Java Regular Expression for detecting class/interface/etc declaration

Question

I\'m trying to create a regular expression that detect a new class for example:

public interface IGame {

or

private class Game         


        

Answer 1

This regex is an incomplete specification of Java class and interface declaration. However, it can match declarations like this:

abstract class X>extends java.util.ArrayListimplements java.util.Queue,Serializable{}

public@Deprecated interface K extends Comparable{}

So it should be sufficient for most purpose.

Please read the code to generate the regex at the end to know exactly what was skipped. In summary, ReferenceType, Annotation are not fully incorporated into this regex since they would require recursive regex to parse properly.

The Java regex to match Java class declaration in all its g(l)ory details:

(?:((?:public|protected|private|abstract|static|final|strictfp|@[ \t\f\r\n]*+\p{javaJavaIdentifierStart}\p{javaJavaIdentifierPart}*+)(?:[ \t\f\r\n]*+(?:(?)?+(?:[ \t\f\r\n]*+[.][ \t\f\r\n]*+\p{javaJavaIdentifierStart}\p{javaJavaIdentifierPart}*+(?:[ \t\f\r\n]*+<[ \t\f\r\n]*+(?:\p{javaJavaIdentifierStart}\p{javaJavaIdentifierPart}*+|[?][ \t\f\r\n]*+(?:(?:extends|super)[ \t\f\r\n]++\p{javaJavaIdentifierStart}\p{javaJavaIdentifierPart}*+)?+)(?:[ \t\f\r\n]*+,[ \t\f\r\n]*+(?:\p{javaJavaIdentifierStart}\p{javaJavaIdentifierPart}*+|[?][ \t\f\r\n]*+(?:(?:extends|super)[ \t\f\r\n]++\p{javaJavaIdentifierStart}\p{javaJavaIdentifierPart}*+)?+))*+[ \t\f\r\n]*+>)?+)*+|\p{javaJavaIdentifierStart}\p{javaJavaIdentifierPart}*+))?+(?:[ \t\f\r\n]*+,[ \t\f\r\n]*+\p{javaJavaIdentifierStart}\p{javaJavaIdentifierPart}*+(?:[ \t\f\r\n]++extends[ \t\f\r\n]++(?:\p{javaJavaIdentifierStart}\p{javaJavaIdentifierPart}*+(?:[ \t\f\r\n]*+[.][ \t\f\r\n]*+\p{javaJavaIdentifierStart}\p{javaJavaIdentifierPart}*+)*+(?:[ \t\f\r\n]*+<[ \t\f\r\n]*+(?:\p{javaJavaIdentifierStart}\p{javaJavaIdentifierPart}*+|[?][ \t\f\r\n]*+(?:(?:extends|super)[ \t\f\r\n]++\p{javaJavaIdentifierStart}\p{javaJavaIdentifierPart}*+)?+)(?:[ \t\f\r\n]*+,[ \t\f\r\n]*+(?:\p{javaJavaIdentifierStart}\p{javaJavaIdentifierPart}*+|[?][ \t\f\r\n]*+(?:(?:extends|super)[ \t\f\r\n]++\p{javaJavaIdentifierStart}\p{javaJavaIdentifierPart}*+)?+))*+[ \t\f\r\n]*+>)?+(?:[ \t\f\r\n]*+[.][ \t\f\r\n]*+\p{javaJavaIdentifierStart}\p{javaJavaIdentifierPart}*+(?:[ \t\f\r\n]*+<[ \t\f\r\n]*+(?:\p{javaJavaIdentifierStart}\p{javaJavaIdentifierPart}*+|[?][ \t\f\r\n]*+(?:(?:extends|super)[ \t\f\r\n]++\p{javaJavaIdentifierStart}\p{javaJavaIdentifierPart}*+)?