gsamokovarov / jump Goto Github PK

Similar to this plugin that works with autojump: https://github.com/fdw/ranger-autojump

Hi! Thanks for the excellent tool.

I'm syncing some remote directories with Onedrive that have hyphens in their names. Jumping to these does not seem to work, at least on my MacOS installation. E.g. "j search" -> "j /some/remote/dirname - with hyphen/search". Whitespace alone seems to be handled fine, as in "j search" -> /some/dir with whitespace/search". I also noted that an apostrophe also does not work, as in "j search" -> "j /not/sure/why/you'd make a dirname like this but it's what I get".

Is it possible to escape these odd but allowed characters in directory names?

Hi, after the atomic rename commit, I observed this error

renameat(AT_FDCWD, "/tmp/jump033116582", AT_FDCWD, "/home/amos/.jump/scores.json") = -1 EXDEV (Invalid cross-device link)

In archlinux, /tmp is a tmpfs which i suppose may conflict with the renaming here.

Hi,

Is it possible to install jump locally without sudo permissions?

It would be best if jump was a substitute for cd. In order to do that, it needs to fallback to my existing habits when it can't provide a better match.

For example, I just started with jump 5 minutes ago. The first thing I did was j /mn<tab> because a) my database is empty b) I need to get to /mnt/path/to/foo. Normally I would cd /mn<tab> p<tab> etc... to tab complete each path.

But what I actually get is:

buchanan01 home
j /mn <cursor here after tab>

It's hard to display here, but my pattern of tab completion just inserted a space. I guess for now, the workaround is to rely on cd until I feel jump is doing a good job, but in reality switching between the two commands might not work out, and I might forget about jump.

I would like to be able to use j to jump around my commonly used directories but then use popd to navigate back through my dir stack.

The spec miss the block device's num_queues field.

Proposed in https://lists.oasis-open.org/archives/virtio-comment/201909/msg00004.html

I am getting this error "err: no entries in the database" on every attempt to use your program.
installed via homebrew.

I'm using zsh and oh-my-zsh with the Windows Subsystem Layer for Ubuntu.

I installed using the wget command listed for installing on ubuntu and install the deb etc. It all seemed to go fine. I added the eval thing to my .zshrc file but when I source my config I get this response:

┌─    /mnt/c/Users/David                                                                                                                       
└─➤ source ~/.zshrc
/home/david/.zshrc:5: job table full or recursion limit exceeded
/home/david/.zshrc:12: job table full or recursion limit exceeded
/home/david/.zshrc:13: job table full
/mnt/c/Users/David/Documents/Configs/zsh/.oh-my-zsh/oh-my-zsh.sh:14: job table full
is-at-least:23: job table full or recursion limit exceeded
/mnt/c/Users/David/Documents/Configs/zsh/.oh-my-zsh/lib/completion.zsh:20: job table full or recursion limit exceeded
/mnt/c/Users/David/Documents/Configs/zsh/.oh-my-zsh/lib/grep.zsh:9: job table full or recursion limit exceeded
/mnt/c/Users/David/Documents/Configs/zsh/.oh-my-zsh/lib/history.zsh:26: job table full or recursion limit exceeded
/mnt/c/Users/David/Documents/Configs/zsh/.oh-my-zsh/lib/key-bindings.zsh:14: job table full or recursion limit exceeded
/mnt/c/Users/David/Documents/Configs/zsh/.oh-my-zsh/lib/misc.zsh:5: job table full or recursion limit exceeded
colors:4: job table full or recursion limit exceeded
is_plugin:1: job table full or recursion limit exceeded
is-at-least:31: job table full or recursion limit exceeded
/mnt/c/Users/David/Documents/Configs/zsh/.oh-my-zsh/lib/completion.zsh:23: job table full or recursion limit exceeded
grep-flag-available:1: job table full or recursion limit exceeded
/mnt/c/Users/David/Documents/Configs/zsh/.oh-my-zsh/lib/misc.zsh:6: job table full or recursion limit exceeded
/mnt/c/Users/David/Documents/Configs/zsh/.oh-my-zsh/lib/theme-and-appearance.zsh:31: job table full or recursion limit exceeded
compinit:77: job table full or recursion limit exceeded
is-at-least:33: job table full or recursion limit exceeded
grep-flag-available:1: job table full
/mnt/c/Users/David/Documents/Configs/zsh/.oh-my-zsh/lib/misc.zsh:7: job table full or recursion limit exceeded
/mnt/c/Users/David/Documents/Configs/zsh/.oh-my-zsh/lib/theme-and-appearance.zsh:35: job table full
compinit:83: job table full or recursion limit exceeded
/mnt/c/Users/David/Documents/Configs/zsh/.oh-my-zsh/lib/misc.zsh:7: job table full or recursion limit exceeded
(eval):1: job table full or recursion limit exceeded
compinit:506: job table full or recursion limit exceeded
is-at-least:23: job table full or recursion limit exceeded
compaudit:176: job table full or recursion limit exceeded
compinit:509: job table full or recursion limit exceeded
is-at-least:31: job table full or recursion limit exceeded

I'm sorry if I don't exactly understand what is going on here. I'm a designer learning to code so I've tried to provide as much information as I could. Is there a reason why j won't work with my set up? I know some Ubuntu things don't work on WSL despite being WSL but my understand is that only relates to gpu related aspects of the kernel or something like that.

