WBS

Filtering Categories

1. Iterate over the root keys of the dictionary categories, where each key represents a category.
2. For every key, get its value object, which represent a list of the category's subcategories, i.e. the subcategories_list object.
3. For every subcategory:
   1. check if its value (a string) is present in a blacklist of categories, i.e. the category_blacklist.
   2. If a subcategory is present in category_blacklist:
      1. Remove it from the subcategories_list.
      2. Extract (i.e. pop) it's key-value pair from categories.
      3. For every subcategory in its subcategories_list, recursively extract it and its subcategories from categories.

Unfiltered Dictionary

{
  "category_a": ["category_b", "category_c"],
  "category_b": ["category_d"],
  "category_c": ["category_e"],
  "category_d": [],
  "category_e": []
}

Blacklist

{
  "blacklist": ["category_b", "category_e"]
}

Filtered Dictionary

{
  "category_a": ["category_c"],
  "category_c": [],
}

Notes

Retrieving All Subcategories without Filtering In place

If you intend to generate the full subcategory tree for a given domain (i.e. filter_in_place == False) for a depth greater than 2 (i.e. degree > 2), you could run into a RecursionError: maximum recursion depth exceeded.

To potentially solve this, you will have to increase the maximum depth of recursive calls in Python with sys.setrecursionlimit(limit), e.g.:

import sys
sys.setrecursionlimit(3000)

Also, as per here take into account that limit in sys.setrecursionlimit(limit: int) actually refers to the max stack depth and not really recursion depth.