hugo-mastodon-comments/static/comments/getcomments.php

<?php
/*
  SPDX-License-Identifier: AGPL-3.0-or-later
  SPDX-FileCopyrightText: 2019 Max Mehl <mail@mehl.mx>
*/

/* load config. You normally don't want to edit something here */
require_once 'config.php';
$instance = $config['mastodon-instance'];
$uid = $config['user-id'];
$searchurl = $config['search-url'];
$search = isset($_GET['search']) ? strtolower($_GET['search']) : '';
$debug_on = $config['debug'];
/* cache files */
$ctt = $config['cache_toots'];
$dbt = "cache-toots.json";
$ctc = $config['cache_comments'];
$dbc = "cache-comments_%id.json";

/* Exit if search empty */
if (empty($search)) {
  debug("No proper search given");
  die();
}

/* MISC FUNCTIONS */
function debug($data) {
  global $debug_on;
  if ($debug_on === true) {
    error_log("[getcomments.php] " . print_r($data, TRUE));
  }
}

/* CACHE FUNCTIONS */
/* write data to file */
function write_db($db, $data, $id) {
  // if $id is given, it's a comments file. Replace placeholder in filename
  if ($id) {
    $db = str_replace('%id', $id, $db);
  }
  $file['toots'] = $data;
  $file['timestamp'] = time();
  // encode and write file
  $encoded = json_encode($file, JSON_PRETTY_PRINT);
  file_put_contents($db, $encoded, LOCK_EX);
}
/* delete file */
function delete_db($db, $id) {
  // if $id is given, it's a comments file. Replace placeholder in filename
  if ($id) {
    $db = str_replace('%id', $id, $db);
  }
  unlink($db);
}
/* access data from file */
function read_db($db, &$data, $cachetime, &$cachebreak, $id) {
  // if $id is given, it's a comments file. Replace placeholder in filename
  if ($id) {
    $db = str_replace('%id', $id, $db);
  }
  // if DB does not exist, create it with empty array
  if (! file_exists($db)) {
    // if $data empty (usually with $toots, not with comment's $result), populate with empty array
    if (empty($data)) {
      $data = array();
    }
    touch($db);
    write_db($db, $data, $id);
    $cachebreak = true;
  }
  $file = file_get_contents($db, true);
  $data = json_decode($file, true);

  // check if timestamp in cache file too old
  if (empty($data['timestamp']) || ($data['timestamp'] + $cachetime < time())) {
    $cachebreak = true;
  }

  $data = $data['toots'];
}

/* TOOT FUNCTIONS */
function collectToots($instance, $uid, $min_id, $searchurl) {
  $raw = file_get_contents("$instance/api/v1/accounts/$uid/statuses?exclude_reblogs=true&exclude_replies=true&limit=50&min_id=$min_id");
  $json_complete = json_decode($raw, true);
  $json = array();
  foreach ($json_complete as $toot) {
    $json[] = array('id' => $toot['id'], 'date' => $toot['created_at'] ,'url' => analyzeToot($instance, $toot['id'], $searchurl));
  }
  return($json);
}
/* Find out if a toot contains the searched URL */
function analyzeToot($instance, $id, $searchurl) {
  debug("Searching for $searchurl in $id");
  $raw = file_get_contents("$instance/api/v1/statuses/$id");
  $json = json_decode($raw, true);

  // search for $searchurl inside of <a> tags, until (and excluding) a "
  preg_match("|$searchurl.+?(?=\")|i", $json['content'], $matches);

