python替换中的正则表达式

Question

Is there any way using regular expression in python to replace all the occurrences of , (comma) after the flower braces {

Data is of the following format in a file - abc.json

{
"Key1":"value1",
"Key2":"value2"
},

{
"Key1":"value3",
"Key2":"value4"
},

{
"Key1":"value5",
"Key2":"value6"
}

This should result in following -

{
"Key1":"value1",
"Key2":"value2"
}

{
"Key1":"value3",
"Key2":"value4"
}

{
"Key1":"value5",
"Key2":"value6"
}

As you can see the , (comma) has been removed after every braces } .

Would be helpful if this can be achieved via jq as well, apart from python REGEX

Answer 1

Test Source: https://regex101.com/r/wT6uU2/1

import re
p = re.compile(ur'},')
test_str = u"{\n\"Key1\":\"value1\",\n\"Key2\":\"value2\"\n},\n\n{\n\"Key1\":\"value3\",\n\"Key2\":\"value4\"\n},\n\n{\n\"Key1\":\"value5\",\n\"Key2\":\"value6\"\n}"

re.findall(p, test_str)

But use replace instead

replace }, -> }

Answer 2

This works:

import re

s="""{
"Key1":"value1",
"Key2":"value2"
},

{
"Key1":"value3",
"Key2":"value4"
},

{
"Key1":"value5",
"Key2":"value6"
}"""

pattern=re.compile(r'(?P<data>{.*?}),', re.S)

print pattern.findall(s)

s1=pattern.sub(r'\g<data>', s)
print s1

Answer 3

If you intend to process the resulting JSON in jq, it's probably easier to wrap it in brackets [{...}, {...}] to make it a JSON array. Then, you can use .[] in jq to unwrap the array.

Answer 4

Before you even consider other options, you really should go back to the source that generated that file and make sure it actually outputs valid json.

That said, you could use JQ to manipulate the contents as a raw string to add brackets, then parse it as an array to them spit out the contents.

$ jq -Rs '"[\(.)]" | fromjson[]' abc.json