[PHP] dawnwing.co.za Tracking scrapper

dawnwing-co-za-api.php

<?php
/**
 * Tracking package via cURL from www.dawnwing.co.za
 *
 * @author   Kundan K Singh <[email protected]>
 */

define('DOMAIN_REQUEST_URL', 'http://www.dawnwing.co.za/business-tools/online-parcel-tracking/');

include_once 'simple_html_dom.php';

libxml_use_internal_errors(true);

/*
|--------------------------------------------------------------------------
| Helper functions
|--------------------------------------------------------------------------
*/
function getDelimiterValues($html)
{
	$response = array();
	foreach ($html as $k => $s) {
		if( strpos( $s, ':' ) !== false ) {
			$a = explode(' : ', trim($s));
			$response[$a[0]] = $a[1];
		}
	}
	return $response;
}
function extractTd($html)
{
	$response = array();
	$content = str_get_html($html);
	$li_count = substr_count(strtoupper($content), strtoupper('<td'));
	for ($i = 0; $i < $li_count; $i++) {
		$response[] = $content->find('td', $i)->innertext;
	}
	return $response;
}
function extractTh($html)
{
	$response = array();
	$content = str_get_html($html);
	$li_count = substr_count(strtoupper($content), strtoupper('<th'));
	for ($i = 0; $i < $li_count; $i++) {
		$response[] = $content->find('th', $i)->innertext;
	}
	return $response;
}
function getTableDelivery($html)
{
	$response = array();
	$content = str_get_html($html);
	$row_th = $content->find('tr', 1)->innertext;
	$row_td = $content->find('tr', 2)->innertext;
	$td = extractTd($row_td);
	$th = extractTh($row_th);
	foreach ($th as $key => $c) {
		$response[$th[$key]] = $td[$key];
	}
	return $response;
}
function getDetailedTrackTrace($html)
{
	$response = array();
	$content = str_get_html($html);
	$tr_count = substr_count(strtoupper($content), strtoupper('<tr'));

	$row_th = $content->find('tr', 0)->innertext;
	$th = extractTh($row_th);

	for ($i = 1; $i < $tr_count; $i++) {
		$row_td = $content->find('tr', $i)->innertext;
		$td = extractTd($row_td);
		foreach ($th as $key => $c) {
			$response[$i - 1][$th[$key]] = $td[$key];
		}
	}
	
	return $response;
}
/*
|--------------------------------------------------------------------------
| Prepare request data and fetch tracking information
|--------------------------------------------------------------------------
*/
if (isset($_REQUEST['id'])) {
	$waybill_no = $_REQUEST['id'];


	// Prepare request data
	$post_data = array(
		'WaybillNo' => $waybill_no,
		'parcel-search' => 'go',
		);

	// get tracking information form server
	$ch = curl_init();
	curl_setopt($ch, CURLOPT_URL, DOMAIN_REQUEST_URL); 
	curl_setopt($ch, CURLOPT_POST, true);
	curl_setopt($ch, CURLOPT_POSTFIELDS, $post_data);
	curl_setopt($ch, CURLOPT_HEADER, false);
	curl_setopt($ch, CURLOPT_RETURNTRANSFER, 1);
	$html_dom = curl_exec($ch);

	/*
	|--------------------------------------------------------------------------
	| Filter Tracking data
	|--------------------------------------------------------------------------
	*/
	$html = str_get_html($html_dom);
	$track_data = $html->find('div#track-trace', 0)->innertext;
	$html = str_get_html($track_data);

	$track_summary = explode("\n", trim(strip_tags($html->find('div', 0)->innertext)));
	$track_summary2 = explode("\n", trim(strip_tags($html->find('div', 1)->innertext)));

	$response_data['status'] = array_merge(getDelimiterValues($track_summary), getDelimiterValues($track_summary2));
	$response_data['delivery_information'] = getTableDelivery($html->find('div', 2)->innertext);
	$response_data['delivery_trace'] = getDetailedTrackTrace($html->find('table', 1)->innertext);

	header('Cache-Control: no-cache, must-revalidate');
	header('Expires: Mon, 26 Jul 1997 05:00:00 GMT');
	header('Content-type: application/json');
	echo json_encode($response_data, JSON_PRETTY_PRINT);
} else {
	echo json_encode(['message' => 'Sorry no ID Specified'], JSON_PRETTY_PRINT);
	exit();
}