Parsing HTML files

Question

Hello,
I have a question about parsing HTML files. Usually when I get an HTML file and I need to find all the text in it I do this. This stuff just collects all of the hyperlinks and ignores all the html tags just keeping the actual text. It's fine for smaller files but occasionally I'll hit a large online text file and it will work but its way to slow for large files. I don't need to do all of this HTML tag stripping however for text files. Is there a way to still grab all the text without doing any tag searching to make it faster?
thanks,
private void find() throws IOException
        //Really slow for large text files.  Need a way to just use a regular scanner on an internet text file
        new ParserDelegator().parse(new InputStreamReader(myBase.openStream()),
                new ParserListener(),
                true); 
     * Inner class for processing all "<a href.."> tags when reading a base URL.
    private class ParserListener extends HTMLEditorKit.ParserCallback
        final String IGNORED_LINKS = "^(http|mailto|\\W).*";
        public void handleStartTag (HTML.Tag t, MutableAttributeSet a, int pos)
            if (t == HTML.Tag.A)
                String href = (String)(a.getAttribute(HTML.Attribute.HREF));
                //System.out.println(href);
                //System.out.println(href.matches(IGNORED_LINKS) + "\t" + href);
                if (! (href == null || href.matches(IGNORED_LINKS)) && !myURLs.contains(href))
                    myURLs.add(href);
            //TODO fix
            if (t == HTML.Tag.TITLE)
                String title = (String) (a.getAttribute(HTML.Attribute.TITLE));
                if (!(title == null))
                    myTitle = title;
                else myTitle = "No title was found";
        public void handleText (char[] data, int pos)
            myText.append(" ");
            myText.append(data);
    }

Answer 1

JFactor2004 wrote:
My question is. If I know an html file is actually just a txt fileThis isn't a question. HTML files are text by definition.
is it possible to look through it (maybe use something similar to a regular scanner) without doing anything with html.That depends on what you mean by "doing something with HTML". You can certainly read it one line at a time.