bash scripting, getting a counter to align with output?

elfoozo · 08-26-2019, 03:06 AM

I'm trying to write something that retrieves total lines of "myfile.txt" then prefixes each iteration of "myfile.txt" with a number as it counts down processing each line of "myfile.txt". I don't know what this programming concept is called so my challenge is figuring out how to search for what this is even called. So I tried nested for loops but it isn't what I'm looking for.

I've tried in my script:

Code:

#!/bin/bash
totallines="$(cat myfile.txt | wc -l)"
for ((eachline="${totallines}"; eachline>=1; eachline--))
   do echo -n "${eachline}: " 
   for line in "$(cat myfile.txt)"    
      do echo "${line}" 
   done 
done

Contents of myfile.txt:

Code:

Line one
Line two
Line three
Line four
Line five

What I'm producing: (wrong)

Code:

5: Line one
Line two
Line three
Line four
Line five
4: Line one
Line two
Line three
Line four
Line five
3: Line one
Line two
Line three
Line four
Line five
2: Line one
Line two
Line three
Line four
Line five
1: Line one
Line two
Line three
Line four
Line five

What I'm trying to produce:

Code:

5: Line one
4: Line two
3: Line three
2: Line four
1: Line one

Is there an official name for this kind of coding logic?
This is not a schoolwork assignment. Any pointers are greatly appreciated.

GazL · 08-26-2019, 04:55 AM

Quote:

Is there an official name for this kind of coding logic?

"wrong" springs to mind.

If you're looking for a bash solution,

Code:

count=$( wc -l < inputfile.txt )
while read line
do
 printf "%4d:%s\n" $(( count-- )) "$line"
done < inputfile.txt

If you don't mind using GNUisms, you could also do something like
tac inputfile.txt | cat -n | tac

The 'nl' utility from coreutils is also an option.

Firerat · 08-26-2019, 05:48 AM

you can also do

Code:

printf "%4d:%s\r" $(( count-- )) "$line"

which would 're-write' the same line
you would want to pad the end if the length varries

elfoozo · 08-27-2019, 12:28 AM

Thank you for the replies. Hopefully my original ask wasn't too vague. I wasn't trying to be cagey and of course I realize what I was doing was wrong. I just don't have the experience to know better (yet). I wasn't looking for someone to write the solution but instead was trying to understand the mechanics of how one could count down as a script works through a list, and give me the observer a visual indicator how many iterations were left to go from that inputfile.txt list.

Based on your code example, if I were not wanting to just display (printf) contents of inputfile.txt but do some kind of processing on what is within inputfile.txt... Say as example I wanted to do something like dig a list of hostnames provided from inputfile.txt or do some simplistic housekeeping based off of what's in inputfile.txt? Is it likely to still be a while loop for that leg of a processing.

GazL · 08-27-2019, 04:31 AM

There's almost always more than one way to do something in shell-script. If you really wanted to, you could put the read in the body of a for loop, but when sequentially processing an input file line by line, while read; do ... seems like the more natural fit.

If the input file is relatively small, one might even want to do a Perlesque slurp into an array (see bash mapfile command).

Short answer: pick what feels right.

Sefyir · 08-27-2019, 11:24 AM

This will read strange to anyone using bash. a while loop will go over every line

Code:

for ((eachline="${totallines}"; eachline>=1; eachline--))

You prepend the line count (as expected)
The next line you put out will have this value. Since you put out multiple lines after one iteration of this, this gives you your undesired behavior

Code:

   do echo -n "${eachline}: "

You iterate through every line in the whole file, echoing out the line. Probably not what you want

Code:

   for line in "$(cat myfile.txt)"    
      do echo "${line}"

Quote:

Based on your code example, if I were not wanting to just display (printf) contents of inputfile.txt but do some kind of processing on what is within inputfile.txt... Say as example I wanted to do something like dig a list of hostnames provided from inputfile.txt or do some simplistic housekeeping based off of what's in inputfile.txt? Is it likely to still be a while loop for that leg of a processing.

GazL gave you a good answer for this but to simplify it slightly,

Code:

count=$( wc -l < inputfile.txt )
while read line
do
 other_processing_commands
 dig +short "$line"
 echo "$count": "$line"
 (( count-- ))
done < inputfile.txt
28: Line
27: Line
26: Line
25: Line
24: Line
...