如何在条件语句和循环中使用diff命令

Question

I've created a script which recursively reads a path and prints its directory, and subdirectory into a csv file to write into excel columns. Now I want to modify this script such that first it will recursively read a path and print its content and then find the recursive difference of this path along with another path and I want to print if any directory is changed or not.

First code:

find path1 | while read file; do
  if [[ -d "$file" ]]; then
    echo $file > out.csv
  else
    echo...
  fi
done

and second code is:

diff -rq $path1 $path2 | while read file1; do
  if [[ "$file" != ]]; then
    echo Changed >> out.csv
  else
    echo no >> out.csv
  fi
done

Now I want to merge these codes such that fist it will recursively prints all content of path1 in csv file, then compare both paths and print changed if difference occur.

Answer 1

It's not overly clear what you're trying to accomplish. Your first part can be simplified to a single line:

find path1 -type d > out.csv

which only finds directories. (Note that the use of > overwrites out.csv , so previous contents don't matter (you made a mistake with that in your example)).

For you second part, you could do:

diff <( cd path1 && find . -type d ) <( cd path2 && find . -type d )

This grabs all the directories in path1, and all the directories in path2, and would output the difference in standard diff format...

I'm using process substition <( ) so bash treats the output as files. Also the cd to path1 or path2 before the find prevents find from outputting path1/ or path2/ in its results (otherwise every line would be different). You could do something similar with sed to strip off the path names.

---

If on the otherhand, your goal is to learn bash loops, you would do this as follows:

rm out.csv
find path | while read file; do 
    [[ -d $file ]] && echo $file >> out.csv
done

cat out.csv | while read file; do 
    [[ -e path2/${file#path1/} ]] || echo "file changed: $file"
done

Note that unlike the former example, this does not print any paths in path2 that are not in path1.

Here, I used ${file#path1/} , which strips path1/ from the front of each filename (as you're now looking in path2 ).

Hope this helps.