Bash 脚本优化：行复制

Question

I have a working script but it is quite slow on large files. It simply copy a bunch of line from a file to another file at a given position. Here the script:

#!/bin/bash

source_file="..."
dest_file="..."

first_line_to_copy=...
last_line_to_copy=...

dest_line=...

cat "$source_file" | sed -n "$first_line_to_copy,$last_line_to_copy p" | while IFS= read -r line;
do
    sed -i "$dest_line i \\$line" $dest_file
    dest_line=$(($dest_line+1))
done

I didn't succeed to manage the copy in one sed command. So it always opens/closes destination file, seeks at dest_line which it really time consumming with a huge number of line.

Thanks for any help

Answer 1

First copy the lines in the range [$first_line_to_copy,$last_line_to_copy] to a temporary file and then insert that file to the specified location:

tempfile=/tmp/tempfile.$$
sed -n "$first_line_to_copy,$last_line_to_copy p" "$source_file" > "$tempfile"
sed -i.backup "$dest_line r $tempfile" "$dest_file"
rm -f "$tempfile"