Is it possible to ensure the same order when fetching mails from POP3 servers? I'm trying to interface with GMail using the code below, but the order changes (i.e. message #1 isn't always the same). In between calls of the script, I make no changes (including email receipt) to GMail, so I'm not sure why the order is changing or, more importantly, how to stop it.

Any help is greatly appreciated.

PHP Code:

<?php

require_once "Net/POP3.php";
require_once "Mail/mimeDecode.php";

$pop3 =& new Net_POP3();

$ia_configuration_data['ia_email'] = '********';
$ia_configuration_data['ia_password']  = '++++++++++';
$ia_configuration_data['ia_server']  = 'pop.gmail.com';
$ia_configuration_data['port']  = '995';
if (!$pop3->connect('ssl://'.$ia_configuration_data['ia_server'], $ia_configuration_data['port'])) {
  //unable to connect to mail server.
  echo __file__.':'.__line__.'<br />';
  print '<pre>'.print_r($pop3,true).'</pre>';
  exit();
}
if ($pop3->login($ia_configuration_data['ia_email'], $ia_configuration_data['ia_password']) !== true) {
  //invalid login credentials
  echo __file__.':'.__line__.'<br />';
  print '<pre>'.print_r($pop3,true).'</pre>';
  exit();
}

$params['include_bodies'] = true;
$params['decode_bodies']  = true;
$params['decode_headers'] = true;

$num_messages = $pop3->numMsg();
$start_message = 1;
$end_message = min(10,$num_messages);

for ($counter=$start_message;$counter<=$end_message;$counter++) {
  $message_details = $pop3->getMsg($counter);
  $decoder = new Mail_mimeDecode($message_details);
  $structure = $decoder->decode($params);
  
  //Do stuff
}
/*
* Disconnect
*/
$pop3->disconnect();
?>

Couldn't you pop them into an array and sort them by the date?

Well, I'm writing this for someone who hasn't checked their email in a year and has 10k+ emails, so I need to do them a little bit at a time and can't load all into memory b/c I'd run out or else that'd be a perfect solution.

Its possible to merge sort without having all of the items in memory at once.

OK. You've piqued my interest -- what are you talking about?

http://en.wikipedia.org/wiki/Merge_sort

Merge sort works by taking a list of unsorted items and splitting it into sublists. Normally you would do this by taking your origional list and recursively dividing it in half until your ideal list size is reached. But in your case you will most likely have to just come up with a maximum list size and then gather that many items from pop3. Then sort that list however you choose and repeat. Once you have all of your sub lists sorted, compare all of the smallest entries from your sublists, the smallest of these will be the smallest entry in the entire list. You can just continue doing this until you have a completely sorted list.

In your case the "smallest" entry will be the one that was sent most recently. Your biggest challenge will be storing all of the sublists, but this can be accomplished using a database.

What might work better than merge sort in this case would be to just feed all of your emails into a database and then let the database sort them.

Oh. Thanks for the clarification. I should have been more clear. I'm wanting to retrieve them already sorted (I don't care by what, so long as message #1 is always message #1) so that I can fetch only messages 1-10 the first time the script runs, 11-20 the second, and so on. I read up on the POP3 spec and apparently this isn't supported, so I don't have a choice here.

Thanks again for the clarification.

Could you mark them read after you retrieve them and then only retrieve unread messages?