How to group log files by hours using grep and uniq in Linux

For example, you have a file which contains log of some events.

Input data:

Dec  9 00:33:42 some log
Dec  9 00:56:49 some log
Dec  9 01:13:12 some log
Dec  9 01:22:02 some log
Dec  9 01:35:52 some log
Dec  9 03:15:52 some log
Dec  9 12:17:52 some log

And you want to group this file by hours. The code below will help group and display data:

grep -oP "Dec\s+9\s(\d{2})" | sort | uniq -c

Output data:

2 Dec 9 00
3 Dec 9 01
1 Dec 9 03
1 Dec 9 12

First column - it's count of unique lines with the date.

Let's break down the command:

• Flag -o is required to output only the requested substring (Dec\s+9\s(\d{2}))
• Flag -P is required for enable perl-style regexes

The pipeline works as follows:

1. Command sort sorts the data
2. And uniq -c finds count of unique lines

Thank you for reading. Best regards, Ildar.