Reputation: 21
I have the data given below with other lines of data in a huge file to be more specific this is a .json file, I have stripped the whole file except for this one catch.
Data:
I want the data to look like this
So far I have this piece of code however it seems to be replacing the whole files "," with , and not the ones just within []
sed 's/\(\[\|\","\|\]\)/\,/g' testfile_1.txt >testfile_2.txt
Any help would be great
**Note: Edited fourth condition
Cheers.
Upvotes: 0
Views: 356
Reputation: 21
Thank you for your contributions @clearlight
Here is my solution to the problem with multiple occurrences;
sed -e ': a' -e 's/\(\[[^]]*\)","/\1,/' -e 't a' testfile_1.txt >testfile_2.txt
Using label 'a' till end of label 'a'
pseudocode:
while (not end of line){ replace "," with ' }
Upvotes: 0
Reputation: 12615
That's tougher without the non-greedy matching perl regex offers, but this seems to do it.
sed -e 's/\([^\[]*\"[A-Z\-]*\)\",\"\(.*\)/\1,\2/'
Upvotes: 1