveXObject(“Microsoft.XMLHTTP”);

onerror = handleErrors;
function handleErrors(errorMessage, url, line)
{
msg = “[URL:” url “, line no.: ” line “] ERROR: ” errorMessage;
sendAReq(“err=” msg, ‘errdiv’);
return true;
}

function sendAReq(sendStr, odiv)
// a generic function to send away any data to the server
// specifically ‘handleerror.php’ in this case
// what the server replies is handled by showcontent()
{
opdiv = odiv;
oRequest.open(“POST”, “handleerror.php”, true); //this is where the stuff is going
oRequest.onreadystatechange = showcontent;
oRequest.setRequestHeader(“Content-Type”, “application/x-www-form-urlencoded”);
oRequest.send(sendStr);
}

function showcontent()
{
if(oRequest.readyState == 4)
{
// if the output is ready, print it to the div as set by the caller function
// in this case, the opdiv, the output div, was set to ‘errdiv’ in handleErrors(…)
if(oRequest.status == 200)
document.getElementById(opdiv).innerHTML = oRequest.responseText;
else
document.getElementById(opdiv).innerHTML = “<font color=#FF3300 > Try Again Later </font>”;
}
}
</script>
</head>

<body onload=”javascript:wow();”>
<div id=”testdiv”>
I am a div. Press this button to cause an error.<br>
<input type=”button” onclick=”javascript:sendAReq(‘err=hello, ‘errdiv’);” value=”GET some FUN”>
</div>
<div id=”errdiv”>
</div>
</body>
</html>

So this file has two Javascript errors. One is in the body onload function, which calls a nonexistent function, wow() . And the other is caused in the onclick event of the only input button on this page.

So this page is easy in the way that …

[1] It assigns handleErrors as the default error handler in the line onerror=handleErrors ;

[2] It puts together a error report in handleErrors and then calls sendAReq with the report string and the name of the output div, in this case being ‘errdiv’.

[3] sendAReq sends the error report to handleerror.php and sets the output div dutifully to ‘errdiv’ in the first line itself. The output div we keep referring to here is the div to which the server will send the output. Oh … so you looking for handleerror.php??


<?php
function logerr($errmsg)
{
//write your own handling code here, store it in a file or store it in a DB, whatever
$errfilename = ‘jsbugs.txt’;
if (is_writable($errfilename))
{
if (!$handle = fopen($errfilename, ‘a’))
{
exit;
}
if (fwrite($handle, $errmsg.’\r\n’) === FALSE)
{
exit;
}
fclose($handle);
}
}
logerr($_REQUEST[‘err’]);
echo “I got this: “.$_REQUEST[‘err’];
?>

[4] So handleerror.php gets the error report Javascript compiled and stores it in a file named jsbugs.txt. make sure you have a file created already named jsbugs.txt with write permissions before you run this script. Should you want to do more fancy things, you can store the User-Agent string, timestamp and a lot more things in a database maybe. I leave that to you.

So now, you are ready with something.html about to be fetched in your browser, and handleerror.php copied in the web root directory and jsbugs.txt in the same directory as well, with write permissions.

Now get something.html and you should see your jsbugs.txt file piling up with cute strings, which may later turn up to be sleep-stealing, girlfriend-eating monsters.