Faster glossary generation

Question

I am trying to make a table of contents for my queryset in Django like this:

def get_toc(self):
    toc = {}
    qs = self.get_queryset()
    idx = set()
    for q in qs:
        idx.add(q.title[0])
    idx = list(idx)
    idx.sort()
    for i in idx:
        toc[i] = []
        for q in qs:
            if q.title[0] == i:
                toc[i].append(q)
    return toc

But it has time complexity O(n^2). Is there a better way to do it?

UPDATE I meant glossary, not table of contents.

Answer 1

This doesn't look like a table of contents, but a glossary, where you map the first character of a term to a list of terms.

We can work with .groupby(…) ^[python-doc] here:

from itertools import groupby

result = {
    k: list(vs)
    for k, vs in groupby(
        self.get_queryset().order_by('title'), lambda x: x.title[0]
    )
}