After profiling startup time, I found that "pyenv rehash" is
by far the slowest piece. This change skips rehashing on init.
See https://github.com/pyenv/pyenv/issues/784#issuecomment-404850327
Also pass the shell explicitly, as suggested in
https://github.com/pyenv/pyenv/issues/784#issuecomment-404944318,
which provides a modest improvement.
In total, this improves startup time from 1s to 0.3s on my machine.
Using the following command:
python -m timeit "__import__('subprocess').Popen(['zsh', '-i', '-c', 'echo']).communicate()"
Result before this change:
10 loops, best of 3: 1 sec per loop
Result after this change:
10 loops, best of 3: 334 msec per loop
This reverts commit 6d00fdf8c8.
As unfortunate as it is, this workaround was a hack and doesn't take
into account prompts like spaceship which don't use $editor_info but
rely on $KEYMAP directly. We'll need to find a more consistent solution
to fix this.
There are a number of things happening here.
- Extra support for yaourt has been removed
- Docs have been updated to explicitly call out that AUR helpers are not
officially supported
- aurutils has been suggested to make common operations easier
- A utility function called aurget (similar to aurfetch from aurutils)
has been added to make cloning AUR repos easier.
Fixes#1531
This was ported from Oh-My-Zsh and since have been disabled in it.
Explained in more detail
rvm/rvm/issues/3091#issuecomment-60083194
Related: #998, #1081
* Added conda virtualenv support to python module
* Added instructions for Python module options to README
Thanks to @egpbos for the original feature and @ickc for fixing the merge conflicts.
* archive: fix unrar check when using unrar-free
unrar-free returns the error code 1 when run without arguments, thus
failing the presence check. Replacing the current presence check with
(( $+commands[unrar] )) fixes the problem.
* archive: add unar support for lsarchive and unarchive
No `archive` support with unar.
* Add zsh-help function for easily searching the zsh documentation
Looks up things in the zsh documentation.
Usage: zsh-help [--all] search term(s)
Option --all will seach for the term anywhere, not just at the start of a
line. When not using --all it will search nicely for terms at the beginning
of the line, which in the zsh man pages is where terms that are explained
are located, allowing you to search the zsh man pages easily.
* Improve zsh-help to search section headings before other text
Provides a much easier way to search and access ZSH's manual. First checks for
terms at the start of the manual, then checks if it's at start of a line allowing
whitespace.
Clean up some of the code a bit and format it to have a proper header for the
zprezto project with author/email and description of the function.
Now that pyenv plugins availability is detected by directly probing
'pyenv', we need to rely on the same mechanism consistently.
Further, we perform available pyenv plugin scan with native zsh
techniques instead of relying on external commands.
If the pyenv virtualenv plugin is installed using the [pyenv-installer][1]
app script or directly via a [Git clone][2], then the pyenv-virtualenv-init
executable, that the `pyenv virtualenv-init` command uses, will not
exist in the user's PATH and therefore cannot be found using
`$commands[pyenv-virtualenv-init]`.
Installing the pyenv-virtualenv plugin in this manner is common among Linux users.
Using the pyenv `commands` command, which lists all commands pyenv can
run, we can find if the virtualenv-init command is available to pyenv
and subsequently the virtualenv plugin, without relying on
pyenv-virtualenv-init to exist in the user's PATH.
[1]: https://github.com/pyenv/pyenv-installer
[2]: https://github.com/pyenv/pyenv-virtualenv#installing-as-a-pyenv-plugin
This was a change in zsh 5.4.1 where prompt_sp is now reset on calls to
prompt, unlike previous behavior where it was set by default. This
restores the previous default behavior.
Refs #1423, but is not a complete fix because we need to wait for
external prompts to update as well.
Just having 'pyenv' available doesn't imply availability of virtualenv pyenv
plugins. Check for availability of virtualenv plugins as well before attempting
to use pyenv wrapper for virtualenv.
By keeping the variable `_etc_host_ignores` around, it can be resolved
lazily which is turn should allow selective ignores on a per invocation
basis.
Also, `'\#'` doesn't need to be added to `_etc_host_ignores` anymore.
It is enforced inline.
Entries from static '/etc/hosts' can now be ignored via 'zstyle' based
configuration. Both IP address and corresponding hostname will be ignored
during host completion. However, some of the entries ignored from '/etc/hosts'
still might appear during completion because of their presence in 'ssh'
configuration ('~/.ssh/config') or history ('~/.ssh/ssh_hosts',
'~/.ssh/known_hosts' etc.).
Changes:
* Simplify zstyle name `skip-virtualenvwrapper-init` to `initialize`
avoiding double negation in name
* Always perform `eval (pyenv virtualenv-init -)` at initialization
* Prefer `virtualenvwrapper_lazy` over `virtualenvwrapper` when available
* Honor `VIRTUALENVWRAPPER_VIRTUALENV` if it is defined.
