[bash,zsh] Work around a quirk of macOS awk

macOS awk is a variant of nawk, but it contains a unique patch for the
UTF-8 support.  However, this patch causes the problem.  If the input
contains any non-UTF-8 data, macOS awk stops processing and does not
do anything, instead of ignoring the unrecognized data and continue
the processing.  However, the contents of the ssh configuration and
/etc/hosts is not under the control of fzf, so we cannot fix the input
when those files contain non-UTF-8 data.  To work around this
behavior, one can set the locale to LC_ALL=C to treat the input data
with the plain 8-bit encoding.

4 files changed, 119 insertions, 21 deletions

diff --git a/shell/completion.bash b/shell/completion.bash
index 85e269aa..b62cf004 100644
--- a/shell/completion.bash
+++ b/shell/completion.bash
@@ -39,6 +39,32 @@ __fzf_defaults() {
   echo "${FZF_DEFAULT_OPTS-} $2"
 }
 
+# This function performs `exec awk "$@"` safely by working around awk
+# compatibility issues.
+#
+# Note: To reduce an extra fork, this function performs "exec" so is expected
+# to be run as the last command in a subshell.
+#
+# Note: This function is included with {completion,key-bindings}.{bash,zsh} and
+# synchronized.
+__fzf_exec_awk() {
+  if [[ -z ${__fzf_awk-} ]]; then
+    __fzf_awk=awk
+
+    # choose the faster mawk if: it's installed && build date >= 20230322 &&
+    # version >= 1.3.4
+    local n x y z d
+    IFS=' .' read n x y z d <<< $(command mawk -W version 2> /dev/null)
+    [[ $n == mawk ]] && (( d >= 20230302 && (x * 1000 + y) * 1000 + z >= 1003004 )) && __fzf_awk=mawk
+  fi
+
+  # Note: macOS awk has a quirk that it stops processing at all when it sees
+  # any data not following UTF-8 in the input stream when the current LC_CTYPE
+  # specifies the UTF-8 encoding.  To work around this quirk, one needs to
+  # specify LC_ALL=C to change the current encoding to the plain one.
+  LC_ALL=C exec "$__fzf_awk" "$@"
+}
+
 __fzf_comprun() {
   if [[ "$(type -t _fzf_comprun 2>&1)" = function ]]; then
     _fzf_comprun "$@"
@@ -364,7 +390,7 @@ _fzf_complete() {
   fi
 
   local cur selected trigger cmd post
-  post="$(caller 0 | command awk '{print $2}')_post"
+  post="$(caller 0 | __fzf_exec_awk '{print $2}')_post"
   type -t "$post" > /dev/null 2>&1 || post='command cat'
 
   trigger=${FZF_COMPLETION_TRIGGER-'**'}
@@ -443,7 +469,7 @@ _fzf_proc_completion() {
 }
 
 _fzf_proc_completion_post() {
-  command awk '{print $2}'
+  __fzf_exec_awk '{print $2}'
 }
 
 # To use custom hostname lists, override __fzf_list_hosts.
@@ -475,7 +501,7 @@ if ! declare -F __fzf_list_hosts > /dev/null; then
         shopt -u dotglob nocaseglob failglob
         shopt -s nullglob
 
