Compare strings to accents, UTF-8

Question

Something is escaping me here, I am doing Curl to a weather page, if the results have accents and comparing with exactly the same string in full this condition returns false (not equal). This is merely for testing:

function get_page($url) {

   $curl = curl_init();
   curl_setopt($curl, CURLOPT_RETURNTRANSFER, True);
   curl_setopt($curl, CURLOPT_URL, $url);
   /*curl_setopt($curl, CURLOPT_TIMEOUT_MS, 1000);*/
   $return = curl_exec($curl);
   curl_close($curl);
  return $return;

}

$weather = get_page("http://www.accuweather.com/pt/pt/cascais/274007/weather-forecast/274007");

preg_match('/<span class="cond">(.*?)<\/span>/s', $weather, $cond);
preg_match('/<strong class="temp">(.*?)<span>/s', $weather, $temp);
$condition = trim($cond[1]); //Céu Limpo (hoje)
$temp = trim($temp[1]); //27 (hoje)

In today’s case (30-06-2015) the condition we have is "Clear Sky", but when I test the following condition:

if(strtolower($condition) == "céu limpo") {
   ....
}

False returns (if commands are not executed)

But if it does:

$hey = "Céu Limpo";
if(strtolower($hey) == "céu limpo") {
   ....
}

Already returns true and the code within the condition is already executed. I would like to know why of this and how to solve

Answer 1

Your problem is related to Html Entities, if you do it here:

$arrayCondition = str_split($condition);
$arrayString = str_split("Céu Limpo");
var_dump($arrayCondition);
var_dump($arrayString);

You’ll notice the difference in their exit:

array(14) { [0]=> string(1) "C" [1]=> string(1) "&" [2]=> string(1) "#" [3]=> string(1) "2" [4]=> string(1) "3" [5]=> string(1) "3" [6]=> string(1) ";" [7]=> string(1) "u" [8]=> string(1) " [9]=> string(1) "L" [10]=> string(1) "i" [11]=> string(1) "m" [12]=> string(1) "p" [13]=> string(1) "o" }

array(10) { [0]=> string(1) "C" [1]=> string(1) " [2]=> string(1) " [3]=> string(1) "u" [4]=> string(1) " [5]=> string(1) "L" [6]=> string(1) "i" [7]=> string(1) "m" [8]=> string(1) "p" [9]=> string(1) "o" }

The first that comes from your Curl is coming with Htmlentities, your handwriting is is coming with the value é

To solve this you can use the html_entity_decode, example:

if (strtolower(html_entity_decode($condition)) == "céu limpo") {
    echo 'funcionou!!!';
}

Answer 2

The problem occurs because it is probably not in UTF-8 the CURL answer.

For this try to convert to utf-8 before making the comparison.

Functions that can be used: utf8_encode or utf8_decode

Example:

if(strtolower(utf8_encode($condition)) == "céu limpo") {
....
}
//Caso não funciona tente ao contrario as vezes seu arquivo não esta em utf-8

if(strtolower(utf8_decode($condition)) == "céu limpo") {
....
}