11 replies to this topic

[chili5]

I have an input file that looks like this:

Quote

<a>sample link</a>
<a rel="" href="http://dwite.ca/">link with rel</a>
<a href="http://compsci.ca/" rel="nofollow">link with no follow</a>
<a href="http://compsci.ca/blog" rel="external">more rels</a>
<a href="http://compsci.ca/v3/viewforum.php?f=131" title="">link</a>

I'm trying to extract from the String the substring rel="" and anything that could possibly be in the quotes inside rel.

My current code is:

/*

 * To change this template, choose Tools | Templates

 * and open the template in the editor.

 */


package dwite;


import java.io.*;

import java.util.*;

import java.util.regex.*;

/**

 *

 * @author brocj1112

 */

public class links {

    public static void main(String[] args) throws IOException {

        //Code goes here

        Scanner fin = new Scanner(new FileReader("IN/links.in"));

        String sHTML;

        String sLink;

        String sRel = "";

        


        while (fin.hasNextLine()) {

            sHTML = fin.nextLine();


            // get just the link part

            sLink = sHTML.substring(sHTML.indexOf("<a"),sHTML.indexOf("</a>")+4);


            System.out.println(sLink);

            if (!sLink.contains("rel=")) {

                // rel is not found

                // so add it to the end of the link

                sLink = sLink.substring(0,sLink.indexOf(">")) + " rel=\"nofollow\">"

                        + sLink.substring(sLink.indexOf(">")+1);

            } else if (sLink.contains("rel=\"\""))  {

                sLink = sLink.replace(" rel=\"\"", "");

                sLink = sLink.substring(0,sLink.indexOf(">")) + " rel=\"nofollow\">"

                        + sLink.substring(sLink.indexOf(">")+1);

            } else if (!sLink.contains("nofollow")) {

                // rel already exists so add nofollow to the end of the rel

                sRel = sLink.substring(sLink.indexOf("rel=\""),sLink.indexOf("rel=\"", sLink.indexOf("rel=\"")));

            }

            

        }

        

        fin.close();

    }

}

The code that is trying to get the rel="" portion is:

sLink.substring(sLink.indexOf("rel=\""),sLink.indexOf("rel=\"", sLink.indexOf("rel=\"")));

However this code only returns:

rel=

and nothing else. Any ideas as to how to accomplish this?

[John]

There are a few ideas that come to mind, but your best bet is probably using regular expressions.

[chili5]

Is using regular expressions with Java, the same as in PHP?

[John]

There is a 95% chance the answer is yes.

[WingedPanther]

Pretty much. Java includes a regex library that is very robust.

[chili5]

Thanks guys! :)

[Skel]

if(sLink.contains("rel=")) {

    String[] temp = sLink.split(" ");

    String rel = temp[1];

}

Personally I'd do it like that.

[chili5]

No you can't do that. It won't work. Since rel can appear anywhere in the string.

[Turk4n]

Skel said:

if(sLink.contains("rel=")) {

    String[] temp = sLink.split(" ");

    String rel = temp[1];

}

Personally I'd do it like that.

I personally think you did it wrong on purpose !

[chili5]

Skel, there is no chance at all that your code would work. If my link was:

<a href="#">test</a>

Your code would return "<a". Which is way wrong!

I'm looking into how I can do this using the regular expressions but I'm almost certain my regexp will look something like this:

String sPattern = "rel=\"[\s\w]{1,}+\"";

Haven't tried it yet but I'm thinking it will look a little like that.

[chili5]

Thanks everyone, I got it. :D

String sRegex = "rel=\"[a-zA-Z\\.]{1,}+\"";

// ...

else if (!sLink.contains("nofollow")) {
                // rel already exists so add nofollow to the end of the rel
                mMatch = p.matcher(sLink);
                
                while (mMatch.find()) {
                    sRel = sLink.substring(mMatch.start(), mMatch.end());
                    sRel = sRel.substring(0,sRel.length()-1) + " nofollow\"";
                }
                sLink = sLink.replaceAll(sRegex, sRel);
            }

:)

[Turk4n]

chili5 said:

Thanks everyone, I got it. :D

String sRegex = "rel=\"[a-zA-Z\\.]{1,}+\"";


// ...


else if (!sLink.contains("nofollow")) {

                // rel already exists so add nofollow to the end of the rel

                mMatch = p.matcher(sLink);

                

                while (mMatch.find()) {

                    sRel = sLink.substring(mMatch.start(), mMatch.end());

                    sRel = sRel.substring(0,sRel.length()-1) + " nofollow\"";

                }

                sLink = sLink.replaceAll(sRegex, sRel);

            }

:)

Great to hear :)