CODEWITHSUNDEEP
bashdiff
Paolo
Paolo

Reputation: 1637

diff two batches of files

I would like to diff two batches of files. If I simply put them in two different directories and diff by directory, the comparisons will be alphabetical which I do not want.

Another approach would be to list files in text1.txt and list files in text2.txt:

text1:

a1
b1
c1

text2:

c2
a2
b2

How can I approach this such that my loop will be:

diff a1 c2
diff b1 a2
diff b2 c1

Upvotes: 0

Views: 120

Answers (3)

Idriss Neumann
Idriss Neumann

Reputation: 3838

Another solution :

#!/bin/bash

file1="..."
file2="..."

getSize(){
    wc -l "$1"|cut -d " " -f1
}

getValueFromLineNumber(){
    sed -n "$1p" "$2"
}

diffFromLineNumber(){
    f1=$(getValueFromLineNumber "$1" "$file1")
    f2=$(getValueFromLineNumber "$1" "$file2")
    diff "$f1" "$f2"
}

# get min size
[[ $(getSize "$file1") -le $(getSize "$file2") ]] && min=$s1 || min=$s2

for (( i=1 ; i <= "$min" ; i++)); do
    diffFromLineNumber "$i"
done

This solution takes care of the case where the two files don't have the same number of lines.

Upvotes: 0

rici
rici

Reputation: 241671

In bash, you can use the -u flag on read to read from a different fd. This allows you to read from two files in parallel:

while read -r -u3 file1 && read -r -u4 file2; do
  diff "$file1" "$file2"
done 3<file1.txt 4<file2.txt

Upvotes: 2

Austin Phillips
Austin Phillips

Reputation: 15746

You can use paste to join the two files, then a bash loop to process.

paste text1 text2 | while read file1 file2; do diff "$file1" "$file2"; done

Upvotes: 2

Related Questions