fish-shell/share/functions/__fish_print_hostnames.fish
Wilke Schwiedop d9ee5d3863 fix trailing comments in __fish_print_hostnames
/etc/hosts specifies, that everything after a #-character is to be
treated as a comment. The current __fish_print_hostnames however only
considers #-characters at the beginning of a line.
Thus the comment from following valid hosts-entry would end up in the
completion output:

1.2.3.4  myhost # examplecomment

getent hosts properly handles comments.
2022-07-17 17:39:58 -05:00

137 lines
6.2 KiB
Fish

function __fish_print_hostnames -d "Print a list of known hostnames"
# This function used to primarily query `getent hosts` and only read from `/etc/hosts` if
# `getent` did not exist or `getent hosts` failed, based off the (documented) assumption that
# the former *might* return more hosts than the latter, which has never been officially noted
# to be the case. As `getent` is several times slower, involves shelling out, and is not
# available on some platforms (Cygwin and at least some versions of macOS, such as 10.10), that
# order is now reversed and `getent hosts` is only used if the hosts file is not found at
# `/etc/hosts` for portability reasons.
begin
test -r /etc/hosts && read -z </etc/hosts | string replace -r '#.*$' ''
or type -q getent && getent hosts 2>/dev/null
end |
# Ignore own IP addresses (127.*, 0.0[.0[.0]], ::1), non-host IPs (fe00::*, ff00::*),
# and leading/trailing whitespace. Split results on whitespace to handle multiple aliases for
# one IP.
string replace -irf '^\s*?(?!(?:0\.|127\.|ff0|fe0|::1))\S+\s*(.*?)\s*$' '$1' |
string split ' '
# Print nfs servers from /etc/fstab
if test -r /etc/fstab
string match -r '^\s*[0-9]{1,3}\.[0-9]{1,3}\.[0-9]{1,3]:|^[a-zA-Z\.]*:' </etc/fstab |
string replace -r ':.*' ''
end
# Check hosts known to ssh.
# Yes, seriously - the default specifies both with and without "2".
# Termux puts these in the android data directory if not rooted.
# The directory is available as $PREFIX/etc, but that variable name is so generic that
# it would cause false-positives.
# Also, some people might use /usr/local/etc.
set -l known_hosts ~/.ssh/known_hosts{,2} \
{/data/data/com.termux/files/usr,/usr/local,}/etc/ssh/{,ssh_}known_hosts{,2}
# Check default ssh configs.
set -l ssh_config ~/.ssh/config
# Inherit settings and parameters from `ssh` aliases, if any
if functions -q ssh
# Get alias and commandline options.
set -l ssh_func_tokens (functions ssh | string match '*command ssh *' | string split ' ')
set -l ssh_command $ssh_func_tokens (commandline -cpo)
# Extract ssh config path from last -F short option.
if contains -- -F $ssh_command
set -l ssh_config_path_is_next 1
for token in $ssh_command
if contains -- -F $token
set ssh_config_path_is_next 0
else if test $ssh_config_path_is_next -eq 0
set ssh_config (eval "echo $token")
set ssh_config_path_is_next 1
end
end
end
end
# Extract ssh config paths from Include option
function _ssh_include --argument-names ssh_config
# Relative paths in Include directive use /etc/ssh or ~/.ssh depending on
# system or user level config. -F will not override this behaviour
set -l relative_path $HOME/.ssh
if string match '/etc/ssh/*' -- $ssh_config
set relative_path /etc/ssh
end
function _recursive --no-scope-shadowing
set -l paths
for config in $argv
if test -r "$config" -a -f "$config"
set paths $paths (
# Keep only Include lines and remove Include syntax
string replace -rfi '^\s*Include\s+' '' <$config \
# Normalize whitespace
| string trim | string replace -r -a '\s+' ' ')
end
end
set -l new_paths
for path in $paths
set -l expanded_path
# Scope "relative" paths in accordance to ssh path resolution
if string match -qrv '^[~/]' $path
set path $relative_path/$path
end
# Use `eval` to expand paths (eg ~/.ssh/../test/* to /home/<user>/test/file1 /home/<user>/test/file2),
# and `set` will prevent "No matches for wildcard" messages
eval set expanded_path $path
for path in $expanded_path
# Skip unusable paths.
test -r "$path" -a -f "$path"
or continue
echo $path
set new_paths $new_paths $path
end
end
if test -n "$new_paths"
_recursive $new_paths
end
end
_recursive $ssh_config
end
set -l ssh_configs /etc/ssh/ssh_config (_ssh_include /etc/ssh/ssh_config) $ssh_config (_ssh_include $ssh_config)
for file in $ssh_configs
if test -r $file
# Don't read from $file twice. We could use `while read` instead, but that is extremely
# slow.
read -alz -d \n contents <$file
# Print hosts from system wide ssh configuration file
# Multiple names for a single host can be given separated by spaces, so just split it explicitly (#6698).
string replace -rfi '^\s*Host\s+(\S.*?)\s*$' '$1' -- $contents | string split " " | string match -rv '[\*\?]'
# Also extract known_host paths.
set known_hosts $known_hosts (string replace -rfi '.*KnownHostsFile\s*' '' -- $contents)
end
end
# Read all files and operate on their combined content
for file in $known_hosts
if test -r $file
read -z <$file
end
end |
# Ignore hosts that are hashed, commented or @-marked and strip the key
# Handle multiple comma-separated hostnames sharing a key, too.
#
# This one regex does everything we need, finding all matches including comma-separated
# values, but fish does not let us print only a capturing group without the entire match,
# and we can't use `string replace` instead (because CSV then fails).
# string match -ar "(?:^|,)(?![@|*!])\[?([^ ,:\]]+)\]?"
#
# Instead, manually piece together the regular expressions
string match -v -r '^\s*[!*|@#]' | string replace -rf '^\s*(\S+) .*' '$1' |
string split ',' | string replace -r '\[?([^\]]+).*' '$1'
return 0
end