Skip text in output with grep in perl mode

Question

I have text like this:

blah, blah <foo:ImportantText> blah blah time=1.234 blah blah
blah, blah <foo:AlsoImportant> blah blah blah time=9.9 blah blah
blah, blah <foo:ImportantText> blah blah time=0.987 blah blah

I want to get:

<foo:ImportantText>=1.234
<foo:AlsoImportant>=9.9
<foo:ImportantText>=0.987

I use this line:

grep -Po '(<foo:.+>).+time=(\d+.\d+)' logfile.txt

Note I don't need to worry about false positives as <foo: and time= do not appear elsewhere in the text. Also blah blah is random text, not a literal.

This gives me:

<foo:ImportantText> blah blah time=1.234
<foo:AlsoImportant> blah blah blah time=9.9
<foo:ImportantText> blah blah time=0.987

How do I drop the intermediate text? I thought '(<foo:.+>)(?=.+time)=(\d+.\d+)' might work, but it doesn't.

Update:

grep -Po '(<foo:.+>).+time=(\d+.\d+)' logfile.txt
                | awk -F ' ' '{print $1substr($NF,4)}'

This works, but is there a grep-only solution?

score 2 · Accepted Answer · 2018-08-31 00:36:03Z

2

Better use sed:

$ sed -E 's/.*(<foo:.+>).+time=([0-9.]+).*/\1=\2/' logfile.txt 
<foo:ImportantText>=1.234
<foo:AlsoImportant>=9.9
<foo:ImportantText>=0.987

edited Aug 31, 2018 at 0:36

answered Aug 30, 2018 at 2:42

user232326

I tend to only use grep and awk for everything, but that is a much easier solution.

Ken Y-N
– Ken Y-N

2018-08-30 06:46:32 +00:00
Commented Aug 30, 2018 at 6:46
I can't do one character edits here, but that's sed not ed.

Ken Y-N
– Ken Y-N

2018-08-31 00:33:23 +00:00
Commented Aug 31, 2018 at 0:33

Add a comment |

Stack Exchange Network

Skip text in output with grep in perl mode

1 Answer 1

You must log in to answer this question.

Hot Network Questions

Skip text in output with grep in perl mode

1 Answer 1

You must log in to answer this question.

Related

Hot Network Questions