如何在字典内的列表中找到随机对象？

Question

I'm working on doing hangman and have simplified my code for the sake of the question.

Essentially I am trying to randomly select within a dictionary which is inputted by the user, a random object within whichever option the user chooses. For example if the user put a for the categ, and FIRST for the opt, randomOpt would be set to either a, b, or c.

categ = input('hello OR hi: ')
opt = input('FIRST, SECOND, OR LAST: ')

hello = {'FIRST':['a','b','c'],'SECOND':['z','x','y'],'LAST':['t','u','v']}

hi = {'FIRST':[1, 2, 3], 'SECOND':[20, 19, 18], 'LAST': [10, 11, 12]}

import random
randomOpt = random.choice(categ[opt])
print(randomOpt)

Whenever I run this, Python returns 'string indices must be integers' My friend suggested doing

categ = input('hello OR hi: ')
opt = input('FIRST, SECOND, OR LAST: ')

hello = {'FIRST':['a','b','c'],'SECOND':['z','x','y'],'LAST':['t','u','v']}

hi = {'FIRST':[1, 2, 3], 'SECOND':[20, 19, 18], 'LAST': [10, 11, 12]}

import random

if categ == 'hello':
   randomOpt = random.choice(hello[opt])
elif categ == 'hi':
   randomOpt = random.choice(hi[opt])

print(randomOpt)

But doing so feels a lot less 'dynamic' I suppose Could anybody help me figure out why this code doesn't work, and how I could edit to to fix it?

Answer 1

You want another dict, with hello and hi as keys.

stuff = {
    'hello': hello,
    'hi': hi
}

randomOpt = random.choice(stuff[categ][opt])

Answer 2

Just add another level of nestedness.

choices = {'hello': hello,
           'hi': hi
}

You can also use the .keys method of dictionaries to dynamically present the choices to the user.

So perhaps something like

user_choice = input('Choose one: ' + ' '.join(choices.keys())
choice = choices[user_choice]
options = ' '.join(choice.keys())
user_opt = input('Choose one: ' + options)
population = choice[user_opt]
print(random.choice(population))

When dealing with nested data, it can help to name variables as you traverse nested structures, IE nested = parent[key] , more_nested = nested[other_key] , etc. which is, to me, easier to reason about compared to obj[key][index][other_key]