Looping with XPath in php to get tables values - php

Question

I have this html table:
<tbody>
<tr>..</tr>
<tr>
<td class="tbl_black_n_1">1</td>
<td class="tbl_black_n_1" nowrap="" align="center">23/07/14 08:10</td>
<td class="tbl_black_n_1">
<img src="http://www.betonews.com/img/SportId389.gif" width="10" height="10" border="0" alt="">
</td>
<td class="tbl_black_n_1"></td>
<td class="tbl_black_n_1" nowrap="" align="center">BAK WS</td>
<td class="tbl_black_n_1" nowrap="" align="right">M. Eguchi</td>
<td class="tbl_black_n_1" align="center">-</td>
<td class="tbl_black_n_1" nowrap="">Radwanska U. </td>
<td class="tbl_black_n_1" align="center" title=" ">1,02</td>
<td class="tbl_black_n_1" align="center">
<td class="tbl_black_n_1" align="center" title=" "> </td>
<td class="tbl_black_n_1" align="center">
<td class="tbl_black_n_1" align="center" title=" ">55,00</td>
<td class="tbl_black_n_1" align="center">
<td class="tbl_black_n_1" align="right">86%</td>
<td class="tbl_black_n_1" align="right">-</td>
<td class="tbl_black_n_1" align="right">14%</td>
<td class="tbl_black_n_1" align="center" title=" ">524.647</td>
<td class="tbl_black_n_1" nowrap="">
<img src="http://www.betonews.com//img/i_betfair.gif" width="12" height="10" border="0" alt="">
<img src="http://www.betonews.com//img/i_history.gif" width="12" height="10" border="0" alt="">
</td>
</tr>
<tr>..</tr>
<tr>..</tr>
<tr>..</tr>
...
</tbody>
There are more than one hundred <tr> structured at the same way, which contain lots of <td>. How can I loop with xpath to store all data in a database? I don't want to get the first <tr>: the query has to begin with the second <tr> (that I have showed).
This is my php code, but I can not go on.. help!
<?php
$url = 'http://www.betonews.com/table.asp?tp=2001&lang=en&dd=23&dm=7&dy=2014&df=1&dw=3';
$ch = curl_init();
curl_setopt($ch, CURLOPT_URL, $url);
curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
$response = curl_exec($ch);
curl_close($ch);
$document = new DOMDocument();
$document->loadHTML($response);
$xpath = new DOMXPath($document);
$expression = '/html/body/table[2]/tbody/tr/td[2]/table/tbody/tr/td[2]/table/tbody/tr[2]/td/table/tbody/tr/td/table/tbody/tr[3]/td/table/tbody/tr';
$rows = $xpath->query($expression);
$results = array();
foreach ($rows as $row) {
$result = array();
???
}
This is what I want to be the final result:
[0] => Array
(
[date] => 23/07/14 08:10
[image] => http://www.betonews.com/img/SportId389.gif
[team1] => M. Eguchi
[team2] => Radwanska U.
[1] => 1,02
[x] => 0
[2] => 55,00
[1%] => 86%
[x%] => 0
[2%] => 14%
[total] => 524.647
)

Answer 1

I would use a different XPath to select the table. First, there is always a problem using absolute paths with tables like this, because often tbody elements are just added by the browser, but they are not actually present in the document, i.e. not visible to the PHP code. Also, because if anything in the source HTML changes in terms of styleing, your code breaks. Now I select the first table with a cellpadding of 3 - This is not optimal, but there wasn't any obvious unique identifier.
Apart from that, you can simply iterate over the DOMNodeList result and then get the correct child nodes. Notice, that the items are increased by two, because whitespace-only elements in between are also a node in XML.
$xpath = new DOMXPath($document);
$expression = '(//table[#cellpadding="3"])[1]/tr[position() > 1]';
$rows = $xpath->query($expression);
$results = array();
foreach ($rows as $row) {
$result = array();
$td = $row->childNodes;
$result["date"] = $td->item(2)->nodeValue;
$result["image"] = $td->item(4)->firstChild->attributes->getNamedItem("src")->nodeValue;
$result["team1"] = $td->item(10)->nodeValue;
$result["team2"] = $td->item(12)->nodeValue;
$result["1"] = $td->item(14)->nodeValue;
$result["x"] = $td->item(16)->nodeValue;
$result["2"] = $td->item(18)->nodeValue;
$result["1%"] = $td->item(20)->nodeValue;
$result["x%"] = $td->item(22)->nodeValue;
$result["2%"] = $td->item(24)->nodeValue;
$result["total"] = $td->item(26)->nodeValue;
$results[] = $result;
}
For the image, you have to do same more proccesing, because you do not want the actual text, but the src attribute of the <img> element instead.