From 115892ccd270cce4f6b1a90810d7bf2b5bb5b3a2 Mon Sep 17 00:00:00 2001
From: Fabian Homborg <FHomborg@gmail.com>
Date: Fri, 29 Nov 2019 20:09:02 +0100
Subject: [PATCH] alias: Use `read --tokenize`

This did some weird unescaping to try to extract the first word.

So we're now more likely to be *correct*, and the alias benchmark is
about 20% *faster*.

Call it a win-win.
---
 benchmarks/benchmarks/aliases.fish | 16 +++-------------
 share/functions/alias.fish         | 16 +++-------------
 tests/checks/alias.fish            |  2 ++
 3 files changed, 8 insertions(+), 26 deletions(-)

diff --git a/benchmarks/benchmarks/aliases.fish b/benchmarks/benchmarks/aliases.fish
index 0eb865842..e6a366685 100644
--- a/benchmarks/benchmarks/aliases.fish
+++ b/benchmarks/benchmarks/aliases.fish
@@ -44,22 +44,12 @@ function alias --description 'Creates a function wrapping a command'
         return 1
     end
 
-    # Extract the first command from the body. This is supposed to replace all non-escaped (i.e.
-    # preceded by an odd number of `\`) spaces with a newline so it splits on them. See issue #2220
-    # for why the following borderline incomprehensible code exists.
-    set -l tmp (string replace -ra -- "([^\\\ ])((\\\\\\\)*) " '$1\n' $body)
-    set first_word (string trim -- $tmp[1])
-    # If the user does something like `alias x 'foo; bar'` we need to strip the semicolon.
-    set base_command (string trim -c ';' -- $first_word)
-    if set -q tmp[2]
-        set body $tmp[2..-1]
-    else
-        set body
-    end
+    # Extract the first command from the body.
+    printf '%s\n' $body | read -lt first_word body
 
     # Prevent the alias from immediately running into an infinite recursion if
     # $body starts with the same command as $name.
-    if test $base_command = $name
+    if test $first_word = $name
         if contains $name (builtin --names)
             set prefix builtin
         else
diff --git a/share/functions/alias.fish b/share/functions/alias.fish
index 0651cc6ac..ea3c15863 100644
--- a/share/functions/alias.fish
+++ b/share/functions/alias.fish
@@ -44,22 +44,12 @@ function alias --description 'Creates a function wrapping a command'
         return 1
     end
 
-    # Extract the first command from the body. This is supposed to replace all non-escaped (i.e.
-    # preceded by an odd number of `\`) spaces with a newline so it splits on them. See issue #2220
-    # for why the following borderline incomprehensible code exists.
-    set -l tmp (string replace -ra -- "([^\\\ ])((\\\\\\\)*) " '$1\n' $body)
-    set first_word (string trim -- $tmp[1])
-    # If the user does something like `alias x 'foo; bar'` we need to strip the semicolon.
-    set base_command (string trim -c ';' -- $first_word)
-    if set -q tmp[2]
-        set body $tmp[2..-1]
-    else
-        set body
-    end
+    # Extract the first command from the body.
+    printf '%s\n' $body | read -lt first_word body
 
     # Prevent the alias from immediately running into an infinite recursion if
     # $body starts with the same command as $name.
-    if test $base_command = $name
+    if test $first_word = $name
         if contains $name (builtin --names)
             set prefix builtin
         else
diff --git a/tests/checks/alias.fish b/tests/checks/alias.fish
index fb3bb3f7c..569ee125e 100644
--- a/tests/checks/alias.fish
+++ b/tests/checks/alias.fish
@@ -10,9 +10,11 @@ my_alias
 
 alias a-2='echo "hello there"'
 
+alias foo '"a b" c d e'
 # Bare `alias` should list the aliases we have created and nothing else
 # We have to exclude two aliases because they're an artifact of the unit test
 # framework and we can't predict the definition.
 alias | grep -Ev '^alias (fish_indent|fish_key_reader) '
 # CHECK: alias a-2 'echo "hello there"'
+# CHECK: alias foo '"a b" c d e'
 # CHECK: alias my_alias 'foo; and echo foo ran'