Check if you have the string and include its tags

Question

I would like to take the content of the site, remove only the text and insert my tags, but in this code I did, when he finds the text "Art" he does not leave the if, and then, only the first ones get the tag li, the rest all get the tag ul.

Could someone help me

    # Use the Curl extension to query Google and get back a page of results
    $url = "www.planalto.gov.br/ccivil_03/constituicao/constituicaocompilado.htm";
    $ch = curl_init();
    $timeout = 5;
    curl_setopt($ch, CURLOPT_URL, $url);
    curl_setopt($ch, CURLOPT_RETURNTRANSFER, 1);
    curl_setopt($ch, CURLOPT_CONNECTTIMEOUT, $timeout);
    $html = curl_exec($ch);
    curl_close($ch);

    # Create a DOM parser object
    $dom = new DOMDocument();

    # Parse the HTML from Google.
    # The @ before the method call suppresses any warnings that
    # loadHTML might throw because of invalid HTML in the page.
    @$dom->loadHTML($html);




    # Iterate over all the  tags
    foreach($dom->getElementsByTagName('font') as $link) {

    $mystring = $link->nodeValue;
    $findme   = 'Art';
    $pos = strpos($mystring, $findme);

        if ($pos === false) {

            echo "li";
            echo $link->nodeValue;
            echo "/li";

        } else { 

            echo "/ul";
            echo "ul id='' class='artigo'";
            echo "li";
            echo $link->nodeValue;
            echo "/li";

        }
    }

for the end result to be so

    _ul id="titulo1" class="titulo">
        _h3>TÍTULO I_/h3>
        _p>Dos Princípios Fundamentais_/p>
    _/ul>
    _ul id="titulo1_artigo1" class="artigo">
        _li>
            _ul class="caput">
                _li>
                    Art. 1º A República ... tem como fundamentos:
                _/li>
            _/ul>
        _/li>
        _li>
            _ul class="incisos">
                 _li> I - a soberania;_/li>
                 _li> II - a cidadania_/li>
                 _li> III - o pluralismo político._/li>
            _/ul>
        _/li>
        _li>
            _ul class="paragrafos">
                _li>Parágrafo único. Todo o ... desta Constituição.
                _/li>
            _/ul>
        _/li>

    _/ul>
    _ul id="titulo1_artigo2" class="artigo">
        _li>
            _ul class="caput">
                _li>
                    Art. 2º São Poderes da União, independentes e harmônicos entre si, o Legislativo, o Executivo e o Judiciário.
                _/li>
            _/ul>
        _/li>   
    _/ul>

Answer 1

Testing your code better, I noticed that the text repeats a few times, this is because of the use of getElementsByTagName which takes the parent element and the child element and the loop presents both with nodeValue, the texts will always repeat. I thought about using Xpath, but the whole problem occurs because this specific HTML document does not have division per block for each content, it simply works with line breaking.

It may be possible to use Xpath or something like that, but apparently it’s quite laborious.

So thinking of line breaks, I thought the following, instead of reading it as DOM, you can read it as text, line by line and detect where the article starts and ends.

To read line by line recommend to use tmpfile(), foef() and fgets. The tmpfile() will serve to store the page you are downloading.

//Gravar a página em um arquivo temporário
$handle = tmpfile();
fwrite($handle, $html);
fseek($handle, 0);

$html = NULL;

$initiate = false;
$inTitle = false;

//Função usada para remover elementos das linhas desnecessários
function removeTags($data) {
    $data = trim($data);
    $data = preg_replace('/[<][^>]+[>]|[<][^<>]+$|^[^<>]+[>]/', '', $data);
    return trim($data);
}

//No while iremos verificar linha a linha
while (false === feof($handle)) {
    $buffer = fgets($handle);//Lê a linha

    //Se a linha é vazio ignora e vai para a proxima linha
    if (trim($buffer) === '') {
        continue;
    }

    //Detecta aonde começa o artigo
    $findme = strpos($buffer, '>Art.') !== false;

    //Detecta um "possivel" termino do artigo ou titulo
    $endLine = stripos($buffer, '</p>') !== false;

    if ($findme) {

        //Se for já tiver ao menos um artigo adicionado ao corpo então isto detecta que terminou de listar os itens do artigo anterior
        if ($initiate) {
            echo '<hr>', PHP_EOL, PHP_EOL;
        }

        //Informa que encontrou ao menos um artigo
        $initiate = true;

        //Informa que estamos no titulo do artigo
        $inTitle = true;
        echo '<h1>', removeTags($buffer);
    } else if ($inTitle && $endLine) {
        //Se estiver no titulo e detectou um possivel fechamento do titulo
        $inTitle = false;
        echo removeTags($buffer), '</h1>', PHP_EOL;
    } else if ($initiate) {
        //Se não estiver dentro de um titulo ele imprime os dados
        $data = removeTags($buffer);

        //Se a linha for vazia então pula para a proxima linha
        if ($data === '') {
            continue;
        }

        echo $data, $inTitle ? '' : ('<br>' . PHP_EOL);
    }
}

//Fecha o arquivo temporario
fclose($temp);

Note that you can change the tmpfile for fopen and save formatted HTML so you don’t need to redo the search.

This code is just an example, so I did not make it do everything necessary, still it is necessary some more details, but the process is the same, just you work using the variables to detect where the article starts and ends for example.