2

I developed code down below, I would like improve this code for future needs. So I am thinking that if I use functions, the code could be much better than now it is and It will look organized.

So how I turned to a program or organized code block?

Could you please help me?

Version: Red Hat Enterprise Linux Server release 7.3 (Maipo)

(output must be same)

 #!/bin/bash

 hammer host list >| host.list
 cat host.list | grep "RHEL Server" | awk -F'|' '{ print $3}' |  sort > 
 host.modified1
 grp=`sort host.modified1 | uniq -c`
 echo "Linux Versions Grouped by Count" > host.modified1
 echo "" >> host.modified1
 echo "$grp"  >> host.modified1
 echo "" >> host.modified1
 echo "Linux Versions and Hostnames" >> host.modified1
 echo "" >> host.modified1
 cat host.list | grep "RHEL Server" | awk -F'|' '{ print $3"|"$2  }' | sort 
 >> host.modified1

output;

  Linux Versions Grouped by Count

  8  RHEL Server 6.5  
 21  RHEL Server 6.6  
  1  RHEL Server 6.7  
 10  RHEL Server 6.8  
 39  RHEL Server 6.9  
 19  RHEL Server 7.2  
 34  RHEL Server 7.3  
 30  RHEL Server 7.4  


Linux Versions and Hostnames

RHEL Server 6.5  | xxx.dnsname
Improve this question

2 Answers 2

Reset to default
1

You could definitely put this in a function if you wanted. Here's an example where I've also tried to cut down on the processing that you are doing. Since I don't know what the output from hammer host list looks like, I have made guesses from what your original code did.

do_stuff () {    
    local tmpfile=$(mktemp)
    hammer host list | grep -F 'RHEL Server' >"$tmpfile"

    printf 'Linux Versions Grouped by Count\n\n'
    awk -F '|' '{ c[$3]++ } END { for (h in c) printf("%d\t%s\n", c[h], h) }' "$tmpfile" | sort -k 2

    printf '\n\nLinux Versions and Hostnames\n\n'
    awk -F '|' '{ printf("%s | %s\n", $3, $2) }' "$tmpfile" | sort

    rm -f "$tmpfile"
}

do_stuff >rhel_things.txt

The function writes everything out to standard output, and you redirect that output when calling the function. It uses a temporary file to store the hammer output, and deletes this output once it's done with it.

If you would want to break it down further:

pre_parse () {    
    local tmpfile=$(mktemp)
    hammer host list | grep -F 'RHEL Server' >"$tmpfile"

    printf '%s\n' "$tmpfile"
}

do_group_counts () {
    local infile="$1"

    printf 'Linux Versions Grouped by Count\n\n'
    awk -F '|' '{ c[$3]++ } END { for (h in c) printf("%d\t%s\n", c[h], h) }' "$infile" | sort -k 2
}

do_ver_and_hosts () {
    local infile="$1"

    printf 'Linux Versions and Hostnames\n\n'
    awk -F '|' '{ printf("%s | %s\n", $3, $2) }' "$infile" | sort
}

tmpfile=$( pre_parse )

{
    do_group_counts "$tmfile"
    printf '\n\n'
    do_ver_and_hosts  "$tmpfile"
} >rhel_stuff.out

rm -f "$tmpfile"
Improve this answer
6
  • 1
    Equally, could create the host list first, and then have a function for group by count, followed by a function for version+hostname. Commented Mar 8, 2018 at 8:23
  • 1
    @steve Example of that added. Commented Mar 8, 2018 at 8:38
  • The second example provides a best practice example because functions should be defined as having one purpose and task vs multiple Commented Mar 8, 2018 at 8:41
  • second example is the what I would like to do but it only creates rhel_stuff file which is empty. Commented Mar 8, 2018 at 12:04
  • @bbenzer Well, then something is wrong. I don't have your hammer output to look at so I can't unfortunately help with debugging. This also means that I've written the code above "blindly" and that there is no guarantee that exactly this code will work for you. Commented Mar 8, 2018 at 12:06
0

I think I can do it in a single awk program, as below. At first glance it will seem longer than it actually is, because it uses long descriptive variable names and is conscientious about immediately freeing memory when possible.

  1. awk has powerful native sort functions built in, so there's no need to pipe to an external program.

  2. awk's asort and asorti functions also take an optional third argument, being the sort method, which can be things like @ind_num_asc, @ind_num_desc, @ind_num_asc, @ind_num_desc, etc. for various ways to sort in ascending or descending order. As usual, see the man page.

  3. The program is cluttered with delete statements in order to be considerate with memory usage, which might become an issue as data sets grow.

  4. Shame you didn't give us sample input data to play with. I had a good time coding this, but Lord only knows how close it is to working without input data, so treat the code below as a proof-of-concept that may need work in order to, um, work. The basic take-away is that the most efficient method is probably all internal to a single awk program.

hammer host list \ | awk -F'|' ' \ BEGIN {i=0} /RHEL Server/ { unsorted_count[$3]++ unsorted_list[i++]=$3"|"$2 } END { printf "\nLinux Versions Grouped by Count\n\n" i=0 for (release in unsorted_count) count_list[i++]=unsorted_count[release]"\t"release delete unsorted_count n=asort(count_list,sorted_count) delete count_list for (i=1; i<=n; i++) printf "%s\n", sorted_count[i] delete sorted_count printf "\n\nLinux Versions and Hostnames\n\n" i=0 for (hostname in unsorted_list) host_list[i++]=unsorted_list[hostname]" | "hostname delete unsorted_list n=asort(host_list,sorted_list) delete host_list for (i=1; i<=n; i++) printf "%s\n", sorted_list[i] }'

Improve this answer
3
  • The actual processing done by the script is in a sense unimportant in this question. The main point is how to reorganize the code inte distinct operational parts. Commented Mar 8, 2018 at 9:42
  • @Kusalananda : Oh well. At this point, the only thing I'l do towards that is to insert a few blank lines in the script in order to visually divide the parts. Commented Mar 8, 2018 at 9:44
  • @Kusalananda : [I timed out on editing the prior comment, so this a replaceent] Oh well. A lot of work, and I missed the main point of the question. Still, it should have value for how it deals with the problem, and for how it demonstrates the abilities of awk. Also, although it isn't divided up into functions, it's pretty well organized and linear. Commented Mar 8, 2018 at 9:51

You must log in to answer this question.

Start asking to get answers

Find the answer to your question by asking.

Ask question

Explore related questions

See similar questions with these tags.