+)(?:[ \t\f\r\n]*+,[ \t\f\r\n]*+(?:\p{javaJavaIdentifierStart}\p{javaJavaIdentifierPart}*+|[?][ \t\f\r\n]*+(?:(?:extends|super)[ \t\f\r\n]++\p{javaJavaIdentifierStart}\p{javaJavaIdentifierPart}*+)?+))*+[ \t\f\r\n]*+>)?+)*+|\p{javaJavaIdentifierStart}\p{javaJavaIdentifierPart}*+))?+)*+[ \t\f\r\n]*+>)[ \t\f\r\n]*+)?+(?:(?)?+(?:[ \t\f\r\n]*+[.][ \t\f\r\n]*+\p{javaJavaIdentifierStart}\p{javaJavaIdentifierPart}*+(?:[ \t\f\r\n]*+<[ \t\f\r\n]*+(?:\p{javaJavaIdentifierStart}\p{javaJavaIdentifierPart}*+|[?][ \t\f\r\n]*+(?:(?:extends|super)[ \t\f\r\n]++\p{javaJavaIdentifierStart}\p{javaJavaIdentifierPart}*+)?+)(?:[ \t\f\r\n]*+,[ \t\f\r\n]*+(?:\p{javaJavaIdentifierStart}\p{javaJavaIdentifierPart}*+|[?][ \t\f\r\n]*+(?:(?:extends|super)[ \t\f\r\n]++\p{javaJavaIdentifierStart}\p{javaJavaIdentifierPart}*+)?+))*+[ \t\f\r\n]*+>)?+)*+)[ \t\f\r\n]*+)?+(?:(?)?+(?:[ \t\f\r\n]*+[.][ \t\f\r\n]*+\p{javaJavaIdentifierStart}\p{javaJavaIdentifierPart}*+(?:[ \t\f\r\n]*+<[ \t\f\r\n]*+(?:\p{javaJavaIdentifierStart}\p{javaJavaIdentifierPart}*+|[?][ \t\f\r\n]*+(?:(?:extends|super)[ \t\f\r\n]++\p{javaJavaIdentifierStart}\p{javaJavaIdentifierPart}*+)?+)(?:[ \t\f\r\n]*+,[ \t\f\r\n]*+(?:\p{javaJavaIdentifierStart}\p{javaJavaIdentifierPart}*+|[?][ \t\f\r\n]*+(?:(?:extends|super)[ \t\f\r\n]++\p{javaJavaIdentifierStart}\p{javaJavaIdentifierPart}*+)?+))*+[ \t\f\r\n]*+>)?+)*+(?:[ \t\f\r\n]*+,[ \t\f\r\n]*+\p{javaJavaIdentifierStart}\p{javaJavaIdentifierPart}*+(?:[ \t\f\r\n]*+[.][ \t\f\r\n]*+\p{javaJavaIdentifierStart}\p{javaJavaIdentifierPart}*+)*+(?:[ \t\f\r\n]*+<[ \t\f\r\n]*+(?:\p{javaJavaIdentifierStart}\p{javaJavaIdentifierPart}*+|[?][ \t\f\r\n]*+(?:(?:extends|super)[ \t\f\r\n]++\p{javaJavaIdentifierStart}\p{javaJavaIdentifierPart}*+)?+)(?:[ \t\f\r\n]*+,[ \t\f\r\n]*+(?:\p{javaJavaIdentifierStart}\p{javaJavaIdentifierPart}*+|[?][ \t\f\r\n]*+(?:(?:extends|super)[ \t\f\r\n]++\p{javaJavaIdentifierStart}\p{javaJavaIdentifierPart}*+)?+))*+[ \t\f\r\n]*+>)?+(?:[ \t\f\r\n]*+[.][ \t\f\r\n]*+\p{javaJavaIdentifierStart}\p{javaJavaIdentifierPart}*+(?:[ \t\f\r\n]*+<[ \t\f\r\n]*+(?:\p{javaJavaIdentifierStart}\p{javaJavaIdentifierPart}*+|[?][ \t\f\r\n]*+(?:(?:extends|super)[ \t\f\r\n]++\p{javaJavaIdentifierStart}\p{javaJavaIdentifierPart}*+)?+)(?:[ \t\f\r\n]*+,[ \t\f\r\n]*+(?:\p{javaJavaIdentifierStart}\p{javaJavaIdentifierPart}*+|[?][ \t\f\r\n]*+(?:(?:extends|super)[ \t\f\r\n]++\p{javaJavaIdentifierStart}\p{javaJavaIdentifierPart}*+)?+))*+[ \t\f\r\n]*+>)?+)*+)*+[ \t\f\r\n]*+))?+[{]

