Question

I have some problems with replaceregexp. To be more understandable, i’ll use simple examples, and not the real code.

So presume I’ve 2 java files:

firstClass.java

/**
 * UNDESIRABLE COMMENTS ABCDEFGHIJ ABCDEFGHIJ ABCDEFGHIJ ABCDEFGHIJ ABCDEFGHIJ
 * ABCDEFGHIJABCDEFGHIJABCDEFGHIJABCDEFGHIJABCDEFGHIJABCDEFGHIJ
 */
public class firstClass{
    public String firstFunction(){
         <instructions>
    }

    /**
    * My function's description that I want to keep
    */
    public String secondFunction(){
         <instructions>
    }
}

secondClass.java

/**
 * UNDESIRABLE COMMENTS ABCDEFGHIJ ABCDEFGHIJ ABCDEFGHIJ ABCDEFGHIJ ABCDEFGHIJ
 * ABCDEFGHIJABCDEFGHIJABCDEFGHIJABCDEFGHIJABCDEFGHIJABCDEFGHIJ
 */
public class secondClass{
    public String firstFunction(){
         <instructions>
    }
}

So I want to remove the “UNDESIRABLE COMMENTS” part of each java file in my project, but I want to do it with ANT, not manually. For that, i’m using this:

<replaceregexp flags="s">
    <regexp pattern="UNDESIRABLE COMMENTS.+\*/"/>
    <substitution expression="*/"/>
    <fileset dir=".">
        <filename name="*.java"/>
    </fileset>
 </replaceregexp>

What I want is to get this code:

firstClass.java

/**
 */
public class firstClass{
    public String firstFunction(){
         <instructions>
    }

    /**
    * My function's description that I want to keep
    */
    public String secondFunction(){
         <instructions>
    }
}

secondClass.java

/**
 */
public class secondClass{
    public String firstFunction(){
         <instructions>
    }
}

It works very well for secondClass.java but not for the first file. The problem appears when there are other comments, like functions’ description. The file secondClass.java results well but in firstClass.java I get this:

/**
    */
    public String secondFunction(){
         <instructions>
    }
}

Everything what was between the beginning of “UNDESIRABLE COMMENTS” to the end of my second function’s description was deleted.

So the problem is: when my replaceregexp finds more */, it stops on the last one, and delete everything between “UNDESIRABLE COMMENTS” and this last */.

Do you have any solution to just delete what is between “UNDESIRABLE COMMENTS” and the first */ found?

Answer 1

Your problem is that that + quantifier in regular expressions is greedy, and will consume as much of the input as possible while still matching. Try

UNDESIRABLE COMMENTS.+?\*/

instead – +? means the same as + but is reluctant i.e. it will match as little as possible while still producing an overall match of the whole expression.