forked from blallo/rss-bridge
[KununuBridge] Add new bridge
This commit is contained in:
logmanoriginal
parent
cd7eb1b8fe
commit
f0f2e8f847
1 changed files with 278 additions and 0 deletions
278
bridges/KununuBridge.php
Normal file
278
bridges/KununuBridge.php
Normal file
|
|
@ -0,0 +1,278 @@
|
|||
<?php
|
||||
class KununuBridge extends BridgeAbstract{
|
||||
public function loadMetadatas(){
|
||||
$this->maintainer = "logmanoriginal";
|
||||
$this->name = "Kununu Bridge"; /* This will be replaced later! */
|
||||
$this->uri = "https://www.kununu.com"; /* This will be replaced later! */
|
||||
$this->description = "Returns the latest reviews for a company and site of your choice.";
|
||||
$this->update = "2016-08-10";
|
||||
|
||||
$this->parameters['global'] =
|
||||
'[
|
||||
{
|
||||
"name" : "Site",
|
||||
"identifier" : "site",
|
||||
"type" : "list",
|
||||
"required" : "true",
|
||||
"exampleValue" : "United States",
|
||||
"title" : "Select your site",
|
||||
"values" :
|
||||
[
|
||||
{
|
||||
"name" : "Austria",
|
||||
"value" : "at"
|
||||
},
|
||||
{
|
||||
"name" : "Germany",
|
||||
"value" : "de"
|
||||
},
|
||||
{
|
||||
"name" : "Switzerland",
|
||||
"value" : "ch"
|
||||
},
|
||||
{
|
||||
"name" : "United States",
|
||||
"value" : "us"
|
||||
}
|
||||
]
|
||||
},
|
||||
{
|
||||
"name" : "Load full article",
|
||||
"identifier" : "full",
|
||||
"type" : "checkbox",
|
||||
"required" : "false",
|
||||
"exampleValue" : "checked",
|
||||
"title" : "Activate to load full article"
|
||||
}
|
||||
]';
|
||||
|
||||
$this->parameters[] =
|
||||
'[
|
||||
{
|
||||
"name" : "Company",
|
||||
"identifier" : "company",
|
||||
"type" : "text",
|
||||
"required" : "true",
|
||||
"exampleValue" : "kununu-us",
|
||||
"title" : "Insert company name (i.e. Kununu US) or URI path (i.e. kununu-us)"
|
||||
}
|
||||
]';
|
||||
}
|
||||
|
||||
public function collectData(array $params){
|
||||
|
||||
// Get Site
|
||||
$site = strtolower(trim($params['site']));
|
||||
if(!isset($site) || empty($site) || !$this->site_is_valid($site))
|
||||
$this->returnError('You must specify a valid site (&site=...)!', 400);
|
||||
|
||||
// Get Company (fixing whitespace and umlauts)
|
||||
$company = $this->encode_umlauts(strtolower(str_replace(' ', '-', trim($params['company']))));
|
||||
if(!isset($company) || empty($company))
|
||||
$this->returnError('You must specify a company (&company=...)!', 400);
|
||||
|
||||
$full = false; // By default we'll load only short article
|
||||
if(isset($params['full']))
|
||||
$full = strtolower(trim($params['full'])) === 'on';
|
||||
|
||||
// Get reviews section name (depends on site)
|
||||
$section = '';
|
||||
switch($site){
|
||||
case 'at':
|
||||
case 'de':
|
||||
case 'ch':
|
||||
$section = 'kommentare';
|
||||
break;
|
||||
case 'us':
|
||||
$section = 'reviews';
|
||||
break;
|
||||
default:
|
||||
$this->returnError('The reviews section is not defined for you selection!', 404);
|
||||
}
|
||||
|
||||
// Update URI for the content
|
||||
$this->uri .= "/{$site}/{$company}/{$section}";
|
||||
|
||||
// Load page
|
||||
$html = $this->file_get_html($this->uri);
|
||||
if($html === false)
|
||||
$this->returnError('Unable to receive data from ' . $this->uri . '!', 404);
|
||||
|
||||
// Update name for this request
|
||||
$this->name = $this->extract_company_name($html) . ' - ' . $this->name;
|
||||
|
||||
// Find the section with all the panels (reviews)
|
||||
$section = $html->find('section.kununu-scroll-element', 0);
|
||||
if($section === false)
|
||||
$this->returnError('Unable to find panel section!', 404);
|
||||
|
||||
// Find all articles (within the panels)
|
||||
$articles = $section->find('article');
|
||||
if($articles === false || empty($articles))
|
||||
$this->returnError('Unable to find articles!', 404);
|
||||
|
||||
// Go through all articles
|
||||
foreach($articles as $article){
|
||||
$item = new \Item();
|
||||
|
||||
$item->author = $this->extract_article_author_position($article);
|
||||
$item->timestamp = $this->extract_article_date($article);
|
||||
$item->title = $this->extract_article_rating($article) . ' : ' . $this->extract_article_summary($article);
|
||||
$item->uri = $this->extract_article_uri($article);
|
||||
|
||||
if($full)
|
||||
$item->content = $this->extract_full_description($item->uri);
|
||||
else
|
||||
$item->content = $this->extract_article_description($article);
|
||||
|
||||
$this->items[] = $item;
|
||||
}
|
||||
}
|
||||
|
||||
/**
|
||||
* Returns true if the given site is part of the parameters list
|
||||
*/
|
||||
private function site_is_valid($site){
|
||||
$parameter = json_decode($this->parameters['global'], true);
|
||||
$sites = $parameter[0]['values'];