And for Java interface declaration:

(?:((?:public|protected|private|abstract|static|strictfp|@[ \t\f\r\n]*+\p{javaJavaIdentifierStart}\p{javaJavaIdentifierPart}*+)(?:[ \t\f\r\n]*+(?:(?)?+(?:[ \t\f\r\n]*+[.][ \t\f\r\n]*+\p{javaJavaIdentifierStart}\p{javaJavaIdentifierPart}*+(?:[ \t\f\r\n]*+<[ \t\f\r\n]*+(?:\p{javaJavaIdentifierStart}\p{javaJavaIdentifierPart}*+|[?][ \t\f\r\n]*+(?:(?:extends|super)[ \t\f\r\n]++\p{javaJavaIdentifierStart}\p{javaJavaIdentifierPart}*+)?+)(?:[ \t\f\r\n]*+,[ \t\f\r\n]*+(?:\p{javaJavaIdentifierStart}\p{javaJavaIdentifierPart}*+|[?][ \t\f\r\n]*+(?:(?:extends|super)[ \t\f\r\n]++\p{javaJavaIdentifierStart}\p{javaJavaIdentifierPart}*+)?+))*+[ \t\f\r\n]*+>)?+)*+|\p{javaJavaIdentifierStart}\p{javaJavaIdentifierPart}*+))?+(?:[ \t\f\r\n]*+,[ \t\f\r\n]*+\p{javaJavaIdentifierStart}\p{javaJavaIdentifierPart}*+(?:[ \t\f\r\n]++extends[ \t\f\r\n]++(?:\p{javaJavaIdentifierStart}\p{javaJavaIdentifierPart}*+(?:[ \t\f\r\n]*+[.][ \t\f\r\n]*+\p{javaJavaIdentifierStart}\p{javaJavaIdentifierPart}*+)*+(?:[ \t\f\r\n]*+<[ \t\f\r\n]*+(?:\p{javaJavaIdentifierStart}\p{javaJavaIdentifierPart}*+|[?][ \t\f\r\n]*+(?:(?:extends|super)[ \t\f\r\n]++\p{javaJavaIdentifierStart}\p{javaJavaIdentifierPart}*+)?+)(?:[ \t\f\r\n]*+,[ \t\f\r\n]*+(?:\p{javaJavaIdentifierStart}\p{javaJavaIdentifierPart}*+|[?][ \t\f\r\n]*+(?:(?:extends|super)[ \t\f\r\n]++\p{javaJavaIdentifierStart}\p{javaJavaIdentifierPart}*+)?+))*+[ \t\f\r\n]*+>)?+(?:[ \t\f\r\n]*+[.][ \t\f\r\n]*+\p{javaJavaIdentifierStart}\p{javaJavaIdentifierPart}*+(?:[ \t\f\r\n]*+<[ \t\f\r\n]*+(?:\p{javaJavaIdentifierStart}\p{javaJavaIdentifierPart}*+|[?][ \t\f\r\n]*+(?:(?:extends|super)[ \t\f\r\n]++\p{javaJavaIdentifierStart}\p{javaJavaIdentifierPart}*+)?+)(?:[ \t\f\r\n]*+,[ \t\f\r\n]*+(?:\p{javaJavaIdentifierStart}\p{javaJavaIdentifierPart}*+|[?][ \t\f\r\n]*+(?:(?:extends|super)[ \t\f\r\n]++\p{javaJavaIdentifierStart}\p{javaJavaIdentifierPart}*+)?