CODEWITHSUNDEEP
pythondictionary
GreenRaccoon23
GreenRaccoon23

Reputation: 3833

Python - Add to a dictionary using a string

[Python 3.4.2]
I know this question sounds ridiculous, but I can't figure out where I'm messing up. I'm trying to add keys and values to a dictionary by using strings instead of quoted text. So instead of this, 

dict['key'] = value

this:

dict[key] = value

When I run the command above, I get this error: 

TypeError: 'str' object does not support item assignment

I think Python is thinking that I'm trying to create a string, not add to a dictionary. I'm guessing I'm using the wrong syntax. This is what I'm trying to do: 

dict[string_for_key][string_for_value] = string_for_deeper_value

I want this^ command to do this:

dict = {string_for_key: string_for_value: string_for_deeper_value}

I'm getting this error: 

TypeError: 'str' object does not support item assignment

I should probably give some more context. I'm:

  1. creating one dictionary
  2. creating a copy of it (because I need to edit the dictionary while iterating through it)
  3. iterating through the first dictionary while running some queries
  4. trying to assign a query's result as a value for each "key: value" in the dictionary.

Here's a picture to show what I mean:

key: value: query_as_new_value

-----EDIT-----

Sorry, I should have clarified: the dictionary's name is not actually 'dict'; I called it 'dict' in my question to show that it was a dictionary.

-----EDIT-----

I'll just post the whole process I'm writing in my script. The error occurs during the last command of the function. Commented out at the very bottom are some other things I've tried. 

from collections import defaultdict

global query_line, pericope_p, pericope_f, pericope_e, pericope_g


def _pre_query(self, typ):
    with open(self) as f:
        i = 1
        for line in f:
            if i == query_line:
                break
            i += 1
        target = repr(line.strip())
    ###skipping some code
    pericope_dict_post[self][typ] = line.strip()
    #^Outputs error TypeError: 'str' object does not support item assignment
    return


pericope_dict_pre = {'pericope-p.txt': 'pericope_p',
                     'pericope-f.txt': 'pericope_f',
                     'pericope-e.txt': 'pericope_e',
                     'pericope-g.txt': 'pericope_g'}
pericope_dict_post = defaultdict(dict)
#pericope_dict_post = defaultdict(list)
#pericope_dict_post = {}
for key, value in pericope_dict_pre.items():
    pericope_dict_post[key] = value
        #^Works
    #pericope_dict_post.update({key: value})
        #^Also works
    #pericope_dict_post.append(key)
        #^AttributeError: 'dict' object has no attribute 'append'
    #pericope_dict_post[key].append(value)
        #^AttributeError: 'dict' object has no attribute 'append'
    _pre_query(key, value)

-----FINAL EDIT-----

Matthias helped me figure it out, although acushner had the solution too. I was trying to make the dictionary three "levels" deep, but Python dictionaries cannot work this way. Instead, I needed to create a nested dictionary. To use an illustration, I was trying to do {key: value: value} when I needed to do {key: {key: value}}.

To apply this to my code, I need to create the [second] dictionary with all three strings at once. So instead of this:

my_dict[key] = value1
my_dict[key][value1] = value2

I need to do this:

my_dict[key][value1] = value2

Thanks a ton for all your help guys!

Upvotes: 8

Views: 15015

Answers (3)

Matthias
Matthias

Reputation: 13222

You could create a dictionary that expands by itself (Python 3 required).

class AutoTree(dict):
"""Dictionary with unlimited levels"""

     def __missing__(self, key):
        value = self[key] = type(self)()
        return value

Use it like this.

data = AutoTree()
data['a']['b'] = 'foo'
print(data)

Result

{'a': {'b': 'foo'}}

Now I'm going to explain your problem with the message TypeError: 'str' object does not support item assignment.

This code will work

from collections import defaultdict
data = defaultdict(dict)
data['a']['b'] = 'c'

data['a'] doesn't exist, so the default value dict is used. Now data['a'] is a dict and this dictionary gets a new value with the key 'b' and the value 'c'.

This code won't work

from collections import defaultdict
data = defaultdict(dict)
data['a'] = 'c'
data['a']['b'] = 'c'

The value of data['a'] is defined as the string 'c'. Now you can only perform string operations with data['a']. You can't use it as a dictionary now and that's why data['a']['b'] = 'c' fails.

Upvotes: 6

acushner
acushner

Reputation: 9946

first, do not use dict as your variable name as it shadows the built-in of the same name.

second, all you want is a nested dictionary, no?

from collections import defaultdict

d = defaultdict(dict)
d[string_for_key][string_for_value] = 'snth'

another way, as @Matthias suggested, is to create a bottomless dictionary:

dd = lambda: defaultdict(dd)
d = dd()
d[string_for_key][string_for_value] = 'snth'

Upvotes: 1

Hackaholic
Hackaholic

Reputation: 19743

you can do something like this: 

>>> my_dict = {}
>>> key = 'a'               # if key is not defined before it will raise NameError
>>> my_dict[key] = [1]
>>> my_dict[key].append(2)
>>> my_dict
{'a': [1, 2]}

Note: dict is inbuilt don't use it as variable name

Upvotes: 0

Related Questions