For people who test the return value of the previous command to alter their prompt, ```eval "$(jump shell bash)" will clobber that return value so that every command seems to exit successfully.

Swapping the order:

[[ "$PROMPT_COMMAND" =~ __jump_prompt_command ]] || {
  PROMPT_COMMAND="$PROMPT_COMMAND;__jump_prompt_command"
}
# instead of
[[ "$PROMPT_COMMAND" =~ __jump_prompt_command ]] || {
  PROMPT_COMMAND="__jump_prompt_command;$PROMPT_COMMAND"
}

seems to fix the issue without changing the behavior as far as I can tell.

Hi,

when installing on an M1 Mac via Homebrew, I get

% brew install jump
Warning: Treating jump as a formula. For the cask, use homebrew/cask/jump
Error: Cannot install under Rosetta 2 in ARM default prefix (/opt/homebrew)!
To rerun under ARM use:
    arch -arm64 brew install ...
To install under x86_64, install Homebrew into /usr/local.

When reading the README.md file it says that I should append eval "$(jump shell)" into my ~/.zshrc file.
Yet, doing it and closing and opening the terminal does nothing, as if it wasn't done.

And Jump suggests that I should run jump shell for Integrations. After doing so, it says:

# Put the line below in ~/.zshrc:
#
#   eval "$(jump shell zsh)"
#
# The following lines are autogenerated:

__jump_chpwd() {
  jump chdir
}

jump_completion() {
  reply="'$(jump hint "$@")'"
}

j() {
  local dir="$(jump cd $@)"
  test -d "$dir" && cd "$dir"
}

typeset -gaU chpwd_functions
chpwd_functions+=__jump_chpwd

compctl -U -K jump_completion j

Should the documentation say -> add this eval "$(jump shell zsh)" instead? Or run jump shell the first time and follow the instructions?

Have I missed anything? Thank you!

Thank you for your great project!
I have a request.

https://github.com/gsamokovarov/jump/releases/tag/v0.41.0

There is no binary of jump v0.41.0 for macOS.

On the other hand, there are binaries for macOS in v0.40.0 or earlier.

https://github.com/gsamokovarov/jump/releases/tag/v0.40.0

Could you release the binary of v0.41.0 for macOS?

For some reason, jump no longer learns about the directories I visit.

Running zsh-5.8-1, and jump r277.56fe5f8-1. eval "$(jump shell)" is present in .zshrc.

if i enter a path that doesn't exist in the jump db, i'd rather not jump, than jump to a seemingly random directory.

eg j xyz -> jump to /home/james/dl ???

Not sure if it is possible.

This happens on two computers of mine.

$ sudo dpkg -i ./jump_0.23.0_amd64.deb 
tar: Archive is compressed. Use -j option
tar: Error is not recoverable: exiting now
dpkg-deb: error: tar subprocess returned error exit status 2
dpkg: error processing archive ./jump_0.23.0_amd64.deb (--install):
dpkg-deb --control subprocess returned error exit status 2
Errors were encountered while processing:
 ./jump_0.23.0_amd64.deb

While the same command succed with v0.22.0

$ sudo dpkg -i ./jump_0.22.0_amd64.deb 
(Reading database ... 96811 files and directories currently installed.)
Preparing to unpack ./jump_0.22.0_amd64.deb ...
Unpacking jump (0.22.0) over (0.22.0) ...
Setting up jump (0.22.0) ...
Processing triggers for man-db (2.8.3-2ubuntu0.1) ...

My OS information:
$ cat /etc/os-release
NAME="Ubuntu"
VERSION="18.04.2 LTS (Bionic Beaver)"
ID=ubuntu
ID_LIKE=debian
PRETTY_NAME="Ubuntu 18.04.2 LTS"
VERSION_ID="18.04"
HOME_URL="https://www.ubuntu.com/"
SUPPORT_URL="https://help.ubuntu.com/"
BUG_REPORT_URL="https://bugs.launchpad.net/ubuntu/"
PRIVACY_POLICY_URL="https://www.ubuntu.com/legal/terms-and-policies/privacy-policy"
VERSION_CODENAME=bionic
UBUNTU_CODENAME=bionic

The tag OS in the file jump-0.30.1-1.x86_64.rpm, as reported by,
rpm -qp --qf '%{OS}\n' *.rpm
says darwin, which makes it impossible to install on a regular Fedora Linux.