-        command awk '
+        __fzf_exec_awk '
           tolower($1) ~ /^host(name)?$/ {
             for (i = 2; i <= NF; i++)
               if ($i !~ /[*?%]/)
@@ -484,7 +510,7 @@ if ! declare -F __fzf_list_hosts > /dev/null; then
         ' ~/.ssh/config ~/.ssh/config.d/* /etc/ssh/ssh_config 2> /dev/null
       ) \
       <(
-        command awk -F ',' '
+        __fzf_exec_awk -F ',' '
           match($0, /^[[a-z0-9.,:-]+/) {
             $0 = substr($0, 1, RLENGTH)
             gsub(/\[/, "")
@@ -494,7 +520,7 @@ if ! declare -F __fzf_list_hosts > /dev/null; then
         ' ~/.ssh/known_hosts 2> /dev/null
       ) \
       <(
-        command awk '
+        __fzf_exec_awk '
           /^[[:blank:]]*(#|$)|0\.0\.0\.0/ { next }
           {
             sub(/#.*/, "")
@@ -523,7 +549,7 @@ _fzf_complete_ssh() {
     *)
       local user=
       [[ "$2" =~ '@' ]] && user="${2%%@*}@"
-      _fzf_complete +m -- "$@" < <(__fzf_list_hosts | command awk -v user="$user" '{print user $0}')
+      _fzf_complete +m -- "$@" < <(__fzf_list_hosts | __fzf_exec_awk -v user="$user" '{print user $0}')
       ;;
   esac
 }
diff --git a/shell/completion.zsh b/shell/completion.zsh
index 3fbdbb20..d59796bc 100644
--- a/shell/completion.zsh
+++ b/shell/completion.zsh
@@ -104,6 +104,32 @@ __fzf_defaults() {
   echo -E "${FZF_DEFAULT_OPTS-} $2"
 }
 
+# This function performs `exec awk "$@"` safely by working around awk
+# compatibility issues.
+#
+# Note: To reduce an extra fork, this function performs "exec" so is expected
+# to be run as the last command in a subshell.
+#
+# Note: This function is included with {completion,key-bindings}.{bash,zsh} and
+# synchronized.
+__fzf_exec_awk() {
+  if [[ -z ${__fzf_awk-} ]]; then
+    __fzf_awk=awk
+
+    # choose the faster mawk if: it's installed && build date >= 20230322 &&
+    # version >= 1.3.4
+    local n x y z d
+    IFS=' .' read n x y z d <<< $(command mawk -W version 2> /dev/null)
+    [[ $n == mawk ]] && (( d >= 20230302 && (x * 1000 + y) * 1000 + z >= 1003004 )) && __fzf_awk=mawk
+  fi
+
+  # Note: macOS awk has a quirk that it stops processing at all when it sees
+  # any data not following UTF-8 in the input stream when the current LC_CTYPE
+  # specifies the UTF-8 encoding.  To work around this quirk, one needs to
+  # specify LC_ALL=C to change the current encoding to the plain one.
+  LC_ALL=C exec "$__fzf_awk" "$@"
+}
+
 __fzf_comprun() {
   if [[ "$(type _fzf_comprun 2>&1)" =~ function ]]; then
     _fzf_comprun "$@"
@@ -253,7 +279,7 @@ if ! declare -f __fzf_list_hosts > /dev/null; then
         # when no matching is found.
         setopt GLOB NO_DOT_GLOB CASE_GLOB NO_NOMATCH NULL_GLOB
 
-        command awk '
+        __fzf_exec_awk '
           tolower($1) ~ /^host(name)?$/ {
             for (i = 2; i <= NF; i++)
               if ($i !~ /[*?%]/)
@@ -262,7 +288,7 @@ if ! declare -f __fzf_list_hosts > /dev/null; then
         ' ~/.ssh/config ~/.ssh/config.d/* /etc/ssh/ssh_config 2> /dev/null
       ) \
       <(
-        command awk -F ',' '
+        __fzf_exec_awk -F ',' '
           match($0, /^[[a-z0-9.,:-]+/) {
             $0 = substr($0, 1, RLENGTH)
             gsub(/\[/, "")
@@ -272,7 +298,7 @@ if ! declare -f __fzf_list_hosts > /dev/null; then
         ' ~/.ssh/known_hosts 2> /dev/null
       ) \
       <(
-        command awk '
+        __fzf_exec_awk '
           /^[[:blank:]]*(#|$)|0\.0\.0\.0/ { next }
           {
             sub(/#.*/, "")
@@ -300,7 +326,7 @@ _fzf_complete_ssh() {
     *)
       local user
       [[ $prefix =~ @ ]] && user="${prefix%%@*}@"
-      _fzf_complete +m -- "$@" < <(__fzf_list_hosts | awk -v user="$user" '{print user $0}')
+      _fzf_complete +m -- "$@" < <(__fzf_list_hosts | __fzf_exec_awk -v user="$user" '{print user $0}')
       ;;
   esac
 }
@@ -358,7 +384,7 @@ _fzf_complete_kill() {
 }
 
