I have a dictionary - questions. the key is a number and the value is another dictionary. here is an example of the structure:
questions = {
2313: {"question": "How much is 2 2", "answers": ["3", "4", "2", "1"], "correct": 2},
4122: {"question": "What is the capital of France?", "answers": ["Lion", "Marseille", "Paris", "Montpellier"],
"correct": 3}
}
I need to add more questions to the dictionary from a text file ('questions.txt') where the questions look like that:
0#What is the capital city of USA?#Washington DC#New York#Los Angeles#Detroit#1
After I opened the file, I created a loop to go through all the questions and add them to the dictionary. I used a function from my protocol 'chatlib':
def split_data(data, expected_fields):
splitted = data.split('#')
if len(splitted) == expected_fields:
return splitted
else:
return
so for example when I use it on the question before it will return it as a list that looks like that:
['0', 'What is the capital city of USA?', 'Washington DC', 'New York', 'Los Angeles', 'Detroit', '1']
I tried a lot of different ways to write the main code,
list_new_questions = open("questions.txt").read().split('\n')
for question in list_new_questions:
questionlist = chatlib.split_data(question, 7)
key = int(questionlist[0])
questions[key] = {"question": "", "answers": [], "correct": 0}
questions[key]["question"] = questionlist[1]
questions[key]["answers"] = [questionlist[2], questionlist[3], questionlist[4], questionlist[5]]
questions[key]["correct"] = int(questionlist[6])
but every time it returns an error (TypeError: 'NoneType' object is not subscriptable) and says that the value of int(questionlist[0]) is None but I don't understand why. How can it be None, it is supposed to be the int value of the first element in the list questionlist which is always a number. every time I print int(questionlist[0]) it always prints a number so I don't understand why it says it's None.
CodePudding user response：
when you return out of line 3
questionlist = chatlib.split_data(question, 7) your function can return None if the len(splitted)` is not == to 7.
so questionlist = None when the there are not 7 fields and then in line 4 you are trying to take the index 0 of None.
easy fix & restructure would be something like:
def split_data(data):
# validate data here
splitted = data.split("#")
# validate splitted here if you need to
return splitted
list_new_questions = open("questions.txt").read().split("\n")
for question in list_new_questions:
questionlist = chatlib.split_data(question)
# do your validation here
if len(questionlist) == 7:
# try except maybe if you are unsure about the data split you get
key = int(questionlist[0])
questions[key] = {
"question": questionlist[1],
"answers": questionlist[2:6],
"correct": int(questionlist[6)
}
else:
print("Invalid question format")
CodePudding user response：
Let's look at the error message:
TypeError: 'NoneType' object is not subscriptable
Subscript notation is when you use square brackets like so: X[Y]. On the line that raises the exception, the only place where that happens is questionlist[0]. You cannot subscribe or index None because what would None[0] even mean? So that tells you that questionlist is None.
We can figure out why that is the case by looking at the previous line,
questionlist = chatlib.split_data(question, 7)
So what does split_data return?
Either return splitted, which is a list of string or return, which is the same as return None. So we know the latter has to be have happened, which means you've come across a line with an unexpected number of #s.
This could have happened because 1) there is a typo somewhere, because 2) you have an empty line in the file, or because 3) the file ends with a newline (automatically inserted by some editors).
The next question is, what do you want to happen in those cases? If you just want to ignore any non-matching line, Yash's solution will help you.
But I think while allowing 3 is definitely good and 2 is probably good for your use-case as well, I don't think you should want to just carry on like nothing's wrong in the case of 1. If there's a typo, I think you want your program to complain loudly... but maybe with a more useful error message.
You could do this without modifying split_data at all. Here's a possible solution:
list_new_questions = open("questions.txt").read().split('\n')
for question in list_new_questions:
if question.strip() == '':
# Ignore lines with only whitespace
continue
questionlist = chatlib.split_data(question, 7)
if questionlist is None:
# Not the right number of fields!
raise Exception(f"Invalid question format: {question!r}")
key, question, *answers, correct = questionlist
questions[int(key)] = {"question": question, "answers": answers, "correct": int(correct)}
CodePudding user response：
Your split_data function can return a None value if the actual fields is != expected_fields.
You can do it like below which can also account for if you start providing more answers.
This also removes the need for the split_data function as it can be done as part of the loop.
with open("questions.txt") as f:
new_questions = f.readlines()
for q in new_questions:
q = q.split("#")
if len(q) == 7:
questions[int(q[0])] = {
"question": q[1],
"answers": q[2:-1],
"correct": int(q[-1])}
else:
print(f"Invalid Data {q}")
If you wanted to provide more or less than 4 answers you could do something like below which only assumes the id is the first field, question is the second and answer is the last
for q in new_questions:
q = q.split("#")
try:
questions[int(q[0])] = {
"question": q[1],
"answers": q[2:-1],
"correct": int(q[-1])}
except IndexError:
print(f"Invalid data {q}")