Bash – How to Redirect Shell Stdout to the First Line of a File

bashshell

The shell standard output redirects to the last line of the file, is there a way to write it to the first line of the file?

Since the content of stdout is unpredictable, I suspect that this may require changing the default behavior of stdout, but I am not sure if it is feasible.

Example, redirecting a timestamp to file

echo `date` >> test.txt

Default save to last line of file

Mon Aug 31 00:40:27 UTC 2020
Mon Aug 31 00:40:28 UTC 2020
Mon Aug 31 00:40:29 UTC 2020
Mon Aug 31 00:40:30 UTC 2020

Desired effect, save the output to the first line of the file

Mon Aug 31 00:40:30 UTC 2020
Mon Aug 31 00:40:29 UTC 2020
Mon Aug 31 00:40:28 UTC 2020
Mon Aug 31 00:40:27 UTC 2020

Thanks in advance!

Best Answer

To write the date to the beginning instead of the end of file, try:

{ date; cat file; } >file.new && mv file.new file

Discussion

Adding new text to the beginning of a file requires the whole file to be rewritten. Adding new text to the end of a file only requires writing the new text.
Andy Dalton's suggestion of just appending to the end of a file like normal and then using tac file to view the file is a good one.
echo `date` >> test.txt can be replaced by a simpler and more efficient date >> test.txt.

If one is using bash 4.3 or later then, as Charles Duffy points out, printf '%(%c)T\n' -1 >>test.txt is still more efficient.
The spaces around the curly braces are essential. This is because { and } are shell reserved words (as opposed to shell keywords which do not require spaces).

Related Solutions

Bash Command Line – Capture and Use Last or Nth Line in stdout

Here is a potential solution to your issue that should be reasonably (but not perfectly) safe in the presence of funky file names (doesn't handle filenames with linefeeds in them - probably fixable, but there might be other problems lurking).

Two functions, the first one runs find with the parameters you pass it, saves the output in an array, and displays them. The second is just a helper to access that array.

myfind() {
  IFS=$'\n' __last_find_result=($(find "$@"));
  printf "%s\n" "${__last_find_result[@]}";
}
myget() {
  echo "${__last_find_result[$1]}";
}

Use case:

$ myfind . -name "c*"
./a b/c d
./.git/config
./.git/hooks/commit-msg.sample
$ vim "$(myget 0)"
# This opens the "./a b/c d" file.
$ vim "$(myget 2)"
# This opens ".git/hooks/commit-msg.sample"

Quotes not required around the $(myget index) if you don't have whitespace or other troublesome characters in your filenames.
Pushes the whole output of find to your environment, which might be limited. (Using a temporary file rather than that array would solve that, but has other issues - notably concurrent use from multiple shells.)

Shell – want the last line when duplicate awk

you can do this using this shell script:

#!/bin/bash
awk '/man1/{pos[$0] = NR}
END {
  for(key in pos) reverse[pos[key]] = key
  for(nr=1;nr<=NR;nr++)
    if(nr in reverse) print reverse[nr]
}' yourfile

Output:

[root@host ~]# sh shell.sh
man1 boy1
man1 boy3
man1 boy4
man1 boy2

Source

Best Answer

Discussion

Related Solutions

Bash Command Line – Capture and Use Last or Nth Line in stdout

Shell – want the last line when duplicate awk

Related Question