Styjun

I want to convert a csv file to json file, it's work on my computer but not on my centos server

My python version on my computer is python 3.7.2, and the csv is rightly convert to json locally.

So i've tried to update my python on Centos, my current version is python 2.6.6

#Read CSV File
def read_csv(file, json_file, format):
 csv_rows = []
 with open(file) as csvfile:
 reader = csv.DictReader(csvfile)
 title = reader.fieldnames
 for row in reader:
 csv_rows.extend([title[i]:row[title[i]] for i in range(len(title))])
 write_json(csv_rows, json_file, format)

I expect that my function can read my csv file but in place, python return my this error :

csv_rows.extend([title[i]:row[title[i]] for i in range(len(title))])
 ^
SyntaxError: invalid syntax

Obviously this error doesn't exist on my computer, only on this CentOS server

Do you think i've must update my python 2 to python 3 ?

The code below was tested under python 3.6 and 2.7

(I believe this is what you are looking for)

reader = [['1','2','3'],['11','22','33'],['111','222','333']]
csv_rows = [] 
title = ['T1','T2','T3']
for row in reader:
 csv_rows.extend([ title[i]:row[i] for i in range(len(title)) ])

print(csv_rows)

Output

['T2': '2', 'T3': '3', 'T1': '1', 'T2': '22', 'T3': '33', 'T1': '11', 'T2': '222', 'T3': '333', 'T1': '111']

Little optimization and readability tip: avoid index selection if possible.

If you are iterating on both the index and the value of an iterable, avoid using range(len(x)), and use directly enumerate, is much faster in python and more readable.

Improving @balderman's answer:

reader = [['1','2','3'],['11','22','33'],['111','222','333']]
csv_rows = []
title = ['T1','T2','T3']
for row in reader:
 csv_rows.extend([val: row[idx] for idx, val in enumerate(title)])
print(csv_rows)

Even better and clearer is to avoid using indexes at all:

reader = [['1','2','3'],['11','22','33'],['111','222','333']]
csv_rows = []
title = ['T1','T2','T3']
for row in reader:
 csv_rows.extend([ t_val: r_val for (t_val, r_val) in zip(title, row) ])
print(csv_rows)

Even more compact (and faster):

reader = [['1','2','3'],['11','22','33'],['111','222','333']]
title = ['T1','T2','T3']

csv_rows = [t_val: r_val for (t_val, r_val) in zip(title, row) for row in reader]
print(csv_rows)

List and dict-comprehensions are a bit tricky at the beginning, but if you think of them as "for loops with the for placed after" they are not so weird.