I am working on a CSV file like this
"EF", "12345", "Test, String", "Xyz" and I need output as
"EF", "12345", "Test$$$ String", "Xyz"
basically I need to replace the comma with $$$
4 Answers
You can use gnu-awk for this:
awk 'BEGIN{OFS=", "; FPAT="\"[^\"]+\""} {sub(/,/, "$$$", $3)} 1' file
"EF", "12345", "Test$$$ String", "Xyz"
Comments
You can try something like this:
, ([^"])
And replace by:
$$$ $1
Or even this:
([^"]),
And replace by:
$1$$$
4 Comments
Kent
my sed and awk don't support lookahead/behind. yours does?
jsphdnl
Thank you, but the regex fails for "EF", "12345", ",Test String", "Xyz" Can this case also be handled
NeronLeVelu
Failed if
"field", ", ", ", ", "field" occursed ':a
s/^\(\([[:blank:]]*"[^"]*",\)*[[:blank:]]*"[^",]*\),/\1\$\$\$/
t a
' YourFile
- posix version
- use recursive remove of
;inside string delimited by"(so not the fastest) - assuming there is no escaped
"inside string (if the case, need to adapt a bit)
Comments
It works with sub as well.
awk '{sub(/Test,/,"Test$$$")}1' file
"EF", "12345", "Test$$$ String", "Xyz"
sortwhich is faster, however I have some rows which contain comma(,)in the row values.How do I sort in such case as the number of columns will increase if (,) is used as delimiter. Example: Here I want to sort based on 2 and and 4th column, but if I use (,) as delimiter i will get total 5 columns in 1st row and 4 columns in second row that would spoil the sorting. "wert","DGC","Xyt, temp","3456" "wert","ABC","ppp","1234"|character as delimiter. Good luck.