responseText contains extra whitespace characters (new lines, line feeds), how to prevent and remove them?

Question

I have an ajax script that calls a php file.

The php file echos "yes" or "no", I want to use the strings to do logical comparisons.

In the javascript, I want to compare the string in the responseText to see if it is == to "yes" (or "no"). But the comparison fails.

So I do the alert responseText, and it does show "yes" (or "no") as the string. But I read on here that the responseText might contain hidden whitespace characters, so I did the string length of responseText and it shows that the string length is 4 characters longer than what it should be. So I escaped the responseText alert(escape(responseText)) and it shows that I have %0A and %0D (newlines and line feeds) hidden at the end of the responseText string.

I read that these characters are added on by php, but I also read that the extra characters are different among different php versions/servers.

How to prevent these extra whitespaces without using regex, since regex might remove intentional whitespaces?

Please don't suggest using jquery or mootools as answers.

TIA

Answer 1

I read that these characters are added on by php, but I also read that the extra characters are different among different php versions/servers.

That's wrong. That's simple to verify: create a test.php file, write this and only this: <?php echo "test"; (without ?>) into it and execute it. There will be no whitespace.

These whitespaces most probably come from your scripts. A common error is to leave some trailing newlines after a closing php tags (?>), which results in a new line being printed.

Verify that all files included before or after you do echo "yes"; don't echo anything and don't have a trailing newline after a ?>.

The easiest way to avoid this problem is do not use php close tags at end of files (they are not mandatory).

Answer 2

I know you asked about doing the comparison without regexes, but given the conditions you mentioned above, that's going to be a very quick and effective way to get your answer, and won't necessarily perturb any other processing.

var trimmedResponse = responseText.replace(/^\s*/,'').replace(/\s*$/,'').toLowerCase();
if (trimmedResponse == 'yes') {
  // do your 'yes' case
} else if (trimmedResponse == 'no') {
  // do your 'no' case
} else {
  // do your 'none of the above' case
}

That's going to trim off leading white space, trailing white space (including the CR/LF combo), and convert to lower case just for comparison.

Answer 3

I think you may be going at it the wrong way. Instead of manually trying to create a response why don´t you use PHP arrays as a data structure and JSON for delivery?

<?php
$flag = false
if (condition){
  $flag = true;
}

$arr = array("is_true" => $flag)
$json = json_encode($arr);

// See http://www.geekality.net/2010/06/27/php-how-to-easily-provide-json-and-jsonp/
// Set JSONP header
header('content-type: application/json; charset=utf-8');

// Get callback from $_GET and prepend the JSON data
echo isset($_GET['callback'])
    ? "{$_GET['callback']}($json)"
    : $json;

Answer 4

Did you try using PHP's trim() function before sending the data?

Answer 5

Even if it's possible to look after all PHP script to avoid CRLF before or after php tag, this is not a good option as you can imagine adding in a few days, or month or years (without wanting!) a CRLF in a file and this will impact another part of your web site.

In fact, to avoid this problem you have two options:

Option 1: Let php send the data so with the garbage at beginning, and clean the data in Javascrpt with:

 response = http.responseText;
 response = response.replace(/[\n\r]+/g, '');

In this case this means you clean ALL the CRLF (change the Regex to clean only at beginning if needed or to clean also spaces)

Option 2 (better I think) is to clean the output buffer of PHP immediatly before sending the data to browser. So:

 ... many code here
 ob_end_clean();
 echo $data_for_the_browser;