Add semantics element as option for a CMML representaiton

Change-Id: I937a973585ff978df1653126dd4691f3ac4835ce

Author: physikerwelt

README.md

@@ -7,13 +7,16 @@ XQuery statements in DB2 dialect.
 The content MathML elements may be given as
 
 1. [MathWebSearch](http://search.mathweb.org/) expressions,
-2. annotation-xml elements with the attribute ```encoding="MathML-Content"``` or
-3. MathML tags that contain content MathML only
+2. annotation-xml elements with the attribute ```encoding="MathML-Content"```,
+3. semantic elements, or
+4. MathML tags that contain content MathML only
 
 ##Details 
 
 If a document contains multiple elements, only the first MathWebSearch expression is processed.
 
 If no MathWebSearch element is available, the first annotation-xml element will be used.
 
-If no annotation-xml Math element is available, the first math element will be used.
+If no annotation-xml Math element is available, the first semantic element will be used.
+
+...

src/main/java/de/tuberlin/dima/schubotz/mathmlquerygenerator/XQueryGenerator.java

@@ -65,11 +65,17 @@ public static Node getMainElement (Document xml) {
 				return node;
 			}
 		}
+        // if that fails too interprete content of first semantic element as content MathML
+        expr = xml.getElementsByTagName( "semantics" );
+        if ( expr.getLength() > 0 ) {
+            return new NdLst( expr ).item( 0 );
+        }
 		// if that fails too interprete content of root MathML element as content MathML
 		expr = xml.getElementsByTagName( "math" );
 		if ( expr.getLength() > 0 ) {
 			return new NdLst( expr ).item( 0 );
 		}
+
 		return null;
 	}
 
@@ -158,7 +164,7 @@ private String generateConstraint (Node node, boolean isRoot) {
 					}
 
 				} else if ( child.getNodeType() == Node.TEXT_NODE ) {
-					out = "./text() = '" + child.getNodeValue() + "'";
+					out = "./text() = '" + child.getNodeValue().trim() + "'";
 				}
 			}
 		}