Determine where licenses are coming from

I'm trying to understand the licensing summary produced by pyoxidizer, e.g.

Software Licensing Summary
==========================

71 distinct software components
0 lack a known software license
11 have unknown license expressions
16 distinct SPDX licenses
2 components in the public domain
4 have copyleft licenses

Count   OSI   FSF free   Copyleft   SPDX License
    2   [x]     [ ]        [ ]      BSD Zero Clause License
   28   [x]     [x]        [ ]      Apache License 2.0
    2   [x]     [x]        [ ]      BSD 3-Clause "New" or "Revised" License
    1   [ ]     [x]        [ ]      Creative Commons Zero v1.0 Universal
    1   [x]     [ ]        [ ]      CNRI Python License
    2   [x]     [x]        [x]      GNU General Public License v3.0 only
   39   [x]     [x]        [ ]      MIT License
    2   [x]     [x]        [x]      Mozilla Public License 2.0
    2   [ ]     [x]        [ ]      OpenSSL License
    3   [x]     [x]        [ ]      Python License 2.0
    1   [x]     [x]        [ ]      Sleepycat License
    1   [ ]     [ ]        [ ]      TCL/TK License
    5   [x]     [x]        [ ]      The Unlicense
    4   [ ]     [x]        [ ]      X11 License
    1   [x]     [x]        [ ]      zlib License
    1   [ ]     [ ]        [ ]      bzip2 and libbzip2 License v1.0.6

Specifically: given some license L, what are the components which use L?

The "Noteworthy Licensing Info" section produces some useful information, but it isn't exhaustive.

Version info:

PyOxidizer 0.22.0
commit: 8ce579ac27dd2b09efbb19e98103e9b6d052f337
source: https://github.com/indygreg/PyOxidizer.git
pyembed crate location: version = "0.22.0"

Start by locating the PyOxidizer licensing summary implementation and the code producing the “Noteworthy Licensing Info” section. Trace how components and their licenses are collected, then provide a way to list every component associated with a selected license and verify that the output is exhaustive.