find difference of file line by line

gaurav_s · 07-01-2015, 10:12 AM

I have two files f1 and f2

f1 contains

Tom Harry Anthony

Cat Dog Camel

f2 contains

Tom Cook Anthony

Giraffe Camel Dog

i want to find difference line by line.

Such thatg if i do a diff on line 1 of both files i should get
Cook

and in line2 i should get
Cat Giraffe

I tried with comm command but its not much helpful,its expecting both file should be sorted.

Keith Hedger · 07-01-2015, 11:03 AM

Code:

diff -y f1 f2

see the man page for diff

gaurav_s · 07-01-2015, 11:13 AM

diff -y just prints both the lines.

gaurav_s · 07-01-2015, 11:21 AM

I tried with all the options of diff ,none of the options gives desired output.

TB0ne · 07-01-2015, 11:39 AM

Quote:

Originally Posted by gaurav_s

I tried with all the options of diff ,none of the options gives desired output.

Looks like you'll have to write a script to do this. Should just have to read both lines into an array, and compare the elements.

catatom · 07-01-2015, 08:10 PM

If you want to compare each word instead of each line, this piece of Bash script will translate whitespace (spacebar button) to newlines (Enter button).

$text | tr " " "\n"
## The tr syntax is a bit like grep's.

You could probably write a Bash script that uses diff to find each line with two variations, record each variation by assigning the output to two variables, then compare the values of those variables by applying tr followed by diff. You'll probably need to use an until statement that keeps going down the output of diff until it recieves empty input. All these lovely bash functions can be found on the bash info page. Enter info bash.

HMW · 07-02-2015, 04:12 AM

It is not clear to me what you are really aiming for here, since your example is somewhat ambiguous.

But, it seems as though you do NOT want to compare the files word by word, because then you would get a very different output than what you describe in your example. My guess is that you want to check if a word from file2 exists in the corresponding line of file1.

If that is indeed the case, then this will work:

Code:

#!/usr/bin/env python3

"""
http://www.linuxquestions.org/questions/linux-software-2/find-difference-of-file-line-by-line-4175546931/
"""

# Open the files
with open("file1.txt") as f1, open("file2.txt") as f2: 
    # Read both files. f1 as a list with lines,
    # f2 as a string 
    f1Lines = f1.readlines()
    f2Str = f2.read()

    # Split f2 into list based on WORDS
    f2Split = f2Str.split()

lineNum = 1 
j = 0 
# Now loop
for i in range(0, len(f2Split)):
    # Compare WORD of file2 against WORDS in file1,
    # line by line
    if f2Split[i] not in f1Lines[j]:
        print("Line " + str(lineNum) + ": " + f2Split[i])

    # If you're at the 3rd word in f2, simulate a
    # new 'line' and add 1 to index j
    if (i+1)%3 == 0:
        j+=1
        lineNum+=1

If I run this code on your example files, I get the following output:

Code:

./pydiff.py 
Line 1: Cook
Line 2: Giraffe

But then there's the odd Cat in your example:

Quote:

Such thatg if i do a diff on line 1 of both files i should get
Cook

and in line2 i should get
Cat Giraffe

I really don't understand why you should get 'Cat' as output. If you follow the logic of MY program, it looks for each word in file2, if it is NOT in the same line in file1 it prints it.
So...
file2, line2: Giraffe Camel Dog
file1, line2: Cat Dog Camel

There is no 'Giraffe' in line 2 of file1, therefore it prints.

But by YOUR logic, you should not only get:

Quote:

and in line2 i should get
Cat Giraffe

But you should ALSO get the following output from line1:
Cook Harry

Why? because there are no 'Harry's in line1 of file2.

So... if you want to do THAT kind of a check, well, it can be implemented with a couple of more lines of code, but as I said above, it is far from clear to me what it is you actually want to achieve here.

Best regards