Import bash history to zsh history.

bash_to_zsh_history.rb

#################################################################
# = This script transfers bash history to zsh history
# = Change bash and zsh history files, if you don't use defaults
#
# = Usage: ruby bash_to_zsh_history.rb
#
# = Author: Ankit Goyal
#################################################################

# change if you don't use default values
BASH_HISTORY_FILE_PATH="#{ENV['HOME']}/.bash_history"
ZSH_HISTORY_FILE_PATH="#{ENV['HOME']}/.zsh_history"

# Read the bash history file
bash_hist_file = File.read(BASH_HISTORY_FILE_PATH)

# Get the list of commands from bash history hile
command_list = bash_hist_file.split("\n")

# Open the zsh history file
zsh_hist_file = File.open(ZSH_HISTORY_FILE_PATH, "a")

# Get timestamp required for zsh history file format and update the history file
time = Time.now.to_i
command_list.each do |command|
  time += 1
  zsh_hist_file.write(": #{time}:0;#{command}\n")
end

# Close the file
zsh_hist_file.close

Thx the script is fast and good

Worked phenomenally, thank you!

Just what I was looking for ! Thank you !!

Thanks !

Perfect!

This command seems to ignore the timestamps in the ~/.bash_history file?

I'm getting:

zsh: corrupt history file /home/user/.zsh_history

Did anyone else have this issue?

My .bash_history has 23k lines, I think this might be an issue. I tried to increase ZSH history limit and rerun the converter, but still didn't work.

EDIT: Solution:

mv .zsh_history .zsh_history_bad
strings .zsh_history_bad > .zsh_history
fc -R .zsh_history

I tried this, but I think it was before I increased ZSH history limit, so it was mangling my history to 10k lines. From my .zshrc:

HISTSIZE=1000000  # how many lines of history to keep in memory
SAVEHIST=1000000  # how many lines to keep in the history file

Works fine and fast! Thank you!

Awesome script! Worked like a charm. Thank you for sharing!

Thanks! 🥇

Thank you! For anyone else who uses this, if you get utf-8 errors, try upgrading ruby

Thank you for the script, it worked!

Only thing I had to modify is the time = Time.now.to_i line of your script, because it was giving me future times for the commands transfered from the old ~/.bash_history:

19070* 2018-04-08 15:45  git pull
19071* 2018-04-08 15:45  makepkg -sci    # <--- old .bash_history
19072* 2018-04-08 12:55  bash            # <--- new .zsh_history
19073* 2018-04-08 12:52  echo $SHELL

Just changing it to time = Time.now.to_i - (nr. of lines in old .bash_history) (e.g.: time = Time.now.to_i - 11469) fixed this and now I get the following view with history from zsh:

10534  2018-04-08 12:58  git pull
10535  2018-04-08 12:58  makepkg -sci    # <--- old .bash_history
10536  2018-04-08 13:14  bash            # <--- new .zsh_history
10537  2018-04-08 13:14  echo $SHELL

If you have invalid UTF-8 sequences in your .bash_history, just remove those characters with iconv like

iconv -f UTF-8 -t UTF-8//IGNORE .bash_history > .bash_history-utf8

Thank you.

one line command: awk '{print " :0:;"$1}' .bash_history > .zsh_history

one line command: awk '{print " :0:;"$1}' .bash_history > .zsh_history

Almost... $1 needs to be $0, or there won't be any args, and > should be >>, to append instead of replace. Also, at least for my history config, there's a : missing at the beginning. Finally, I'd not import duplicates, ergo sort ~/.bash_history | uniq | awk '{print ": :0:;"$0}' >> ~/.zsh_history.

Thanks, it's work

I was surprised to see the top Google results for this are Ruby and Python scripts, huh! Then I missed the awk command at the bottom here and did it with Perl. First I checked the format of both my histories:

$ head -n1 ~/.zsh_history
: 1551826967:0;ls -al ~
$ head -n2 ~/.bash_history
#1532234168
exit

Then wrote a Perl command that I ran under Bash so Zsh would not be running and conflict (from Zsh, I ran exec bash and confirmed Zsh was not running with pgrep zsh):

perl -0777 -pe 's/^#(\d+)\n(.*?)$/: \1:0;\2/gm' ~/.bash_history >> ~/.zsh_history

Thanks, it's work fine!

awesome dude

Thanks, works perfectly.

Worked perfectly.

Worth noting that you should remember to increase your SAVEHIST value first.

Cool! This is a node script version

https://gist.github.com/marchrius/71f858c2b032f037d286cbf2394ad8a7

getting this error message - anyone have a solution?

./bash_to_zsh_history.rb: line 15: syntax error near unexpected token (' ./bash_to_zsh_history.rb: line 15: bash_hist_file = File.read(BASH_HISTORY_FILE_PATH)'

Ruby version is ruby 2.6.3p62 (2019-04-16 revision 67580) [universal.x86_64-darwin19]

getting this error message - anyone have a solution?

./bash_to_zsh_history.rb: line 15: syntax error near unexpected token (' ./bash_to_zsh_history.rb: line 15: bash_hist_file = > File.read(BASH_HISTORY_FILE_PATH)'

Ruby version is ruby 2.6.3p62 (2019-04-16 revision 67580) [universal.x86_64-darwin19]

@robertstrom You probably copy-pasted the script wrong. There's an extra ' at the end of line 15.

Worked perfectly! Thanks!

Thanks bigplum & gimpf,

sort ~/.bash_history | uniq | awk '{print ": :0:;"$0}' >> ~/.zsh_history

just what I needed on a VM without ruby installed

Thanks for the script

Awesome work! Thank you!

clean, simple and fast - thx

Personally I found awk and PERL to be to heavy weight :) So I went for

cat ~/.bash_history | cut -d' ' -f1- | sort | uniq | xargs -I % echo ": $(date +%s):0;" % >> ~/.zsh_history

Personally I found awk and PERL to be to heavy weight :) So I went for

cat ~/.bash_history | cut -d' ' -f1- | sort | uniq | xargs -I % echo ": $(date +%s):0;" % >> ~/.zsh_history

This worked like a charm on OSX Catalina from bash to zsh.