What is Best Method to Prepend A Few Lines of Text to a Large File?

I know KM has a nice tool to append: Append Text to File, but I don't see anything to prepend.

I know there are number of shell script and AppleScript gurus in our members, so I'm hoping someone has already solved this problem:

What is Best Method to Prepend A Few Lines of Text to a Large File?

By large I mean 2,000 - 3,000 lines. I know, that's not really large, but I need this to work very fast because I'm using this in a dynamically created file in a Prompt With List action.


Well - @JMichaelTX, I waited until a guru responded to your request. Since they didn't, I'm fine with embarrassing myself to give you my solution to shred up. I'm pretty sure my method won't suffice, for your need, but it works for my prepending needs. My hope is maybe it will trigger a faster/better solution from you or someone else. I copied a whole AppleScript PDF manual (771 pages), to test, on my 2014 iMac and it took about 10 seconds.

Anyway, enough chatter.

Here is the macro but don't embarrass me too much:

Write to Test file (Prepend).kmmacros (6.7 KB)



I think you've got it about right. The general approach I've always used is:

  1. Save new text to a variable.

  2. Open the original file for reading and writing.

  3. Read the file into a variable.

  4. Write the new text followed by the original file text to the original file.


Here's another way.

Search and Replace Action (v9.0.5)

Search and Replace.kmactions (519 B)


@thoffman666 - I like that! :sunglasses:

Many thanks. It seemed like an ideal use for \A.

For those who don't already know, a search and replace of \Z can be used to append.

1 Like

Unfortunately there isn't really a good way to prepend information to a file.

All of the "solutions" that I am aware of are basically variants of this:

  1. Take the existing file's data and store it (either in RAM or in a temp file)

  2. Add the new information to the file

  3. Append the old information to the new file

That being said, I bet that would probably be fast enough for your purposes.

Here's a simple prepend.sh shell script which will take whatever input it is given and prepend it to a file, which is define as FILE=

(You can/should edit that line to the appropriate file on your system.)

#!/usr/bin/env zsh -f
# Purpose: add new information to the top of a file
# From:	Timothy J. Luoma
# Mail:	luomat at gmail dot com
# Date:	2020-07-10

	## Change this to wherever your file is

## Shouldn't need to change anything below this line ##


	# a useful default for macOS

	# if the script is given no arguments, don't do anything
if [[ "$#" -gt "0" ]]

		# does the file exist and have a size greater than zero bytes?
	if [[ -s "$FILE" ]]

			# if we get here, the answer is yes

			# this lets us use `$EPOCHSECONDS` as a variable
		zmodload zsh/datetime

			# a random temporary filename

			# there's no way that file should exist, but let's remove it if it does
		rm -f "$TEMPFILE"

			# move the existing file to the temporary file
		mv -f "$FILE" "$TEMPFILE"

			# now we take whatever input we are given and save it to the file
		echo "$@" >>| "$FILE"

		## OR, remove the previous line and use this with Keyboard Maestro shell variable
		# echo "$KMVAR_Whatever" >>| "$FILE"

			# now we append all of the information which was previously in the file
			# and delete the temp file
		cat "$TEMPFILE" >>| "$FILE" && rm -f "$TEMPFILE"


			# if we get here, the file didn't exist before, so we just create it

		echo "$@" >>| "$FILE"


exit 0


Wow! What GREAT bunch of people we have in this forum, with expertise not only in KM but in many other related tools!

I do not dislike any of your solutions. Because you guys have provided such a good selection of solutions, I'm going to spend some time this weekend analyzing and running some tests on them to see what works best for my use case.

Thanks again!


Perhaps I'm missing something but why not just use cat.

dump your two lines to file_a cat file_a and your file together.

Could you please provide a complete, detailed, example for those of us that are bash-challenged?

# Create sample 'main' file
echo "line 1" > /tmp/main_file.txt
echo "line 2" >> /tmp/main_file.txt
echo "line 3" >> /tmp/main_file.txt
echo "line 4" >> /tmp/main_file.txt

# create file to 'prepend'
echo "line 0" > /tmp/file_to_prepend.txt

# using 'cat' to concatenate the two files
cat /tmp/file_to_prepend.txt /tmp/main_file.txt > /tmp/new_file.txt

# move the new file back over the original file
mv /tmp/new_file.txt /tmp/main_file.txt

If the file size is small enough, then just read it in to a variable, and save it back.

But to avoid it going in to a variable, the cat suggestion is probably the best.

Expanding on that, this looks like a nice solution:

cat - ~/File.txt >~/Temp.txt
mv ~/Temp.txt ~/File.txt


\z is generally better for this - \Z actually matches before the trailing line ending at the end of the file, where \z matches at the very end, mirroring \A.


Absolutely the better use of stdin on cat there ^^^.