CODEWITHSUNDEEP
pythonpython-2.7
Xandros
Xandros

Reputation: 727

Convert a list to a string and back

I have a virtual machine which reads instructions from tuples nested within a list like so:

[(0,4738),(0,36),
 (0,6376),(0,0)]

When storing this kind of machine code program, a text file is easiest, and has to be written as a string. Which is obviously quite hard to convert back.

Is there any module which can read a string into a list/store the list in a readable way?

requirements:

Upvotes: 28

Views: 40340

Answers (7)

Martijn Pieters
Martijn Pieters

Reputation: 1122022

Use the json module:

string = json.dumps(lst)
lst = json.loads(string)

Demo:

>>> import json
>>> lst = [(0,4738),(0,36),
...  (0,6376),(0,0)]
>>> string = json.dumps(lst)
>>> string
'[[0, 4738], [0, 36], [0, 6376], [0, 0]]'
>>> lst = json.loads(string)
>>> lst
[[0, 4738], [0, 36], [0, 6376], [0, 0]]

An alternative could be to use repr() and ast.literal_eval(); for just lists, tuples and integers that also allows you to round-trip:

>>> from ast import literal_eval
>>> string = repr(lst)
>>> string
'[[0, 4738], [0, 36], [0, 6376], [0, 0]]'
>>> lst = literal_eval(string)
>>> lst
[[0, 4738], [0, 36], [0, 6376], [0, 0]]

JSON has the added advantage that it is a standard format, with support from tools outside of Python support serialising, parsing and validation. The json library is also a lot faster than the ast.literal_eval() function.

Upvotes: 46

Slater Victoroff
Slater Victoroff

Reputation: 21914

JSON!

import json

with open(data_file, 'wb') as dump:
    dump.write(json.dumps(arbitrary_data))

and similarly:

source = open(data_file, 'rb').read()
data = json.loads(source)

Upvotes: 21

malohm
malohm

Reputation: 251

eval should do it in a simple way:

>>> str([(0,4738),(0,36),(0,6376),(0,0)])
'[(0, 4738), (0, 36), (0, 6376), (0, 0)]'

>>> eval(str([(0,4738),(0,36),(0,6376),(0,0)]))
[(0, 4738), (0, 36), (0, 6376), (0, 0)]

Upvotes: 20

Sean McSomething
Sean McSomething

Reputation: 6507

If you're just dealing with primitive Python types, you can just use the built-in repr():

Help on built-in function repr in module __builtin__:

repr(...)
    repr(object) -> string

    Return the canonical string representation of the object.
    For most object types, eval(repr(object)) == object.

Upvotes: 1

inspectorG4dget
inspectorG4dget

Reputation: 113965

with open('path/to/file', 'w') as outfile:
     for tup in L:
         outfile.write("%s\n" %' '.join(str(i) for i in tup))

with open('path/to/file) as infile:
    L = [tuple(int(i) for i in line.strip().split()) for line in infile]

Upvotes: 0

NPE
NPE

Reputation: 500367

If these are just two-tuples, you could store them in a CVS file using the csv module. No need for any brackets/parentheses.

Upvotes: 0

Sukrit Kalra
Sukrit Kalra

Reputation: 34493

Just use ast.literal_eval

>>> from ast import literal_eval
>>> a = literal_eval('[(1, 2)]')
>>> a
[(1, 2)]

You can convert it into a string using repr().

>>> repr(a)
'[(1, 2)]'

Upvotes: 27

Related Questions