Orange DM: How does Orange calculates the euclidean distance for categorical data?

Question

I have been using Orange for a while now and am trying to implement some classification methods to classify my data.

I am using the Test and Score widget to test a few classification methods (kNN,ANN, etc.) on my dataset and am using k-fold cross validation.

My dataset is the nursery dataset from the UCI Repository available here which only contains categorical attributes.

Here, is my question: How does the kNN Algorithm calculate the Euclidean distance? I thought that for categorical datasets the "hamming-distance" should be used. However, when I use the hamming distance, which I have to program manually in the python widget:

import Orange
knn = Orange.classification.KNNLearner(n_neighbors=9, metric="hamming")
out_learner=knn

the results with both distances are always the same.