Structured JSON layout

Question

I'm working with JSON by PHP at the moment, when I encode it, it would output as:

{"username":"ND","email":"[email protected]","regdate":"8th June 2010","other":{"alternative":"ND"},"level":"6"}

When I would like it to output like this:

{
    "username": "ND",
    "email": "[email protected]",
    "regdate": "8th June 2010",
    "other":
    {
        "alternative": "ND"
    },
    "level":"6"
}

So that me and my other developers can read it well enough when structured. How can I do this?

Example also like this:

https://graph.facebook.com/19292868552

Cheers

Answer 1

I could find this useful myself, so here's a little function I wrote:

<?php
function json_pretty_encode($obj)
{
    $json = json_encode($obj);
    if (!$json) return $json;

    $f = '';
    $len = strlen($json);

    $depth = 0;
    $newline = false;

    for ($i = 0; $i < $len; ++$i)
    {
        if ($newline)
        {
            $f .= "\n";
            $f .= str_repeat(' ', $depth);
            $newline = false;
        }

        $c = $json[$i];
        if ($c == '{' || $c == '[')
        {
            $f .= $c;
            $depth++;
            $newline = true;
        }
        else if ($c == '}' || $c == ']')
        {
            $depth--;
            $f .= "\n";
            $f .= str_repeat(' ', $depth);
            $f .= $c;
        }
        else if ($c == '"')
        {
            $s = $i;
            do {
                $c = $json[++$i];
                if ($c == '\\')
                {
                    $i += 2;
                    $c = $json[$i];
                }
            } while ($c != '"');
            $f .= substr($json, $s, $i-$s+1);
        }
        else if ($c == ':')
        {
            $f .= ': ';
        }       
        else if ($c == ',')
        {
            $f .= ',';
            $newline = true;
        }
        else
        {
            $f .= $c;
        }
    }

    return $f;
}
?>

It's naive, trusting that PHP will return a valid JSON string. It could be written more concisely, but it's easy to modify this way. (And of course this adds unnecessary overhead in production scenarios where only a machine reads the text.)

Edit: Added an else clause to catch numbers and other unknown chars.