 _fzf_complete_kill_post() {
-  awk '{print $2}'
+  __fzf_exec_awk '{print $2}'
 }
 
 fzf-completion() {
diff --git a/shell/key-bindings.bash b/shell/key-bindings.bash
index 8b0ef76c..946bc2d8 100644
--- a/shell/key-bindings.bash
+++ b/shell/key-bindings.bash
@@ -25,6 +25,32 @@ __fzf_defaults() {
   echo "${FZF_DEFAULT_OPTS-} $2"
 }
 
+# This function performs `exec awk "$@"` safely by working around awk
+# compatibility issues.
+#
+# Note: To reduce an extra fork, this function performs "exec" so is expected
+# to be run as the last command in a subshell.
+#
+# Note: This function is included with {completion,key-bindings}.{bash,zsh} and
+# synchronized.
+__fzf_exec_awk() {
+  if [[ -z ${__fzf_awk-} ]]; then
+    __fzf_awk=awk
+
+    # choose the faster mawk if: it's installed && build date >= 20230322 &&
+    # version >= 1.3.4
+    local n x y z d
+    IFS=' .' read n x y z d <<< $(command mawk -W version 2> /dev/null)
+    [[ $n == mawk ]] && (( d >= 20230302 && (x * 1000 + y) * 1000 + z >= 1003004 )) && __fzf_awk=mawk
+  fi
+
+  # Note: macOS awk has a quirk that it stops processing at all when it sees
+  # any data not following UTF-8 in the input stream when the current LC_CTYPE
+  # specifies the UTF-8 encoding.  To work around this quirk, one needs to
+  # specify LC_ALL=C to change the current encoding to the plain one.
+  LC_ALL=C exec "$__fzf_awk" "$@"
+}
+
 __fzf_select__() {
   FZF_DEFAULT_COMMAND=${FZF_CTRL_T_COMMAND:-} \
   FZF_DEFAULT_OPTS=$(__fzf_defaults "--reverse --walker=file,dir,follow,hidden --scheme=path" "${FZF_CTRL_T_OPTS-} -m") \
@@ -74,13 +100,7 @@ if command -v perl > /dev/null; then
   }
 else # awk - fallback for POSIX systems
   __fzf_history__() {
-    local output script n x y z d
-    if [[ -z $__fzf_awk ]]; then
-      __fzf_awk=awk
-      # choose the faster mawk if: it's installed && build date >= 20230322 && version >= 1.3.4
-      IFS=' .' read n x y z d <<< $(command mawk -W version 2> /dev/null)
-      [[ $n == mawk ]] && (( d >= 20230302 && (x *1000 +y) *1000 +z >= 1003004 )) && __fzf_awk=mawk
-    fi
+    local output script
     [[ $(HISTTIMEFORMAT='' builtin history 1) =~ [[:digit:]]+ ]]    # how many history entries
     script='function P(b) { ++n; sub(/^[ *]/, "", b); if (!seen[b]++) { printf "%d\t%s%c", '$((BASH_REMATCH + 1))' - n, b, 0 } }
     NR==1 { b = substr($0, 2); next }
@@ -90,7 +110,7 @@ else # awk - fallback for POSIX systems
     output=$(
       set +o pipefail
       builtin fc -lnr -2147483648 2> /dev/null |   # ( $'\t '<lines>$'\n' )* ; <lines> ::= [^\n]* ( $'\n'<lines> )*
-        command $__fzf_awk "$script"           |   # ( <counter>$'\t'<lines>$'\000' )*
+        __fzf_exec_awk "$script"               |   # ( <counter>$'\t'<lines>$'\000' )*
         FZF_DEFAULT_OPTS=$(__fzf_defaults "" "-n2..,.. --scheme=history --bind=ctrl-r:toggle-sort --wrap-sign '"$'\t'"↳ ' --highlight-line ${FZF_CTRL_R_OPTS-} +m --read0") \
         FZF_DEFAULT_OPTS_FILE='' $(__fzfcmd) --query "$READLINE_LINE"
     ) || return
diff --git a/shell/key-bindings.zsh b/shell/key-bindings.zsh
index 9c4215e0..51defeab 100644
--- a/shell/key-bindings.zsh
+++ b/shell/key-bindings.zsh
@@ -46,6 +46,32 @@ __fzf_defaults() {
   echo -E "${FZF_DEFAULT_OPTS-} $2"
 }
 
