Find and Replace word in text files based on name of text file

Question

I have 2 files: A.txt, B.txt

Both these files contain the letter NODE in them.

I want to replace NODE in each of the files with A_NODE and B_NODE in A.txt and B.txt, respectively.

I know there is a for loop and sed involved, but not able to figure this one out. Any suggestions?

I was thinking:

for file in *; do sed -i 's/NODE/$file.NODE/g' file; done

See: [Difference between single and double quotes in Bash](https://stackoverflow.com/q/6697753/3776858) — Cyrus, Jun 15 '17 at 17:51
Did you try `set -x` or `echo sed -i 's/NODE/$file.NODE/g' file` for debugging? — Walter A, Jun 15 '17 at 19:45

score 0 · Answer 1 · answered Jun 15 '17 at 22:38

0

for file in *; do sed -i "s/NODE/$(basename -s .txt file)_NODE/g" file; done

The double quotes (") instead of single quotes (') enable bash variable interpolation (replacing the $ expression).
$(...) executes an arbitrary shell command and inserts the result.
basename -s .txt removes the directory part and the suffix .txt from the filename.

answered Jun 15 '17 at 22:38

sapanoia

That exact command did not work, but based of your comment, I modified a bit and this did: `for file in *; do sed -i "s/NODE/$(basename -s .txt "$file")_NODE/g" $file; done` – aram Jun 16 '17 at 23:52

score 0 · Answer 2 · answered Jun 16 '17 at 23:54

0

Thank you @sapanoia:

This works:

for file in *; do sed -i "s/NODE/$(basename -s .txt "$file")_NODE/g" $file; done

answered Jun 16 '17 at 23:54

aram

2 Answers2