Faster math typesetting in Jupyter and Python (Part 3)

Part 3 of my programming series about the Python package swiftex for writing math in Jupyter as fast as with pencil and paper. For the previous part, follow this link.

The plan for today: enable swiftex to tag and retrieve subexpressions from complicated expressions like fractions and square roots! At the end of the session, we can do the following in Jupyter.

At the end of this article, our Python package will allow Jupyter to do things like this

Fractions and Roots

First of all, let’s allow for defining fractions by giving the numerator and denominator as strings in the constructor:

Frac('a + b', 'c + d')

The corresponding test for setting this up reads

def test_simple_fraction_as_strings(self):
    actual = Frac("a + b", "c + d")
    self.assertEqual(r'\frac{a + b}{c + d}', actual)

As implementation, we define a new class inheriting from Symbol . Probably it will turn out to be convenient to have access to the numerator and denominator as symbols, so I store them as Symbol objects separately. Actually, I’m violating the TDD approach here when I implement something that’s not really needed yet, but hey, let’s not be dogmatic. ;-)

class Frac(Symbol):

    def __init__(self, numer, denom):
        self.numer = Symbol(numer)
        self.denom =…