Created
March 24, 2015 15:34
-
-
Save jimbojsb/8c46ef720643af3f118b to your computer and use it in GitHub Desktop.
Entertainment.com Crawler
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
<?php | |
/** | |
{ | |
"require": { | |
"symfony/dom-crawler": "2.*", | |
"symfony/css-selector": "2.*" | |
} | |
} | |
*/ | |
require_once __DIR__ . '/../vendor/autoload.php'; | |
use \Symfony\Component\DomCrawler\Crawler; | |
$books = []; | |
$context = stream_context_create([ | |
"http" => [ | |
"user_agent" => "Offers.com Entertainment Bot (bot@offers.com)" | |
] | |
]); | |
$baseUrl = "https://www.entertainment.com"; | |
$handle = fopen($baseUrl . "/coupon/view_all_products.cmd", 'r', false, $context); | |
$rootHtml = stream_get_contents($handle); | |
$rootDomCrawler = new Crawler($rootHtml); | |
$books = []; | |
$rootDomCrawler->filter(".region")->each(function(Crawler $node) use ($baseUrl, $context, &$books) { | |
$state = $node->filter(".region-toggle")->first()->extract("_text")[0]; | |
echo "Processing $state\n"; | |
$node->filter(".product")->each(function(Crawler $node) use ($baseUrl, $context, &$books) { | |
$book["name"] = $node->filter(".locations")->first()->extract(["_text"])[0]; | |
echo "Loading " . $book["name"] . "\n"; | |
$book["url"] = $node->filter(".getDetails")->first()->extract(["href"])[0]; | |
$book["retail_price"] = trim(str_replace("$", "", strip_tags($node->filter(".retailPrice")->first()->extract(["_text"])[0]))); | |
$book["sale_price"] = str_replace("$", "", $node->filter(".yourPrice")->first()->extract(["_text"])[0]); | |
$bookUrl = $baseUrl . $book["url"]; | |
$handle = fopen($bookUrl, 'r', false, $context); | |
$bookHtml = stream_get_contents($handle); | |
if ($bookHtml) { | |
$crawler = new Crawler($bookHtml); | |
$crawler->filter(".col-sm-6.col-md-4 .thumbnail")->each(function(Crawler $node, $i) use (&$book) { | |
if ($i > 0) { // only way i could find to select only the right .thumbnails due to invalid html | |
$feature["type"] = $node->filter(".feature-type")->first()->extract(["_text"])[0]; | |
$feature["name"] = $node->filter("h3")->first()->extract(["_text"])[0]; | |
$feature["offer"] = $node->filter("h4")->first()->extract(["_text"])[0]; | |
$feature["detail"] = $node->filter(".featured-details")->first()->extract(["_text"])[0]; | |
$feature["location"] = $node->filter(".featured-location")->first()->extract(["_text"])[0]; | |
$book["features"][] = $feature; | |
} | |
}); | |
$crawler->filter(".panel")->each(function(Crawler $node, $i) use (&$book, $crawler) { | |
$area["name"] = $node->filter("a")->extract(["_text"])[0]; | |
$modalId = "modal" . ($i + 1); | |
$crawler->filter("#$modalId")->each(function(Crawler $node) use (&$area, $book) { | |
$area["cities"] = $node->filter("li.cities")->extract(["_text"]); | |
}); | |
$area["companies"] = $node->filter("li")->extract(["_text"]); // no exception here because they don't all have them | |
$book["areas"][] = $area; | |
}); | |
if (stripos($bookHtml, 'free shipping')) { | |
$book["free_shipping"] = true; | |
} | |
} | |
$books[] = $book; | |
}); | |
}); | |
print_r($books); |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment