regexps refactored

8 files changed, 80 insertions, 7 deletions

diff --git a/sources/scala/util/regexp/Alphabet.scala b/sources/scala/util/regexp/Alphabet.scala
new file mode 100644
index 0000000000..0ae119e8e5
--- /dev/null
+++ b/sources/scala/util/regexp/Alphabet.scala
@@ -0,0 +1,3 @@
+package scala.util.regexp ;
+
+trait Alphabet ;
diff --git a/sources/scala/util/regexp/Base.scala b/sources/scala/util/regexp/Base.scala
new file mode 100644
index 0000000000..9eb93f4659
--- /dev/null
+++ b/sources/scala/util/regexp/Base.scala
@@ -0,0 +1,31 @@
+package scala.util.regexp ;
+
+/** basic regular expressions */
+
+trait Base {
+
+  type regexp <: RegExp;
+
+  abstract class RegExp;
+
+  /** Alt( R,R,R* ) */
+  case class  Alt(rs: regexp*)  extends RegExp {
+
+    // check rs \in R,R,R*
+
+    if({ var i = 0;
+        val it = rs.elements;
+        while( it.hasNext ) {it.next; i=i+1};
+        i} < 2)
+      throw new SyntaxErrorExpression("need at least 2 branches in alt")
+  }
+  case class  Sequ(rs: regexp*) extends RegExp ;
+  case class  Star(r: regexp)   extends RegExp ;
+  case object Eps               extends RegExp {
+    override def toString() = "Eps";
+  }
+
+  /** this class can be used to add meta information to regexps */
+  class Meta( r:regexp )       extends RegExp ;
+
+}
diff --git a/sources/scala/util/regexp/PointedHedgeExp.scala b/sources/scala/util/regexp/PointedHedgeExp.scala
new file mode 100644
index 0000000000..b3cccf2198
--- /dev/null
+++ b/sources/scala/util/regexp/PointedHedgeExp.scala
@@ -0,0 +1,15 @@
+package scala.util.regexp ;
+
+/** pointed regular hedge expressions, a useful subclass of full rhe */
+trait PointedHedgeExp[ A <: Alphabet ] extends Base {
+
+  type label = A;
+
+  type regexp <: RegExp;
+
+  case class  Node(label: A, r: regexp)       extends RegExp ;
+  case class  TopIter(r1: regexp, r2: regexp) extends RegExp ;
+
+  case object Point extends RegExp ;
+
+}
diff --git a/sources/scala/util/regexp/SyntaxErrorException.scala b/sources/scala/util/regexp/SyntaxErrorException.scala
new file mode 100644
index 0000000000..e1ada9daca
--- /dev/null
+++ b/sources/scala/util/regexp/SyntaxErrorException.scala
@@ -0,0 +1,6 @@
+package scala.util.regexp ;
+
+/** this runtime exception is thrown if an attempt to instantiate a
+  * syntactically incorrect expression is detected */
+class SyntaxErrorExpression(e: String)
+  extends java.lang.RuntimeException(e);
diff --git a/sources/scala/util/regexp/WordExp.scala b/sources/scala/util/regexp/WordExp.scala
new file mode 100644
index 0000000000..1151b8b11f
--- /dev/null
+++ b/sources/scala/util/regexp/WordExp.scala
@@ -0,0 +1,17 @@
+package scala.util.regexp ;
+
+/** regular word expressions. use them with an alphabet: <pre>
+abstract class IntLabels extends Alphabet;
+object IntWordExp extends WordExp[IntLabels] {
+  type regexp = RegExp;
+};
+ * </pre>
+ */
+
+trait WordExp[ A <: Alphabet ] extends Base {
+
+  type label = A;
+
+  case class Letter(a: label)    extends RegExp;
+
+}
diff --git a/sources/scala/xml/dtd/Decl.scala b/sources/scala/xml/dtd/Decl.scala
index dd0a71f435..246ee4a5a0 100644
--- a/sources/scala/xml/dtd/Decl.scala
+++ b/sources/scala/xml/dtd/Decl.scala
@@ -21,9 +21,9 @@ case class ElemDecl( name:String ,
                      attribs:Map[String,AttrDecl] )
      extends MarkupDecl {
 
-       final val parsedContentModel:RegExp = {
+       final val parsedContentModel:ContentModel.RegExp = {
          try {
-           RegExp.parse( contentModel );
+           ContentModel.parse( contentModel );
          } catch {
            case _:Error =>
              Console.println( "error parsing declaration of " + name );
diff --git a/sources/scala/xml/dtd/Parser.scala b/sources/scala/xml/dtd/Parser.scala
index b0e9e369ec..410e82ce06 100644
--- a/sources/scala/xml/dtd/Parser.scala
+++ b/sources/scala/xml/dtd/Parser.scala
@@ -3,6 +3,7 @@ package scala.xml.dtd ;
 /** Parser for regexps (content models in DTD element declarations) */
 
 object Parser with Scanner { // a bit too permissive concerning #PCDATA
+  import ContentModel._ ;
 
   /** parses the argument to a regexp */
   def parse(s:String):RegExp = { initScanner( s ); contentspec }
@@ -61,7 +62,7 @@ object Parser with Scanner { // a bit too permissive concerning #PCDATA
       PCDATA_
     else {
       val t = choiceRest( PCDATA_ );
-      if( !t.mixed )
+      if( !isMixed( t ) )
         error("mixed content models must be like (#PCDATA.|.|.|.)*");
       accept( RPAREN );
       // lax: (workaround for buggy Java XML parser in JDK1.4.2)
@@ -115,14 +116,14 @@ object Parser with Scanner { // a bit too permissive concerning #PCDATA
     //Console.println("particle, token="+token2string(token));
     token match {
     case LPAREN => nextToken; sOpt; regexp;
-    case NAME   => val a = RNode( value ); nextToken; maybeSuffix( a )
+    case NAME   => val a = Letter(ElemName(value)); nextToken; maybeSuffix( a )
     case _      => error("expected '(' or Name, got:"+token2string( token ));
     }
   }
 
   //                                     atom ::= name
   def atom = token match {
-    case NAME   => val a = RNode( value ); nextToken; a
+    case NAME   => val a = Letter(ElemName(value)); nextToken; a
     case _      => error("expected Name, got:"+token2string( token ));
   }
 }
diff --git a/sources/scala/xml/dtd/Validation.scala b/sources/scala/xml/dtd/Validation.scala
index 6073b5aaa8..0055b0844c 100644
--- a/sources/scala/xml/dtd/Validation.scala
+++ b/sources/scala/xml/dtd/Validation.scala
@@ -3,7 +3,7 @@ package scala.xml.dtd ;
 import scala.collection.mutable;
 import scala.collection.Set;
 import scala.collection.Map;
-
+import ContentModel._ ;
 
 class ElementValidator(namespace$$: String, elemSpec: RegExp) {
   final val namespace = namespace$$.intern();
@@ -56,7 +56,7 @@ class ElementValidator(namespace$$: String, elemSpec: RegExp) {
     case Star(z:Alt) if(( z.rs.length == 1 )&& (z.rs(0) match {
       case PCDATA_   => true
       case _         => false
-    })) => new MixedModeAutom( elemSpec.getLabels );
+    })) => new MixedModeAutom(getLabels( elemSpec ));
     case _    => TrivialAutom;
   }
   def validate( nodes:Seq[Node] ):Unit = autom.run( nodes );