I just noticed this and thought I'd mention it:

➜ sudo apt install ./jump_0.41.0_amd64.deb
➜ jump
Failed to execute process '/usr/bin/jump'. Reason:
exec: unknown error (errno was 8)
The file '/usr/bin/jump' is marked as an executable but could not be run by the operating system.
➜  file '/usr/bin/jump'
/usr/bin/jump: Mach-O 64-bit x86_64 executable

Is it possible that the wrong executable slipped into the package?

Thanks for your great tool!

I'm seeing this error in the terminal, whenever I run any command.

err: json: cannot unmarshal string into Go value of type scoring.Entries

Is it by design that the windows binary is not published for the latest couple of releases ?

I've tried building from source on windows, and everything seems to work fine. It would be easier for windows users to grab the binary directly, instead of building from source.

It's super easy to build cross-platform binaries in Go, and makes it really easy to download a command as a user.

There is an old AUR package here: https://aur.archlinux.org/packages/jump-git

But it builds from source and hence requires Go toolchain to be setup on your machine.

The release page only has .deb & .rpm files. If it also had the raw binary itself, then I could create an AUR package (it's pretty easy!)

So this kind of depends on #16

Please support https://wiki.archlinux.org/index.php/XDG_Base_Directory by default if $XDG_DATA_HOME is set or provide environment key to modify path to database.

Thanks

brew install jump installs https://jumpdesktop.com/

I mount and unmount drives regularly, so I would like to disable automatic cleaning, since the directory is probably just not mounted.

autojump used to have a jc shortcut to only jump to children of the current directories.

In #8 it was suggested that we could use / to distinguish among two parents, but I feel that if I'm in the current directory, I shouldn't need to distinguish it!

Having a command that only looks into the children of current seems to make more sense.

What are your thoughts?

Hi,

Big fan of jump. I think it's great.

It's also a bit hard to find, since it's name is so similar to autojump.

I think it's also worth getting some real branding work done. I'm a designer who has switched to code, but I'm still very much interested in design and one of the things I'm trying to do to help give back to the open source community is through a project I'm working on called Open Source Branding: https://github.com/jcklpe/Open-Source-Branding

I'm specifically working on this stuff through a repo and using github issues to format things because I'm interested in using this project as a means of demonstrating traditional, decentralized developer tools repurposed for design work. Please consider opening a ticket here and I'd love to help out anyway I can! : https://github.com/jcklpe/Open-Source-Branding/issues/new?template=branding-request.md

Hi there,

I like your tool a lot, so it would be nice, to have this on my raspi too. It's a raspberry 3B+ which running dietpi:

dietpi@DietPi:~$ uname -a
Linux DietPi 4.14.79-v7+ #1159 SMP Sun Nov 4 17:50:20 GMT 2018 armv7l GNU/Linux

root@DietPi:/var/# dpkg --print-architecture
armhf

Would be great, if you can provide a Debian package for armhf architecture or at least try to explain me how I get it compiled.

Thanks Micha

brew install jump
==> Downloading https://mirror.jumpdesktop.com/downloads/jdm/JumpDesktopMac-8.8.28.zip
Already downloaded: /Users/username/Library/Caches/Homebrew/downloads/17dfbdd22426b7ad2cec567f30e1f8fdb2a8cb6f41d3b0d7f647f3cec8284d4d--JumpDesktopMac-8.8.28.zip
==> Installing Cask jump
==> Moving App 'Jump Desktop.app' to '/Applications/Jump Desktop.app'
🍺  jump was successfully installed!

Interestingly, official brew docs also suggest this exact command: https://formulae.brew.sh/formula/jump

The installed application: https://formulae.brew.sh/cask/jump-desktop-connect#default

Brew info:

➜  ~ brew -v
Homebrew 3.3.15
Homebrew/homebrew-core (no Git repository)
Homebrew/homebrew-cask (git revision a497e739b2; last commit 2022-02-21)

Hi,

inspired by https://github.com/fdw/ranger-autojump I want to propose a jump ranger plugin that you can add to the jump README or to a jump wiki:

import ranger.api
import subprocess
from ranger.api.commands import *

class j(Command):
    """:j

    Uses jump to set the current directory.
    """

    def execute(self):
        directory = subprocess.check_output(["jump", "cd", self.arg(1)])
        directory = directory.decode("utf-8", "ignore")
        directory = directory.rstrip('\n')
        self.fm.execute_console("cd " + directory)

Just drop this jump.py in the ranger plugin directory and eventually add a keybinding like map cj console j%space to ranger's rc.conf and enjoy jumping from ranger.

