Building a dictionary from a tab delimited file in a pythonic way

Question

I've written the following function that takes a tab delimited file (as a string) and turns it into a dictionary with an integer as a key and a list of two floats and the value:

def parseResults(self, results):
    """
    Build a dictionary of the SKU (as key), current UK price and current Euro price
    """
    lines = results.split('\n')
    individual_results = []
    for i in range(1,len(lines)-1):
        individual_results.append(lines[i].split('\t'))
    results_dictionary = {}
    for i in range(len(individual_results)):
        results_dictionary[int(individual_results[i][0])] = [float(individual_results[i][1]), float(individual_results[i][2])]
    return results_dictionary

I've been reading about using list comprehension and also dictionary comprehension but I don't really know what the best way to build this is.

I think I can simplify the first list build using:

individual_results = [results.split('\t') for results in lines[1:]]

but I don't then know the best way to create the dictionary. I've got the feeling this might be possible in a neat way without even creating the intermediate list.

Thanks,

Matt

Answer 1

Like this:

import csv
import StringIO
results = "sku\tdelivered-price-gbp\tdelivered-price-euro\tid\n32850238\t15.53\t35.38\t258505\n"

data = list(csv.DictReader(StringIO.StringIO(results), delimiter='\t'))
print(data)

Output:

[{'sku': '32850238', 'delivered-price-euro': '35.38', 'delivered-price-gbp': '15.53', 'id': '258505'}]

Of course, if you can read from an actual file, you can skip the stringIO part.

To build the type of dictionary you want, you would do this:

data = {}
for entry in csv.DictReader(StringIO.StringIO(results), delimiter='\t'):
    data[entry['sku']] = [entry['delivered-price-gbp'], entry['delivered-price-euro']]

Or even as a dictionary comprehension:

import csv
import StringIO
results = "sku\tdelivered-price-gbp\tdelivered-price-euro\tid\n32850238\t15.53\t35.38\t258505\n10395850\t35.21\t46.32\t3240582\n"

data = {entry['sku']: [entry['delivered-price-gbp'], entry['delivered-price-euro']] 
        for entry in csv.DictReader(StringIO.StringIO(results), delimiter='\t')}
print(data)

But that's now getting highly difficult to read.

The output would in those two last cases be:

{'32850238': ['15.53', '35.38'], '10395850': ['35.21', '46.32']}

Answer 2

使用标准库中的CSV模块，它具有直接读取字典csv.DictReader的方法

Answer 3

Try something like this:

In [8]: from collections import defaultdict

In [9]: with open("filename") as f:
   ...:     dic=defaultdict(list)
   ...:     next(f)                #skip the first line 
   ...:     for line in f:
   ...:         k,v=line.split(None,1)
   ...:         dic[int(k)].extend( map(float,v.split()[:2]) )
   ...:         

In [10]: dic
Out[10]: defaultdict(<type 'list'>, {32850238: [15.53, 35.38]})

Answer 4

Your code can simply be:

def parseResults(self, results):  
    lines = results.split('\n')
    li_results = [x.split('\t') for x in lines]
    results_dict = {int(x[0]):map(float,[x[1],x[2]]) for x in li_results[1:]} # skip the header
    return results_dict

or if you want it shorter(not recommended):

def parseResults(self, results):
    return {int(x[0]):map(float,[x[1],x[2]]) for x in [i.split('\t') for i in results.split('\n')][1:]}

Output(from the string you have given):

{32850238: [15.53, 35.38]}