See README for instructions on how to run the software.
This is first and foremost a visualization project, i.e. a project focused on providing a visualization of how sorting algorithms work for students and other interested parties. Because of this, you not only have to implement a given algorithm, but also decide on how to illustrate key concepts.
For inspiration on the currently available sorting algorithms, take a look at the Algorithms directory. GnomeSort is a simple example of a sorting algorithm, with simple visualization due to the nature of the algorithm. Quicksort is a more involved visualization; showing the splits performed in running the algorithm.
Algorithm implementations must subclass from SortAlgorithm. The items you wish to sort are available in self.items, however, you should rarely need to access them directly. You will also need to add the algorithm to the algorithm list in SortDisplay.py.
All comparisons must use self.cmp, which is an instance of Comparator and has the most common comparison operations available. Examples: If you want to say self.items[i] > 0, instead say self.cmp.gt(self.items[i], 0). If you want to say self.items[i] >= self.items[j], instead say self.cmp.gteI(i, j). See the Comparator module for documentation of the available comparison operations.
To draw colored lines for illustration, use self.markers, which is an instance of Markers. See the module for documentation.
When creating a marker, you can either decide to change the color of an element (above set to False; for instance used for highlighting the current element you're looking at), or draw a separator line above an element (above set to True; for instance used to show that the problem has been split into smaller sub-problems). The return value of addMarker is needed for later manipulation of the marker, so keep track of that.
If you wish to swap element i and j, please use self.items.swap(i, j).
The algorithms are implemeted as iterators, as this allows the program to step through the algorithm at it's own pace, allowing for the simulation to be slowed down. This is done by having the sorting function yield whenever a logical step has occurred in the algorithm. See the other algorithms for examples.