+))*+[ \t\f\r\n]*+>)?+)*+|\p{javaJavaIdentifierStart}\p{javaJavaIdentifierPart}*+))?+)*+[ \t\f\r\n]*+>)[ \t\f\r\n]*+)?+(?:(?)?+(?:[ \t\f\r\n]*+[.][ \t\f\r\n]*+\p{javaJavaIdentifierStart}\p{javaJavaIdentifierPart}*+(?:[ \t\f\r\n]*+<[ \t\f\r\n]*+(?:\p{javaJavaIdentifierStart}\p{javaJavaIdentifierPart}*+|[?][ \t\f\r\n]*+(?:(?:extends|super)[ \t\f\r\n]++\p{javaJavaIdentifierStart}\p{javaJavaIdentifierPart}*+)?+)(?:[ \t\f\r\n]*+,[ \t\f\r\n]*+(?:\p{javaJavaIdentifierStart}\p{javaJavaIdentifierPart}*+|[?][ \t\f\r\n]*+(?:(?:extends|super)[ \t\f\r\n]++\p{javaJavaIdentifierStart}\p{javaJavaIdentifierPart}*+)?+))*+[ \t\f\r\n]*+>)?+)*+(?:[ \t\f\r\n]*+,[ \t\f\r\n]*+\p{javaJavaIdentifierStart}\p{javaJavaIdentifierPart}*+(?:[ \t\f\r\n]*+[.][ \t\f\r\n]*+\p{javaJavaIdentifierStart}\p{javaJavaIdentifierPart}*+)*+(?:[ \t\f\r\n]*+<[ \t\f\r\n]*+(?:\p{javaJavaIdentifierStart}\p{javaJavaIdentifierPart}*+|[?][ \t\f\r\n]*+(?:(?:extends|super)[ \t\f\r\n]++\p{javaJavaIdentifierStart}\p{javaJavaIdentifierPart}*+)?+)(?:[ \t\f\r\n]*+,[ \t\f\r\n]*+(?:\p{javaJavaIdentifierStart}\p{javaJavaIdentifierPart}*+|[?][ \t\f\r\n]*+(?:(?:extends|super)[ \t\f\r\n]++\p{javaJavaIdentifierStart}\p{javaJavaIdentifierPart}*+)?+))*+[ \t\f\r\n]*+>)?+(?:[ \t\f\r\n]*+[.][ \t\f\r\n]*+\p{javaJavaIdentifierStart}\p{javaJavaIdentifierPart}*+(?:[ \t\f\r\n]*+<[ \t\f\r\n]*+(?:\p{javaJavaIdentifierStart}\p{javaJavaIdentifierPart}*+|[?][ \t\f\r\n]*+(?:(?:extends|super)[ \t\f\r\n]++\p{javaJavaIdentifierStart}\p{javaJavaIdentifierPart}*+)?+)(?:[ \t\f\r\n]*+,[ \t\f\r\n]*+(?:\p{javaJavaIdentifierStart}\p{javaJavaIdentifierPart}*+|[?][ \t\f\r\n]*+(?:(?:extends|super)[ \t\f\r\n]++\p{javaJavaIdentifierStart}\p{javaJavaIdentifierPart}*+)?+))*+[ \t\f\r\n]*+>)?+)*+)*+)[ \t\f\r\n]*+)?+[{]

