How to use sed text processing

Question

please i need your help on this topic.

i have 3 temp files and nbf.txt.
i want to copy each hour lines from this 3 temp files to nbf.txt:
inside each temp file we have
temp file for each hour:
tmp1
Receiveid Transaction
======================
2020-05-29 00->12

tmp2
Failed Transaction
===================
2020-05-29 00->1 failed

temp3
Failed Open Sessions Per Hour
==============================
2020-02-29 00->2 sessions

so i need to use sed for each temp file:

i am using the below command in my script:

sed -i -e '1,$p' nbf.txt -e '1,3d' tmp1
sed -i -e '1,$p' nbf.txt -e '1,3d' tmp2
sed -i -e '1,$p' nbf.txt -e '1,3d' tmp3

the result inside nbf.txt NB: please i need the below result using sed or tail or awk

Received Transactions Per Hour
==============================
2020-02-29 00->69646 requests

2020-02-29 23->82666 requests

2361523 Received Transactions


Rejected Transactions Per Hour
===============================

0 Rejected Transactions


Failed Open Sessions Per Hour
==============================
2020-02-29 00->2 sessions
2020-02-29 22->1 sessions

8 Failed Sessions

but not working corrdctly

Could you add a couple of lines from one of the logs, as an example? — Gerard H. Pille, May 28 '20 at 16:21
I don't get where the failed and rejected connections come from. — Hauke Laging, May 28 '20 at 17:17

score 1 · Answer 1 · answered May 28 '20 at 17:43

In general there is no reason to pipe from cat to grep to awk and even to another awk because a single awk can do all that very well.

$ awk -F'[ :]' '/:/ { datestring=$1; a[$2]++;}; END { for(i=0;i<24;i++) '\
'{ ind=sprintf("%02d",i); '\
'printf "%s %2s->%7d\n",datestring,ind,(a[ind]>0) ? a[ind] : 0 ; }; }' \
Received_transaction.log-$Hour
2020-05-28 00->      0
2020-05-28 01->      0
2020-05-28 02->      0
2020-05-28 03->      0
2020-05-28 04->      0
2020-05-28 05->      0
2020-05-28 06->      0
2020-05-28 07->      0
2020-05-28 08->      0
2020-05-28 09->      0
2020-05-28 10->      0
2020-05-28 11->      0
2020-05-28 12->     13
2020-05-28 13->      0
2020-05-28 14->      0
2020-05-28 15->      0
2020-05-28 16->      0
2020-05-28 17->      0
2020-05-28 18->      0
2020-05-28 19->      0
2020-05-28 20->      0
2020-05-28 21->      0
2020-05-28 22->      0
2020-05-28 23->      0

`00:00 ====> 12:59` is 13 hours, not one. Due to race conditions there may be wrong entries in a file. Shall they be detected and ignored? Basically you are just counting the lines which contain a `:`, right? — Hauke Laging, May 28 '20 at 18:06

How to use sed text processing

1 Answers1