Convert Disqus export to a JSON file

disqus-to-json.php

<?php

$xml = simplexml_load_file('php://stdin');

$threads = [];
$posts = [];

$userMap = [
  'evertp' => [
    'url' => 'https://evertpot.com',
    'avatar' => 'https://s.gravatar.com/avatar/9e9b4f80c65dc6076d74cd421084e4fc?s=200',
  ]
];

foreach($xml->thread as $thread) {

  $thread = [
    'id' => (string)$thread->attributes('http://disqus.com/disqus-internals')['id'],
    'url' => (string)$thread->link,
    'post_id' => (string)$thread->id,
    'comments' => [],
  ];

  $threads[$thread['id']] = $thread;

}

foreach($xml->post as $post) {

  $threadId = (string)$post->thread->attributes('http://disqus.com/disqus-internals')['id'];
  $postId = (string)$post->attributes('http://disqus.com/disqus-internals')['id'];
  if ($post->parent) {
    $parentId = (string)$post->parent->attributes('http://disqus.com/disqus-internals')['id'];
  } else {
    $parentId = null;
  }

  $posts[$postId] = [
    'id' => $postId,
    'created' => (string)$post->createdAt,
    'isDeleted' => (string)$post->isDeleted !== 'false',
    'isClosed' => (string)$post->isClosed !== 'false',
    'isSpam' => (string)$post->isSpam !== 'false',
    'name' => (string)$post->author->name,
    'user' => (string)$post->author->username,
    'message' => (string)$post->message,
    'children' => [],
  ];



  if ($parentId) {
    $posts[$parentId]['children'][] =& $posts[$postId];
  } else {
    $threads[$threadId]['comments'][] =& $posts[$postId];
  }

}

function convertComments(array $comments) {

  global $userMap;
  $result = [];
  foreach($comments as $comment) {

    if($comment['isDeleted']) {
      continue;
    }
    if($comment['isSpam']) {
      continue;
    }

    $newComment = [
      'created' => $comment['created'],
      'name' => $comment['name'],
      'disqusUser' => $comment['user'],
      'message' => $comment['message'],
      'children' => convertComments($comment['children']),
    ];
    if (isset($userMap[$newComment['disqusUser']])) {
      $newComment['url'] = 'https://evertpot.com/';
      $newComment['avatar'] = 'https://s.gravatar.com/avatar/9e9b4f80c65dc6076d74cd421084e4fc?s=200';
    }
    $result[] = $newComment;

  }
  return $result;

}

function convertThreads(array $threads) {

  $result = [];
  foreach($threads as $thread) {

    if (!$thread['comments']) {
      continue;
    }

    $result[$thread['post_id']] = convertComments($thread['comments']);

  }
  return $result;

}

$threads = convertThreads($threads);
echo json_encode($threads, JSON_PRETTY_PRINT | JSON_UNESCAPED_SLASHES);

'post_id' => (string)$thread->id,
Needs to be turned into
'post_id' => (string)$thread->attributes('http://disqus.com/disqus-internals')['id'],

on recent Disqus exports. Otherwise this works well, thanks!!

@sven337 That doesn't seem right, as there was both a post_id and id field, and the id field already has this line:

    'id' => (string)$thread->attributes('http://disqus.com/disqus-internals')['id'],

I know, but it seems to work. My XML export has:

<thread dsq:id="2390033662">
<id />

The "id" for all threads is empty. Using the dsq:id seems to make the export produce something that makes sense.

EDIT: actually this is broken too, the commetns don't thread properly. I need to debug more.

Yea that's what I would expect to break =) there's no reference anymore to what post anything belongs to

Actually, this works just fine. The reason I had a threading issue was related to URIs changing on my blog.
So we're back to: my change seems to work even though I have no idea why (and don't really care to find out).

Thanks a lot for your helpful article!!