Python: Create Dictionary from Text/File that's in Dictionary Format
Question:
I’d like to create a dictionary from a text file that I have, who’s contents are in a ‘dictionary’ format. Here’s a sample of what the file contains:
{‘fawn’: [1], ‘sermersheim’: [3], ‘sonji’: [2], ‘scheuring’: [2]}
It’s exactly this except it contains 125,000 entries. I am able to read in the text file using read(), but it creates a variable of the literal text of the file even when I initialize the variable with
dict = {}
Answers:
This looks like json to me. Use the json module if so.
This looks like yaml to me. Use the pyyaml module if so. (As suggested by @jcollado).
You can’t use the json module because it is strict about its input.
You can use the eval built-in. For example, this would work if each dictionary entry is on a different line:
dicts_from_file = []
with open('myfile.txt','r') as inf:
for line in inf:
dicts_from_file.append(eval(line))
# dicts_from_file now contains the dictionaries created from the text file
Alternatively, if the file is just one big dictionary (even on multiple lines), you can do this:
with open('myfile.txt','r') as inf:
dict_from_file = eval(inf.read())
This is probably the most simple way to do it, but it’s not the safest. As others mentioned in their answers, eval has some inherent security risks. The alternative, as mentioned by JBernardo, is to use ast.literal_eval which is much safer than eval since it will only evaluate strings which contain literals. You can simply replace all the calls to eval in the above examples with ast.literal_eval after importing the ast module.
If you’re using Python 2.4 you are not going to have the ast module, and you’re not going to have with statements. The code will look more like this:
inf = open('myfile.txt','r')
dict_from_file = eval(inf.read())
inf.close()
Don’t forget to call inf.close(). The beauty of with statements is they do it for you, even if the code block in the with statement raises an exception.
I highly discourage using eval though. It may result in security issues if you don’t have full control on the input file. Just import your dictionary and save them by using the json or pickle module.
Using eval might be dangerous. If json doesn’t work, then I’d recommend using yaml which seems to work fine with your example input:
>>> import yaml
>>> yaml.load("{'fawn': [1], 'sermersheim': [3], 'sonji': [2], 'scheuring': [2]}")
{'fawn': [1], 'scheuring': [2], 'sermersheim': [3], 'sonji': [2]}
It’s not a production ready solution and may not work well with a file of your size, but if you need a simple way and can prepend you file to
my_dict = {'fawn': [1], 'sermersheim': [3], 'sonji': [2], 'scheuring': [2]}
then you can rename it to a python file and simply import
from my_file import my_dict
I’d like to create a dictionary from a text file that I have, who’s contents are in a ‘dictionary’ format. Here’s a sample of what the file contains:
{‘fawn’: [1], ‘sermersheim’: [3], ‘sonji’: [2], ‘scheuring’: [2]}
It’s exactly this except it contains 125,000 entries. I am able to read in the text file using read(), but it creates a variable of the literal text of the file even when I initialize the variable with
dict = {}
This looks like json to me. Use the json module if so.
This looks like yaml to me. Use the pyyaml module if so. (As suggested by @jcollado).
You can’t use the json module because it is strict about its input.
You can use the eval built-in. For example, this would work if each dictionary entry is on a different line:
dicts_from_file = []
with open('myfile.txt','r') as inf:
for line in inf:
dicts_from_file.append(eval(line))
# dicts_from_file now contains the dictionaries created from the text file
Alternatively, if the file is just one big dictionary (even on multiple lines), you can do this:
with open('myfile.txt','r') as inf:
dict_from_file = eval(inf.read())
This is probably the most simple way to do it, but it’s not the safest. As others mentioned in their answers, eval has some inherent security risks. The alternative, as mentioned by JBernardo, is to use ast.literal_eval which is much safer than eval since it will only evaluate strings which contain literals. You can simply replace all the calls to eval in the above examples with ast.literal_eval after importing the ast module.
If you’re using Python 2.4 you are not going to have the ast module, and you’re not going to have with statements. The code will look more like this:
inf = open('myfile.txt','r')
dict_from_file = eval(inf.read())
inf.close()
Don’t forget to call inf.close(). The beauty of with statements is they do it for you, even if the code block in the with statement raises an exception.
I highly discourage using eval though. It may result in security issues if you don’t have full control on the input file. Just import your dictionary and save them by using the json or pickle module.
Using eval might be dangerous. If json doesn’t work, then I’d recommend using yaml which seems to work fine with your example input:
>>> import yaml
>>> yaml.load("{'fawn': [1], 'sermersheim': [3], 'sonji': [2], 'scheuring': [2]}")
{'fawn': [1], 'scheuring': [2], 'sermersheim': [3], 'sonji': [2]}
It’s not a production ready solution and may not work well with a file of your size, but if you need a simple way and can prepend you file to
my_dict = {'fawn': [1], 'sermersheim': [3], 'sonji': [2], 'scheuring': [2]}
then you can rename it to a python file and simply import
from my_file import my_dict