逗号分隔的csv文件中的python dict

Question

Python dict is in a format like this:

'{"a":1, "b":2, "c":3}'

Notice it use comma to separate different key:value pairs.

The problem is I have a CSV file, which is separate columns by comma too :

'
  "id",   "gender",   "age",    "name"
 "001",     "male",    "14",    "{"first":"Mike", "last":"Green"}"
 "002",   "female",    "15",    "{"first":"Kate", "last":"Spear"}"
'

When I do
pandas.read_csv('csvfile.csv', sep = ',', names=["id", "gender", "age", "name"])

I got:

'
  "id",   "gender",   "age",    "name"
 "001",     "male",    "14",    "{"first":"Mike"
 "002",   "female",    "15",    "{"first":"Kate"
'

The reason I guess is csv reader regards the comma follows first name in dict as a separator in csv files. Since I only specified 4 columns named " "id", "gender", "age", "name"", so it ignore last names.

Any thoughts or possible solution to this? Thanks!

Answer 1

You can change the delimiter that read_csv uses. If you can change the csv files to use a semicolon for separating columns, you can then use read_csv(file.csv, sep=';'...)

Alternatively you can fix the quoting from

"001",     "male",    "14",    "{"first":"Mike", "last":"Green"}"

to

"001",     "male",    "14",    "{'first':'Mike', 'last':'Green'}"

Of course both methods mean editing the csv file.

The second looks sounder. The regular expression (\\{[^"]*)(")([^}]*\\}) could be used to match quotes inside braces. (untested)