Add workflow algo to fish out sample thickness and use it to normalise. Re #3426

Code/Mantid/Framework/PythonAPI/PythonAlgorithms/WorkflowAlgorithms/NormaliseByThickness.py

+from MantidFramework import *
+from mantidsimple import *
+
+class NormaliseByThickness(PythonAlgorithm):
+    """
+        Normalise detector counts by the sample thickness
+    """
+    
+    def category(self):
+        return "Workflow\\SANS"
+
+    def name(self):
+        return "NormaliseByThickness"
+
+    def PyInit(self):
+        # Input workspace
+        self.declareProperty("InputWorkspace", "", Direction=Direction.Input, Description="Workspace to be normalised")
+        self.declareProperty("OutputWorkspace", "", Direction=Direction.Input, Description="Name of the workspace that will contain the transmission histogram")
+        self.declareProperty("SampleThickness", 1.0, Direction=Direction.Input, Description="Optional sample thickness value. If not provided the sample-thickness run propertywill be used.")
+        self.declareProperty("OutputMessage", "", Direction=Direction.Output)
+
+    def PyExec(self):
+        input_ws_name = self.getProperty("InputWorkspace")
+        if mtd[input_ws_name].getAxis(0).getUnit().name() != "Wavelength":
+            raise RuntimeError, "NormaliseByThickness expects an input workspace with units of Wavelength"        
+        
+        # Determine whether we should use the input thickess or try to read it from the run properties
+        thickness = self.getProperty("SampleThickness")
+        if thickness <= 0:
+            if mtd[input_ws_name].getRun().hasProperty("sample-thickness"):
+                thickness = mtd[input_ws_name].getRun().getProperty("sample-thickness").value
+                if thickness <= 0:
+                    raise RuntimeError, "NormaliseByThickness could not get the sample thickness"
+            else:
+                raise RuntimeError, "NormaliseByThickness could not get the sample thickness"
+
+        output_ws_name = self.getProperty("OutputWorkspace")                
+        self.executeSubAlg(Scale, input_ws_name, OutputWorkspace=output_ws_name, Factor=1.0/thickness, Operation="Multiply")
+        
+        self.setProperty("OutputMessage", "Normalised by thickness [%g cm]" % thickness)
+
+mtd.registerPyAlgorithm(NormaliseByThickness())