+# This function performs `exec awk "$@"` safely by working around awk
+# compatibility issues.
+#
+# Note: To reduce an extra fork, this function performs "exec" so is expected
+# to be run as the last command in a subshell.
+#
+# Note: This function is included with {completion,key-bindings}.{bash,zsh} and
+# synchronized.
+__fzf_exec_awk() {
+  if [[ -z ${__fzf_awk-} ]]; then
+    __fzf_awk=awk
+
+    # choose the faster mawk if: it's installed && build date >= 20230322 &&
+    # version >= 1.3.4
+    local n x y z d
+    IFS=' .' read n x y z d <<< $(command mawk -W version 2> /dev/null)
+    [[ $n == mawk ]] && (( d >= 20230302 && (x * 1000 + y) * 1000 + z >= 1003004 )) && __fzf_awk=mawk
+  fi
+
+  # Note: macOS awk has a quirk that it stops processing at all when it sees
+  # any data not following UTF-8 in the input stream when the current LC_CTYPE
+  # specifies the UTF-8 encoding.  To work around this quirk, one needs to
+  # specify LC_ALL=C to change the current encoding to the plain one.
+  LC_ALL=C exec "$__fzf_awk" "$@"
+}
+
 # CTRL-T - Paste the selected file path(s) into the command line
 __fzf_select() {
   setopt localoptions pipefail no_aliases 2> /dev/null
@@ -117,13 +143,13 @@ fzf-history-widget() {
       FZF_DEFAULT_OPTS=$(__fzf_defaults "" "-n2..,.. --scheme=history --bind=ctrl-r:toggle-sort --wrap-sign '\t↳ ' --highlight-line ${FZF_CTRL_R_OPTS-} --query=${(qqq)LBUFFER} +m --read0") \
       FZF_DEFAULT_OPTS_FILE='' $(__fzfcmd))"
   else
-    selected="$(fc -rl 1 | awk '{ cmd=$0; sub(/^[ \t]*[0-9]+\**[ \t]+/, "", cmd); if (!seen[cmd]++) print $0 }' |
+    selected="$(fc -rl 1 | __fzf_exec_awk '{ cmd=$0; sub(/^[ \t]*[0-9]+\**[ \t]+/, "", cmd); if (!seen[cmd]++) print $0 }' |
       FZF_DEFAULT_OPTS=$(__fzf_defaults "" "-n2..,.. --scheme=history --bind=ctrl-r:toggle-sort --wrap-sign '\t↳ ' --highlight-line ${FZF_CTRL_R_OPTS-} --query=${(qqq)LBUFFER} +m") \
       FZF_DEFAULT_OPTS_FILE='' $(__fzfcmd))"
   fi
   local ret=$?
   if [ -n "$selected" ]; then
-    if [[ $(awk '{print $1; exit}' <<< "$selected") =~ ^[1-9][0-9]* ]]; then
+    if [[ $(__fzf_exec_awk '{print $1; exit}' <<< "$selected") =~ ^[1-9][0-9]* ]]; then
       zle vi-fetch-history -n $MATCH
     else # selected is a custom query, not from history
       LBUFFER="$selected"