It is generated by building up the pattern piece by piece. I reference Java Language Specification for Java SE 7 for the grammar of NormalClassDeclaration and NormalInterfaceDeclaration.

Here is the code to generate the monster above, including comments on which pattern is skipped:

// All rules never end with WhiteSpace. This allows us to check for error easier.

final static String Identifier = "\\p{javaJavaIdentifierStart}\\p{javaJavaIdentifierPart}*+";
// Heuristic to exclude any position where adding space would split an Identifier or a keyword apart
final static String TokenBoundary = "(?:(?";

final static String TypeName = Identifier + "(?:" + WhiteSpaceOpt + "[.]" + WhiteSpaceOpt + Identifier + ")*+";
// Expanded definition of ClassOrInterfaceType = ClassType = InterfaceType
//     ClassOrInterfaceType -> TypeName  TypeArguments
//     ClassOrInterfaceType -> ClassOrInterfaceType . Identifier TypeArguments
final static String ClassType = TypeName + "(?:" + WhiteSpaceOpt + TypeArguments + ")?+" + 
    "(?:" + WhiteSpaceOpt + "[.]" + WhiteSpaceOpt + Identifier + "(?:" + WhiteSpaceOpt + TypeArguments + ")?+" + ")*+";
// Definition of ClassType and InterfaceType are identical
final static String InterfaceType = ClassType;
final static String ClassOrInterfaceType = ClassType;

final static String TypeBound = "extends" + WhiteSpaceCom + "(?:" + ClassOrInterfaceType + "|" + TypeVariable + ")";
final static String TypeParameter = TypeVariable + "(?:" + WhiteSpaceCom + TypeBound + ")?+";
final static String TypeParameterList = TypeParameter + "(?:" + WhiteSpaceOpt + "," + WhiteSpaceOpt + TypeParameter + ")*+";
final static String TypeParameters = "<" + WhiteSpaceOpt + TypeParameterList + WhiteSpaceOpt + ">";

final static String Super = "extends" + WhiteSpaceCom + ClassType;
    
final static String InterfaceTypeList = InterfaceType + "(?:" + WhiteSpaceOpt  + "," + WhiteSpaceOpt + InterfaceType + ")*+";
final static String Interfaces = "implements" + WhiteSpaceCom + InterfaceTypeList;

final static String NormalClassDeclaration =
    // Annotation in its fullest form can end in ), so WhiteSpaceOpt is used here for pedantic
    // It can be changed to WhiteSpaceCom to save the TokenBoundary check, since current definition always end with Identifier character
    "(?:" + "(" + ClassModifiers + ")" + WhiteSpaceOpt + ")?+" +
    TokenBoundary + "class" + WhiteSpaceCom +
    // WhiteSpaceOpt is used here, since TypeParameters starts with < and no whitespace is needed to delimit
    "(" + Identifier + ")" + WhiteSpaceOpt + 
    "(?:" + "(" + TypeParameters + ")" + WhiteSpaceOpt + ")?+" +
    // As the result, we need to check for boundary before "extends" and "implements"
    TokenBoundary + "(?:" + "(" + Super + ")"+ WhiteSpaceOpt + ")?+" +
    TokenBoundary + "(?:" + "(" + Interfaces + WhiteSpaceOpt + ")" + ")?+[{]";
    // ClassBody is skipped, and only opening bracket { is matched

final static String InterfaceModifier = "(?:public|protected|private|abstract|static|strictfp|" + Annotation + ")";
// Same as ClassModifiers, except that "final" is no longer a valid modifier
final static String InterfaceModifiers = InterfaceModifier + "(?:" + WhiteSpaceOpt + TokenBoundary + InterfaceModifier + ")*+";

final static String ExtendsInterfaces = "extends" + WhiteSpaceCom + InterfaceTypeList;

final static String NormalInterfaceDeclaration = 
    "(?:" + "(" + InterfaceModifiers + ")" + WhiteSpaceOpt + ")?+" +
    TokenBoundary + "interface" + WhiteSpaceCom +
    // WhiteSpaceOpt is used here, since TypeParameters starts with < and no whitespace is needed to delimit
    "(" + Identifier + ")" + WhiteSpaceOpt + 
    "(?:" + "(" + TypeParameters + ")" + WhiteSpaceOpt + ")?+" +
    // As the result, we need to check for boundary before "extends" here
    TokenBoundary + "(?:" + "(" + ExtendsInterfaces + ")" + WhiteSpaceOpt + ")?+[{]";

^{I am well aware CamelCase and also the fact that some constants differ by pluralization are inappropriate, but it provides a better mapping to the grammar defined in the JLS.}

Here is the demo on ideone.