vim-syncr

Ⅰ. 插件描述

Vim plugin to sync local and remote files and directories.

Ⅱ. 基本信息

创建日期：  2014-03-01
使用用户：  4
Github星：  8
插件作者：  Jacob Malcom

Ⅲ. 安装方法

使用Vundle管理器安装

在你的.vimrc下添加: 
Plugin 'jacob-ogre/vim-syncr'
… 然后在Vim中运行以下命令: 
:source %
:PluginInstall

对于Vundle版本 < 0.10.2，请用上面的Bundle替换Plugin。

使用NeoBundle管理器安装

在你的.vimrc下添加: 
NeoBundle 'jacob-ogre/vim-syncr'
… 然后在Vim中运行以下命令: 
:source %
:NeoBundleInstall

使用VimPlug管理器安装

在你的.vimrc下添加: 
Plug 'jacob-ogre/vim-syncr'
… 然后在Vim中运行以下命令: 
:source %
:PlugInstall

使用Pathogen管理器安装

在终端中运行以下命令: 
cd ~/.vim/bundle
git clone https://github.com/jacob-ogre/vim-syncr

Ⅳ. 文档说明

# vim-syncr

A small plugin to sync local and remote files and directories from Vim.

What?

vim-syncr currently includes just three functions:

  • sync current file local -> remote = :Suplfil
  • sync current dir local -> remote = :Supldir
  • sync current file remote -> local = :Sdownlf

that are mapped by default to:

  • (sync up file) -> <leader>sf
  • (sync up directory) -> <leader>sd
  • (sync down file) -> <leader>sdf

Why?

Our workflow looks something like this:

          worker A   worker B ...
               \       /
        --------------------------
        |          |             |
     remote A   remote B      remote C

Each of the remotes is suited to a particular type of analysis, e.g., remote A
is a super computing cluster where 1000s of nodes can be requested for large
analyses but can have long wait times both for the analysis to start and can
become very laggy when trying to edit remotely. Remotes B and C vary in their
number of processors, RAM, etc., but allow short time-to-run for other analyses.
As a result of these differences it's normal to jump from remote to remote many
times per day, but we want the current code at each remote. While we could set
each remote with its own git repo--commit/push/pull/fetch/etc. is fine, but
takes additional steps--it can lead to silly mistakes when one thinks they're
on version B when they forgot to pull and actually ran version A. Allegedly.
Because I've never done that. Maybe once. In addition, connectivity to B and C
can be a little wonky, again leading to laggy coding.

Our solution is that each person in the group has a single git repo on their
local computer and a copy of the reposity is maintained on their accounts on the

  1. Some folks use Vim, some use Sublime Text, but in both cases we just
  2. the local changes to the remote. SFTP exists for Sublime Text and works

very well, ending up with something like:

               GitHub repo --------- other worker's local repos...
                   |
             local git repo
                   |
        --------------------------
        |          |             |
     remote A   remote B      remote C

That is, the remote copies are not git repos, just sync'd versions of the local
git repos. Vim-syncr is not as full-featured as Sublime Text SFTP, but it works
for the Vimmers in the group.

How?

As one might guess, vim-syncr uses rsync to perform synchronization between a
local repository and a remote copy. The remote machine information is contained
in a configuration file saved at the root of the "project" directory; see
.syncrconf for an example. The current bash functions that drive the sync
operations are hard-coded to exclude certain files (see TODO, below). Note that
this uses the -c flag of rsync to checksum the files rather than just comparing
time stamps.

Setup

Basic operation

  1. Set up SSH login on the remote machine, if possible; otherwise, you will have
    to enter your password for each sync.
  2. Use [Vundle] (https://github.com/gmarik/Vundle.vim) to clone vim-syncr into
    ~/.vim/bundle, or where ever you have your plugins set up. Manual cloning

will work just fine as well.

  1. Copy the .syncrconf from ~/.vim/bundle/vim-syncr/ to the root directory of
    your active repository(ies).
  2. Edit the .syncrconf as necessary to make the connections work.
  3. If you work with the same repo on multiple remote machines, make additional
    copies of .syncrconf but named as, e.g., syncrconf_server2 and edit the

settings as needed. The serv_switch script (see below) permits easy
switching between remote machines from within vim, and will rename the files
given the "name" entry in .syncrconf.

  1. (optional) Add autocmd BufWritePost * :Suplfil to your .vimrc to enable
    automatic upload to the remote server on each :write. Also works with :Gw if

using [Fugitive] (https://github.com/tpope/vim-fugitive).

Switching the remote target

  1. Copy serv_switch from ~/.vim/bundle/vim-syncr to the root of your working
    repository, which will enable simple switching between remote machines.
  2. (optional) add map <Leader>sw :!/path/to/work/repo/serv_switch<cr> to your
    .vimrc to make remote server switching easy.

Usage

Basic operations

  1. Open and edit a file in your working repository, then :w, and last <leader>sf.
  2. Alternately, alter multiple files in a directory, :w, and <leader>sd to
    sync the entire directory (i.e., the parts that have changed).
  3. An entire repository can be synced by opening a file at the working
    repository root (e.g., .syncrconf), then using <leader>sd to perform a

recursive sync.

Switching the remote target

  1. If you added <Leader>sw to your .vimrc, then and if ',' is your <Leader>,
    ,sw will spawn a dialog providing the "name" field of all .syncrconf*

files in the repo root.

  1. Choose the number of the remote machine you want to which you want to sync
    and the change will be made.

Other options?

sftp can also do the job, particularly for single files, and you might check out
[vim-sftp] (https://github.com/hesselbom/vim-hsftp) - from which the config
parsing in vim-syncr was derived! - to see if it better fits your needs. However,
when working in a large repository where only a small number of files (and/or a
small portion of a few files) might change, syncing tends to be faster. There's
also a little more control (and highlighting!) using vim-syncr...but more could
be done to make it even more flexible.

TODO (in general order of importance)

  • Add rsync --exclude items to config file for more flexibility than the
    hard-coded excludes currently in-place.
  • Other things, I'm sure...

添加新评论