My scores.json file is here: https://gist.github.com/60c5f00c08b5769ed99152dbdef9a3da

Some sample lines from the file: (notice how bitcoin has pretty higher score)

{ "Path": "/home/dufferzafar/dev/bitcoin", "Score": {"Weight":48, "Age":"2018-09-03T15:37:08.64200953+05:30"} },
{ "Path": "/home/dufferzafar/.apps/OpenSubtitlesDownload", "Score": {"Weight":14, "Age":"2018-09-03T15:37:08.64200953+05:30"} },
{ "Path": "/home/dufferzafar/dev/@yard/@mtech/sem-1/github-analytics", "Score": {"Weight":14, "Age":"2018-09-03T15:37:08.64200953+05:30"} },

Now,

jump cd bit matches with: /home/dufferzafar/.apps/OpenSubtitlesDownload

jump run btc matches with: /home/dufferzafar/dev/@yard/@mtech/sem-1/github-analytics

I expected both of these to match with bitcoin 😞

I just accidentally fill my disk and crashes jump. It writes a partial scores.json file and all subsequent calls ending up with ❯ err: unexpected EOF

Hi, it seems after using the temp file renaming, jump clean stops working.

Hey, I currently use joelthelion/autojump but don't really like its matching algorithm.

There's also rupa/z, but I haven't used it in a while so don't remember its algorithm.

I was wondering if you have used those tools and can provide some insight as to what the differences are?

This would help people decide which one they want to use.

Thanks :)

Feature Request: Evict last jump.

I find myself running j target-dir; j all too often, and would like the ability to evict the first result returned by j target-dir so that I don't encounter it again in future.

When I put

eval "$(jump shell)" in my ~/.zshrc file, I get zsh: command not found: No error.

Any ideas?

Coming from j being aliased to fasd_cd, I'm trying out Jump to see if I like it more. I'm used to sometimes typing j someth<tab> to quickly go to a directory called "something" when there are multiple on my system and I don't want to think beforehand what the correct search terms would be.

It seems this was a use-case for Jump in the past but it has been removed. Was that a design decision?

Thank you for your great project!
I have a request.

Could you release assets for Linux?

I found assets for Linux aren't released at v0.50.0.

https://github.com/gsamokovarov/jump/releases/tag/v0.50.0

On the other hand, assets for Linux has been released at v0.41.0.

https://github.com/gsamokovarov/jump/releases/tag/v0.41.0

jump_linux_amd64_binary

autojump database is a simple (easily grep-able) text file located at: ~/.local/share/autojump/autojump.txt

24.4948974278	/home/dufferzafar/dev/massren
14.1421356237	/home/dufferzafar/dev/tagtool

jump's database is a JSON file located at: ~/.jump/scores.json

[
    {"Path":"/home/dufferzafar/dev","Score":{"Weight":1,"Age":"2018-09-03T15:37:08.64200953+05:30"}}
]

Creating an importor seems pretty straightforward - only the Age field is new.

What would happen if all entries have the same Age? Will it negatively impact somehow?

Jump 0.19.0 received five out of five stars in a recent review in Linux-Magazin 07/2018. The following URL provides a paywalled link (priced at €0.99) to that review:

• http://www.linux-magazin.de/ausgaben/2018/07/tooltipps-5/

With best wishes, Robbie

Does jump support garbage collection?

I have to use the command which like below:

status --is-interactive; and source (jump shell fish | psub)

The default location of the z database file can be changed with the $Z_DATA environment variable.

jump import will ignore this setting and is unable to find the file in non-default locations.

I want to implement the cd command with python, but I don't know how to modify the working directory of the parent process (shell).😂

Ok, I'm not sure about how to explain this issue, because I have no background.

But, I'll try:

I just installed via brew install jump and then echo 'eval "$(jump shell zsh)"' >> ~/.zshrc

Then I tried with: exec -l $SHELL to replace my current shell and go to a directory j SomeDirectory_A

Nothing happend... I'm still at my home directory.

So... maybe the exec didn't make the magic. I closed all the tabs and terminal.

Open a new one.

j SomeOtherDirectoryAtMyHome_B

Nothing.

If I check: which j

j () {
    local dir="$(jump cd $@)"
    test -d "$dir" && cd "$dir"
}

and if I try: jump cd SomeDirectory_A ... nothing happens.

and then, I tried again: j SomeOtherDirectoryAtMyHome_B and then it moved me to SomeDirectory_A

I really want to use your utillity, because your idea looks great.

I would suggest prioritizing child directories or provide something like jc (as with autojump) to jump only within the children of the current directory.

Also, thanks for this, I'm liking it.

Hi,

Thanks for the awesome J project! Any plans to release pre-build binaries for the latest v0.51.0 ?

Cheers,
PY

Hi, this update breaks all my scripts related to jump top. Could you make an option to print the additional score column and default to the old behavior?

Hi! Great Utility!

Just a small typo on Line 39 of jump/cmd/chdir.go... search for 'scrore'

steve