  if(!empty($matches)) {
    return(strtolower($matches[0]));  // take first match inside toot
  } else {
    return("");
  }
}
/* of context, extract the interesting bits */
function filterComments($descendants, $root, &$result) {
  // go through each comment
  foreach ($descendants as $d) {
    $result['comments'][$d['id']] = [
      'author' => [
        'display_name' => $d['account']['display_name'] ? $d['account']['display_name'] : $d['account']['username'],
        'avatar' => $d['account']['avatar_static'],
        'url' => $d['account']['url']
      ],
      'toot' => $d['content'],
      'date' => $d['created_at'],
      'url' => $d['uri'],
      'reply_to' => $d['in_reply_to_id'],
      'root' => $root,
    ];
  }
  return $result;
}
/* get /context of toot */
function tootContext($instance, $id, &$result) {
  $raw = file_get_contents("$instance/api/v1/statuses/$id/context");
  $json = json_decode($raw, true);
  filterComments($json['descendants'], $id, $result);
}
/* extract stats info from toot */
function filterStats($stats) {
  $result = [
    'reblogs' => (int)$stats['reblogs_count'],
    'favs' => (int)$stats['favourites_count'],
    'replies' => (int)$stats['replies_count'],
    'url' => $stats['url']
  ];
  return $result;
}
/* for toot, extract interesting statistics */
function tootStats($instance, $id, &$result) {
  debug("Checking ID $id");
  $raw = file_get_contents("$instance/api/v1/statuses/$id");
  $json = json_decode($raw, true);
  $newStats = filterStats($json);
  $result['stats']['reblogs'] += $newStats['reblogs'];
  $result['stats']['favs'] += $newStats['favs'];
  $result['stats']['replies'] += $newStats['replies'];
  if (empty($result['stats']['url'])) {
    $result['stats']['url'] = $newStats['url'];
  }
}

/***************
 * START PROGRAM
 ***************/

/* check whether the cached file containing all toots is older than max. cache time */
// this at the same time loads the cached DB, either way
$cachebreak = false;
read_db($dbt, $toots, $ctt, $cachebreak, false);

if ($cachebreak) {
  /* Collect all the toots */
  /* get id of latest cached toot, and set as $min_id */
  debug("Toots cache outdated. Checking for new toots");
  if (!empty($toots['0']['id'])) {
    $min_id_cached = $toots['0']['id'];
    $min_id = $min_id_cached;
  } else {
    /* if cached toots do not exist, start from oldest toot */
    $min_id = "0";
    $min_id_cached = "0";
  }

  /* test whether there are new toots available */
  // Search for toots older than the cached latest toot ID ($min_id)
  $uptodate = false;
  while ($uptodate === false) {
    $toots = array_merge(collectToots($instance, $uid, $min_id, $searchurl), $toots);
    $min_id_new = $toots['0']['id']; // the latest ID of the recent search

    if ($min_id_new === $min_id) {
      // min_id is the latest, let's write the new DB and end this loop
      $uptodate = true;
      debug("Toots up-to-date. Rewrite cache DB.");
      write_db($dbt, $toots, false);
    } else {
      // next round looks for toots newer than the newly found ID
      debug("Newer toots than in cache found. Starting another search for new toots");
      $min_id = $min_id_new;
    }
  }
} else {
  debug("Toots cache is up-to-date");
}

// create empty $result
$result_empty = ['comments' => [], 'stats' => ['reblogs' => 0, 'favs' => 0, 'replies' => 0, 'url' => '', 'root' => 0]];
$result = $result_empty;

/* check if URL from $search exists in $toots */
$id = array_keys(
  array_filter(
    array_column($toots, 'url'),
    function ($value) use ($search) {
      return (strpos($value, $search) !== false);
    }
  )
);
if (empty($id)) {
  debug("Blog URL \"$search\" has not been found");
} else {
  // if multiple toots with the searched URL exist, take the oldest one (largest array index)
  $id = $toots[end($id)]['id'];

  /* read cached comments, or reload new comments if cached data too old */
  $cachebreak = false;
  read_db($dbc, $result, $ctc, $cachebreak, $id);

  if ($cachebreak) {
    debug("Comments cache for $id outdated. Checking for new comments");
    // delete old cache file, otherwise the stats would add up
    delete_db($dbc, $id);
    // re-create empty $result and new cache file
    $result = $result_empty;
    read_db($dbc, $result, $ctc, $cachebreak, $id);
    /* Extract comments and stats from toot */
    tootContext($instance, $id, $result);
    tootStats($instance, $id, $result);
    // FIXME: At the moment the API doesn't return the correct replies count so I count it manually
    $result['stats']['replies'] = count($result['comments']);
    $result['stats']['root'] = $id;

    write_db($dbc, $result, $id);
  } else {
    debug("Comments cache for $id up-to-date. Returning cached comments");
  }
}

// headers for not caching the results
header('Cache-Control: no-cache, must-revalidate');
header('Expires: Mon, 26 Jul 1997 05:00:00 GMT');

// headers to tell that result is JSON
header('Content-type: application/json');

// actually output result as JSON, to be digested by getcomments.js
echo json_encode($result);

?>