Question

I have the following bit of HTML

<div class="article">this is a div article content</div>

which is being “tagged” by an HTML-agnostic program on the words div, class and article, resulting in:

<<hl>div</hl> <hl>class</hl>="<hl>article</hl>">this is a <hl>div</hl> <hl>article</hl> content</<hl>div</hl>>

although what I really need is:

<div class="article">this is a <hl>div</hl> <hl>article</hl> content</div>

Since the output is utter garbage (even tools like HTML Tidy choke on it), I figured a regex replace would help strip out the extra <hl>s inside the HTML tag:

replace(/<([^>]*)<hl>([^<]*?)<\/hl>([^>]*?)>/g, '<$1$2$3>')

Now, this works but only replaces the first occurrence in the tag, that is, the div:

<div <hl>class</hl>="<hl>article</hl>">this is a <hl>div</hl> <hl>article</hl> content</div>

My question is: how do I replace all <hl>s inside the tag, so as to make sure the HTML remains valid?

Additional notes:

• I don’t need the tag attributes at all (i.e. class="article" can disappear)
• I can change <hl> and </hl> for any other strings
• Yes, the output comes from Solr

UPDATE: I accepted jcollado’s answer, but I needed this in Javascript. This is the equivalent code:

var stripIllegalTags = function(html) {

  var output = '',
    dropChar,
    parsingTag = false;

  for (var i=0; i < html.length; i++) {
    var character = html[i];

    if (character == '<') {
      if (parsingTag) {
        do {
          dropChar = html[i+1];
          i++;
        } while (dropChar != '>');
        continue;
      }
      parsingTag = true;
    } else if (character == '>') {
      parsingTag = false;
    }

    output += character;

  }

  return output;

}

Answer 1

Maybe the piece of code below is helpful for you:

class HTMLCleaner(object):
    def parse(self, html):
        output = []
        parsing_tag = False

        html = iter(html)
        for char in html:
            if char == '<':
                if parsing_tag:
                    drop_char = html.next()
                    while drop_char != '>':
                        drop_char = html.next()
                    continue
                parsing_tag = True
            elif char == '>':
                parsing_tag = False

            output.append(char)

        return ''.join(output)

html = '<<hl>div</hl> <hl>class</hl>="<hl>article</hl>">this is a <hl>div</hl> <hl>article</hl> content</<hl>div</hl>>'

parser = HTMLCleaner()
print parser.parse(html)

The output for the given input is:

<div class="article">this is a <hl>div</hl> <hl>article</hl> content</div>

which I believe is what you’re looking for.

The code basically drops all tags when another tag hasn’t been parsed yet.