How to split each individual value between two string in Python

Question

variable=";CREATEDBY~string~1~~72~0~0~0~~~0;CREATEDBYNAME~string~1~~800~0~0~0~~~1;CREATEDBYYOMINAME~string~1~~800~0~0~0~~~2;CREATEDON~date~1~yyyy-MM-dd HH:mm:ss.SSS~26~0~0~0~~~3;CREATEDONUTC~date~1~yyyy-MM-dd HH:mm:ss.SSS~26~0~0~0~~~4"

How can I split each value which is between (; and ~)?

The result will be like CREATEDBY,CREATEDBYNAME,CREATEDBYYOMINAME,...

I have tried the below, but it's giving the first occurrence.

variable[variable.find(";")+1:myString.find("~")]

How do I get the list of strings by using the split?

use split function . mystring.split(';') then mystirng.split('~') — Hassan ALi, Dec 18 '18 at 08:37
Possible duplicate of [How to split a string into a list?](https://stackoverflow.com/questions/743806/how-to-split-a-string-into-a-list) Note that the second answer indicates how to specify the delimiter. — jpmc26, Dec 18 '18 at 18:14

score 21 · Accepted Answer · answered Dec 18 '18 at 08:39

Using str.split

Ex:

variable=";CREATEDBY~string~1~~72~0~0~0~~~0;CREATEDBYNAME~string~1~~800~0~0~0~~~1;CREATEDBYYOMINAME~string~1~~800~0~0~0~~~2;CREATEDON~date~1~yyyy-MM-dd HH:mm:ss.SSS~26~0~0~0~~~3;CREATEDONUTC~date~1~yyyy-MM-dd HH:mm:ss.SSS~26~0~0~0~~~4"

for i in variable.strip(";").split(";"):
    print(i.split("~", 1)[0])
#or
print([i.split("~", 1)[0] for i in variable.strip(";").split(";")])

Output:

CREATEDBY
CREATEDBYNAME
CREATEDBYYOMINAME
CREATEDON
CREATEDONUTC

['CREATEDBY', 'CREATEDBYNAME', 'CREATEDBYYOMINAME', 'CREATEDON', 'CREATEDONUTC']

Tim Biegeleisen · Answer 2 · 2018-12-18T09:33:34.967

14

We can try using re.findall with the pattern ;(\w+)(?=~):

variable = ";CREATEDBY~string~1~~72~0~0~0~~~0;CREATEDBYNAME~string~1~~800~0~0~0~~~1;CREATEDBYYOMINAME~string~1~~800~0~0~0~~~2;CREATEDON~date~1~yyyy-MM-dd HH:mm:ss.SSS~26~0~0~0~~~3;CREATEDONUTC~date~1~yyyy-MM-dd HH:mm:ss.SSS~26~0~0~0~~~4"
result = re.findall(r';(\w+)~', variable)
print(result)

['CREATEDBY', 'CREATEDBYNAME', 'CREATEDBYYOMINAME', 'CREATEDON', 'CREATEDONUTC']

edited Dec 18 '18 at 09:33

answered Dec 18 '18 at 08:38

Tim Biegeleisen

502,043
27
286
360

It's rejecting if the COLUMN name has '_' (i.e CREATED_BY) – Vicky Dec 18 '18 at 09:11
2

@Vicky Then use `\w+`, which includes underscores, to match your words. – Tim Biegeleisen Dec 18 '18 at 09:21
@Vicky: good observation. That's why I hate Regex. You never know what they are really doing and what their intension was. It's very easy to get them wrong accidentally. – Thomas Weller Dec 18 '18 at 10:26
1

@Thomas I'm not sure that's a reason to hate Regex. `[A-Z]` should be self-explanatory that it isn't going to match underscores, and the question could have been clearer that underscores should be matched. – grg Dec 18 '18 at 17:10
@ThomasWeller while I'm sure you can find [degenerate cases](https://www.regextester.com/94502) that I would totally agree with you on, the regex in question is so utterly trivial that any experienced programmer should be able to immediately understand what it does, even if their not familiar with python's particular flavor of regex. And as for getting it wrong, blaming the choice of tool for code not satisfying an incomplete spec is something I just can't get behind. – Jared Smith Dec 18 '18 at 17:15
1

@JaredSmith: trivial or not, I can't judge. The spec is not so bad regarding the underscore: between `;` and `~`, it did not say that there should be a word (`\w`). So a `[^~]` would be closer to the spec. Next, I wonder why he says the pattern should be `;(\w+)(?=~)` but then uses `;(\w+)~` in the code. – Thomas Weller Dec 18 '18 at 18:51

score 5 · Answer 3 · answered Dec 18 '18 at 08:39

You can split() the string and then find() the first ~ for each one:

variable=";CREATEDBY~string~1~~72~0~0~0~~~0;CREATEDBYNAME~string~1~~800~0~0~0~~~1;CREATEDBYYOMINAME~string~1~~800~0~0~0~~~2;CREATEDON~date~1~yyyy-MM-dd HH:mm:ss.SSS~26~0~0~0~~~3;CREATEDONUTC~date~1~yyyy-MM-dd HH:mm:ss.SSS~26~0~0~0~~~4"
result = [item[:item.find('~')]  for item in  variable.split(';')]

print(result)

score 5 · Answer 4 · answered Dec 18 '18 at 08:44

5

Use regular expression with lookahead and lookbehind:

>>> import re
>>> re.findall(r'(?<=;).*?(?=~)', variable)
['CREATEDBY', 'CREATEDBYNAME', 'CREATEDBYYOMINAME', 'CREATEDON', 'CREATEDONUTC']

answered Dec 18 '18 at 08:44

Martin Frodl

667
4
11

Posted the same answer 5 minutes ago. You don't even need lookarounds if you place the name in a capture group. – Tim Biegeleisen Dec 18 '18 at 08:46
3

The two answers are fairly similar, yes, but I wouldn't call them the same. – Martin Frodl Dec 18 '18 at 08:54

Raffi · Answer 5 · 2018-12-18T19:04:28.513

1

import re

variable = ";CREATEDBY~string~1~~72~0~0~0~~~0;CREATEDBYNAME~string~1~~800~0~0~0~~~1;CREATEDBYYOMINAME~string~1~~800~0~0~0~~~2;CREATEDON~date~1~yyyy-MM-dd HH:mm:ss.SSS~26~0~0~0~~~3;CREATEDONUTC~date~1~yyyy-MM-dd HH:mm:ss.SSS~26~0~0~0~~~4"
pattern = re.compile (";(.+?)~")
matches = re.findall ( pattern, variable )
print matches

edited Dec 18 '18 at 19:04

answered Dec 18 '18 at 08:42

Raffi

1,810
2
16
25

How to split each individual value between two string in Python

5 Answers5