* Document about `VIRTUALENVWRAPPER_PYTHON` and `VIRTUALENVWRAPPER_VIRTUALENV`
(this would be particularly important in macOS after recent homebrew update)
* Add additional documentation for `initialize` in _README.md_ and _zpreztorc_
* Add aliases `py2`, `py3` as shortcut for `python2`, `python3` respectively
This includes some improvements by @indrajitr in addition to the main migration.
The first step was to avoid PROMPT and RPROMPT modification when possible (which may help resolve some other issues as well relating to zsh crashes with the sorin prompt) then update the displayed git information in a separate variable rather than a command.
We use zsh-async for creating and running background tasks. The sorin prompt uses it to update git info without blocking the prompt from displaying (because of how long it can take). In the future it may be worth moving more tasks and more prompts to using this.
The move to zsh-async does make the git prompt slower in some circumstances (most noticeable in large repos), but this is a worthwhile tradeoff to avoid the cache file which had a number of potential security holes.
We have also switched to adding zsh-async as an external submodule (rather than the version bundled with pure) which may cause some migration headaches, but it will be worth it in the long run.
Previously I made a change to try and bind these unbound keys to the main
keymap in thoughts that in vicmd or viins mode it would fallback if it wasn't
bound in vicmd/viins mode. This turned out not to be the case.
Explicitly bind the keys in both viins and vicmd mode as works properly.
Changes:
- Rename `rsync_scp_wrap` to `noremoteglob` and make it more generally usable
- Enable completion support for commands wrapped with `noremoteglob`
- Tighten up internal variable usages
- Update documentation for 'noremoteglob' function
Changes:
* In prezto, function names are hyphenated ('-') by convention,
rename `promptpwd` to `prompt-pwd` accordingly.
* Unset local variable `current_pwd` proactively.
* Tweak additional documentation for `prompt-pwd`.
We cannot always assume availability of `pip`, it can be `pip2` or `pip3`
instead. We detect the first available one and use it for generating the
completion cache.
If you are in a tmux session, and auto-start is enabled
Then you `sudo su` and then `su <username>` it will re-attach, in a loop.
This resolves that problem, by forcing a detach before a (re)attach.
Currently rm/cp/ln/mv are aliased to rm/cp/ln/mv -i so that it will
prompt before removing files.
Some people would not like this functionality and wish for an easy way to
disable it. This adds a new option:
zstyle ':prezto:module:utility' safe-ops yes/no
It is enabled by default even if zstyle is not set, but can be set to no
to disable these aliases.
This should resolve issue #205
Somewhere around zsh 5.1, there was a change which turned typeset (and
most likely local) into a keyword. In older versions of zsh, using
`local x=()` will cause the () to be treated as a glob qualifier.
Fixes#1373
Usage of `(( ... ))` over `[[ ... ]]` is preferred for arithmetic expression
since the former is less error prone.
Also, unset local variable whenever possible.
Bind the remaining unbound keys in the main mode so that they
don't result in undefined functionality when pressed. (Often this
will change the casing of one/most/all of the characters you've
typed or even stranger things). In emacs mode this usually just
inserts a tilde, but this will fix that as well so nothing happens
when unbound keys are pressed.
Since there isn't any binding which does "nothing", create a noop
ZLE widget and bind them to that.
The user or other modules are still able to override these bindings
either by overriding the main keymap or setting a keymap in one
of the other modes which will take precedence over the main fallback.
Changes:
- Use git.io over 'https'
- Add extra check to conform to *.github.com URLs
- Use built-in _urls function for completion
- Update readme with GitHub blog URL
* [editor] Add bindkey-all function to show all bound keys
Function which allows you to see all the bound keys for all of
the different keymaps. It does accept arguments and passes them
through to bindkey -m "$keymap" so you could set a shortcut to
all keymaps as well if you wished.
Makes it much easier to see and grep what keys are bound to each of
the different keymaps.
When no arguments are given it will print keymap headers to stderr,
if given arguments it will not print anything extra to the screen.
* [editor] Update README to include information about bindkey-all funct.
Escape all $ except the first $. Escape all backtick `'s. This prevents
variable names or shell expansions placed as branch names from
remotely exploiting code.
Fixes issue #1267 for sorin prompt.
Adds a function which wraps rsync and scp so that remote paths are not globbed
but local paths are globbed. This is because the programs have their own
globbing for remote paths. The wrap function globs args starting in / and ./
and doesn't glob paths with : in it as these are interpreted as remote paths
by these programs unless the path starts with / or ./
Fixes issue #1125
Delete key deletes character in vimcmd cmd mode instead of weird
default functionality. The default functionality in vicmd mode
is the Delete key will change the case of many of the characters
on the screen, which is not the default thing that vim does.
This could be confusing and frustrating to users, so set it to delete
a character instead.
