Vfrdtyky

Here's what my script looks like

#/bin/bash

touch input.txt

touch output.txt

seq -w 0 999 >>input.txt

input=$(cat input.txt)



for i in $input

  do

    if [ $(($i%2)) -eq 0 ]; then

      echo $i "even" >> output.txt

    else

      echo $i "odd" >> output.txt

   fi

 done

Here's the result of running the script and viewing the output.txt file created

000 even

001 odd

002 even

003 odd

004 even

005 odd

006 even

007 odd

I would like the script to do this for all 1,000 lines of the script, but I get an error message on line 9 saying

./tester.sh: line 9: 008: value too great for base (error token is "008")

My end goal is for the script to add each number on a line, and then tell if the number is even or odd, outputting to output.txt for all 1000 lines of the file.

End goal output file:

000 even

001 odd

002 even

003 odd

...

505 even

506 odd

507 even

508 odd

...

998 even

999 odd

From 000 all the way to 999

Use seq as seq and use printf to print your number in the format you like.

Bash arithmetic expansion interprets strings with leading zeros as octal numbers. You can force the number to be in 10th by prefixing it with 10# like (( 10#$i % 2)).

for i in $input

  do

    if [ $(( 10#$i % 2)) -eq 0 ]; then

      echo $i "even" >> output.txt

    else

      echo $i "odd" >> output.txt

   fi

 done

1. Keep in mind that arithmetic expansion (( .. )) can do comparisions. It's clear to if (( 10#$i % 2 == 0 )); then.


2. I find printf "%03d" "$i" to be just clearer in this case.


3. No need to touch a file before >>, should create the file aumatically (this can be turned off with some bash set -o option, but I haven't seen anyone use it).


4. 
   input=$(cat ...); for i in $input is just bad. Don't read lines with for
   


5. I don't like temp files.


6. 
   How to read file line by line.

Your script is just:

seq 0 999 | xargs -i sh -c 'printf "%03d " "$1"; (( $1 % 2 == 0 )) && echo even || echo odd;'  >>output.txt

If you prefer while read:

seq 0 999 | while IFS= read -r num; do 

    printf "%03d " "$num"; 

    if (( num % 2 == 0 )); then 

        echo even

    else

        echo odd

    fi

done >>output.txt

Or if you have to have your input.txt file containing 000n001n002n and so on it's time for a tee:

seq -w 0 999 | tee -a input.txt | while IFS= read -r num; do 

    echo -n "$num "

    if (( 10#$num % 2 == 0 )); then 

        echo even

    else

        echo odd

    fi

done >>output.txt

This is a skeleton code for reading a text file line by line and acting upon the lines... Fill the missing part according to your own needs.

#!/bin/bash

{

  while read -r line; do

    if (( line % 2 == 0 )); then

      # ...

    else

      # ...

    fi

  done < input.txt

} > output.txt

You may also apply pre-processing to the input file with <(cmd ...) notation:

#!/bin/bash

{

  while read -r line; do

    ...

  done < <(arbitrary-cmd input.txt | another-cmd ... )

} > output.txt

This form looks nicer but it spawns a "subshell" and makes it impossible for the code inside the while block to modify variables defined outside it, should you have any.

#!/bin/bash

{

  arbitrary-cmd input.txt | another-cmd ... | while read -r line; do

    ...

  done

} > output.txt

Your script could be something like

#!/bin/ksh

input=$(seq -w 0 999)

for i in $input

  do

    if [ $(($i%2)) -eq 0 ];then

      echo $i "even" >> output.txt

    else

      echo $i "odd" >> output.txt

    fi

  done

then your output will be something like

000 even

001 odd

002 even

003 odd

004 even

005 odd

006 even

007 odd

Then you could grep "even" or "odd" and execute what you need or you could execute your command directly inside the if/else statement.

This works too:

#!/bin/bash

for x in `seq -w 0 999`; do

    if [ $((10#$x%2)) == 0 ]; then

        echo $x even

    else

        echo $x odd

    fi

done > output.txt