|
||||
|
||||
$site_names = array();
|
||||
|
||||
foreach($sites as $site_item)
|
||||
$site_names[] = $site_item['value'];
|
||||
|
||||
return in_array($site, $site_names);
|
||||
}
|
||||
|
||||
/**
|
||||
* Fixes relative URLs in the given text
|
||||
*/
|
||||
private function fix_url($text){
|
||||
return preg_replace('/href=(\'|\")\//i', 'href="https://www.kununu.com/', $text);
|
||||
}
|
||||
|
||||
/**
|
||||
* Encodes unmlauts in the given text
|
||||
*/
|
||||
private function encode_umlauts($text){
|
||||
$umlauts = Array("/ä/","/ö/","/ü/","/Ä/","/Ö/","/Ü/","/ß/");
|
||||
$replace = Array("ae","oe","ue","Ae","Oe","Ue","ss");
|
||||
|
||||
return preg_replace($umlauts, $replace, $text);
|
||||
}
|
||||
|
||||
/**
|
||||
* Returns the company name from the review html
|
||||
*/
|
||||
private function extract_company_name($html){
|
||||
$panel = $html->find('div.panel', 0);
|
||||
if($panel === false)
|
||||
$this->returnError('Cannot find panel for company name!', 404);
|
||||
|
||||
$company_name = $panel->find('h1', 0);
|
||||
if($company_name === false)
|
||||
$this->returnError('Cannot find company name!', 404);
|
||||
|
||||
return $company_name->plaintext;
|
||||
}
|
||||
|
||||
/**
|
||||
* Returns the date from a given article
|
||||
*/
|
||||
private function extract_article_date($article){
|
||||
// They conviniently provide a time attribute for us :)
|
||||
$date = $article->find('time[itemprop=dtreviewed]', 0);
|
||||
if($date === false)
|
||||
$this->returnError('Cannot find article date!', 404);
|
||||
|
||||
return strtotime($date->datetime);
|
||||
}
|
||||
|
||||
/**
|
||||
* Returns the rating from a given article
|
||||
*/
|
||||
private function extract_article_rating($article){
|
||||
$rating = $article->find('span.rating', 0);
|
||||
if($rating === false)
|
||||
$this->returnError('Cannot find article rating!', 404);
|
||||
|
||||
return $rating->getAttribute('aria-label');
|
||||
}
|
||||
|
||||
/**
|
||||
* Returns the summary from a given article
|
||||
*/
|
||||
private function extract_article_summary($article){
|
||||
$summary = $article->find('[itemprop=summary]', 0);
|
||||
if($summary === false)
|
||||
$this->returnError('Cannot find article summary!', 404);
|
||||
|
||||
return strip_tags($summary->innertext);
|
||||
}
|
||||
|
||||
/**
|
||||
* Returns the URI from a given article
|
||||
*/
|
||||
private function extract_article_uri($article){
|
||||
// Notice: This first part is the same as in extract_article_summary!
|
||||
$summary = $article->find('[itemprop=summary]', 0);
|
||||
if($summary === false)
|
||||
$this->returnError('Cannot find article summary!', 404);
|
||||
|
||||
$anchor = $summary->find('a', 0);
|
||||
if($anchor === false)
|
||||
$this->returnError('Cannot find article URI!', 404);
|
||||
|
||||
return 'https://www.kununu.com' . $anchor->href;
|
||||
}
|
||||
|
||||
/**
|
||||
* Returns the position of the author from a given article
|
||||
*/
|
||||
private function extract_article_author_position($article){
|
||||
// We need to parse the aside manually
|
||||
$aside = $article->find('aside', 0);
|
||||
if($aside === false)
|
||||
$this->returnError('Cannot find article author information!', 404);
|
||||
|
||||
// Go through all h2 elements to find index of required span (I know... it's stupid)
|
||||
$author_position = 'Unknown';
|
||||
foreach($aside->find('h2') as $subject){
|
||||
if(stristr(strtolower($subject->plaintext), 'position')){ /* This works for at, ch, de, us */
|
||||
$author_position = $subject->next_sibling()->plaintext;
|
||||
break;
|
||||
}
|
||||
}
|
||||
|
||||
return $author_position;
|
||||
}
|
||||
|
||||
/**
|
||||
* Returns the description from a given article
|
||||
*/
|
||||
private function extract_article_description($article){
|
||||
$description = $article->find('div[itemprop=description]', 0);
|
||||
if($description === false)
|
||||
$this->returnError('Cannot find article description!', 404);
|
||||
|
||||
return $this->fix_url($description->innertext);
|
||||
}
|
||||
|
||||
/**
|
||||
* Returns the full description from a given uri
|
||||
*/
|
||||
private function extract_full_description($uri){
|
||||
// Load full article
|
||||
$html = file_get_html($uri);
|
||||
if($html === false)
|
||||
$this->returnError('Could not load full description!', 404);
|
||||
|
||||
// Find the article
|
||||
$article = $html->find('article', 0);
|
||||
if($article === false)
|
||||
$this->returnError('Cannot find article!', 404);
|
||||
|
||||
// Luckily they use the same layout for the review overview and full article pages :)
|
||||
return $this->extract_article_description($article);
|
||||
}
|
||||
}
|
||||
Loading…
Reference in a new issue