rcviz

rcviz is a Python module to visualize recursion as a tree with arguments and return values at each node.

Features:

• Provides a decorator to instrument target functions (as opposed to trace or debugger based approaches)
• Uses Python module graphviz, provided by the graphviz software package.
• Supports IPython, particular inline viewing in Jupyter Notebooks
• Python 2 and Python 3 compatibility

Usage

1. Create an instance of the CallGraph class:

       cg = CallGraph(filename="factorial.pdf")

   The keyword argument filename is optional, and overrides the default filename. The default output
   filename for the callgraph is callgraph.png, but in the above example, we override it to factorial.pdf. The output file format is determined from the file extension provided in the filename (e.g. .pdf). If rendered inline inside a Jupyter (IPython) Notebook, the filename parameter is
   ignored.

2. Use the @viz(cg) decorator to instrument the recursive function. The decorator takes one required positional argument an instance of the CallGraph class, e.g. cg.

       @viz(cg)
       def factorial(n)
           ...

3. Render the recursion callgraph through an instance of a CallGraph class. Given the above example, such as cg in the above example. To render a graph to a file, call: cg.render(). If called within a Jupyter Notebook cell, the filename parameter (whether default or overridden) is ignored.

   The output file type is derived from the file name. Supported types include .dot (graphviz dot file), .png (png image), .svg (vector graphic)

   Example

   from rcviz import CallGraph, viz
   
   cg = CallGraph(filename="sort.pdf")
   
   @viz(cg)
   def quicksort(items):
       if len(items) <= 1:
           return items
       else:
           pivot = items[0]
           lesser = quicksort([x for x in items[1:] if x < pivot])
           greater = quicksort([x for x in items[1:] if x >= pivot])
           return lesser + [pivot] + greater
   
   print(quicksort(list("helloworld")))
   cg.render()

   Note: If executed inside a Jupyter Notebook cell, either cg.render() or cg will produce an inline callgraph, using IPython's _repr_svg_ method protocol.

Output

Note:

1. The edges are numbered by the order in which they were traversed by the execution.

2. The edges are colored from black to grey to indicate order of traversal : black edges first, grey edges last.

Experimental

Show intermediate values of local variables in the output render by invoking decoratedfunction.track(param1=val1, param2=val2,...). In the quicksort example above you can track the pivot with:

	pivot = items[0]
	quicksort.track(the_pivot=pivot) # shows a new row labelled the_pivot in each node 

Dependencies

This requires graphviz built with Python bindings to work.

On macOS, using Homebrew:

$ brew install --with-bindings graphviz

On Ubuntu:

# sudo apt-get install graphviz libgv-python

Tested on Python 2.7.3 and 3.7.6 Developed by Puneeth K, Rohan Mayya, Pavan Puranik