How to modify a text within string using regular expressions in python?

Question

I'm trying to change a file containing strings like:

Record 1 : 
{ "K1":"value1" , 
  "K2":"value2" 
}

Record 2 :
{ "K1":"value3" , 
  "K2":"value4" 
}

to

{
    "Record_1" : 
        { "K1": "value1", 
          "K2": "value2" 
    }, 

    "Record_2" :
        { "K1": "value3", 
          "K2": "value4" 
        }
}

(to make it into a correct JSON format).
The part of code that I'm having problems with is:

pattern = r"(\s*)Record (\d+):"
all_records_json = re.sub(middle_pattern, "\"Record_"+ ??? + "\" : ",all_records)

And I don't know what to put instead of ??? so that it reads the (\\d+) part that matched the pattern.

Answer 1

First, your pattern doesn't actually match your data in the first place:

>>> all_records = '''Record 2 :
... { "K1":"value3" , 
...   "K2":"value4" 
... }'''
>>> pattern = r"(\s*)Record (\d+):"
>>> re.findall(pattern, all_records)
[]

That's because your data has a space between the digits and the colon. You need to fix that. While we're at it, I have no idea why you're putting a group around the preceding whitespace, so let's not do that. So we get:

>>> pattern = r"\s*Record (\d+)\s*:"
>>> re.findall(pattern, all_records)
[2]

Now, your only capturing group is the \\d+ . So that will be group 1. Which you can include in the substitution as \\1 . So:

>>> print(re.sub(pattern, r'"Record_\1": ', all_records))
"Record_2":
{ "K1":"value3" ,
  "K2":"value4"
}

That still isn't valid JSON, but it's what you wanted, right?

If you read the docs, re.sub explains that "Backreferences, such as \\6 , are replaced with the substring matched by group 6 in the pattern." For full details on back references for both (…) groups and (?P<name>…) groups, look them up in the Regular Expression Syntax . You should also read the Regular Expression HOWTO , which explains all of this in a more novice-friendly way.

Notice that I made the substitution a raw string just like the pattern, so I didn't have to escape the \\1 , and I also used single instead of double quotes, so I didn't have to escape the " characters.