CODEWITHSUNDEEP
regexsed
Jepper
Jepper

Reputation: 1111

sed regex cannot get first match

I give up... of the following

15 Sep 1605.00 (SPX1530U1605-E),0.25,0.0,0.05,0.10,0,87

I want to extract the number 1530 out of the blob. "SPX" can be any combination of capital letters [A-Z] and varies in length, (e.g. GOOG, FB). There is always a capital letter following the number, as in "U" in the example.

Below gets the second number, 1605. I'm at loss on how to extract 1530.

echo "15 Sep 1605.00 (SPX1530U1605-E),0.0,0.0,266.10,284.60,0,0" | \
gsed -r 's/.*[A-Z]([0-9].*)[-][A-Z].*/\1/g'

It would be acceptable to perform the operation on just the string "SPXW1530I1605-E" rather than the entire line.

Upvotes: 1

Views: 65

Answers (2)

hek2mgl
hek2mgl

Reputation: 157947

Usually grep is the tool of choice when you only want to extract data. You can use GNU grep, it offers perl compatile regular expression when you pass the -P option:

grep -oP '\([A-Z]+\K[0-9]+' file

We are searching for a literal ( followed by 1 or more capital (ASCII) letters. Then we are using \K which cleans up the match buffer. (Nice, isn't it?) The following numbers are the final match.

Upvotes: 2

anubhava
anubhava

Reputation: 784938

This should work:

s='15 Sep 1605.00 (SPX1530U1605-E),0.25,0.0,0.05,0.10,0,87'
sed -r 's/.*[A-Z]+([0-9]+)[A-Z].*/\1/' <<< "$s"
1530

Upvotes: 1

Related Questions