From 449fed1c13525c8420b4366014ba34616cb20239 Mon Sep 17 00:00:00 2001
From: Kerin Millar <kfm@plushkava.net>
Date: Tue, 11 Jun 2024 06:26:41 +0100
Subject: Add the is_anyof() and is_subset() functions

Examples follow.

is_anyof  y x y z        # returns 0
is_anoyf  y x w z        # returns 1
is_subset x y   -- x y z # returns 0
is_subset x y z -- z y x # returns 0
is_subset x y   -- x w z # returns 1

Signed-off-by: Kerin Millar <kfm@plushkava.net>
---
 functions.sh   | 62 ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
 test-functions | 49 ++++++++++++++++++++++++++++++++++++++++++++++
 2 files changed, 111 insertions(+)

diff --git a/functions.sh b/functions.sh
index 7d146c8..166f184 100644
--- a/functions.sh
+++ b/functions.sh
@@ -446,6 +446,28 @@ is_int()
 	esac
 }
 
+#
+# Determines whether the first parameter matches any of the parameters that
+# follow it.
+#
+is_anyof()
+{
+	local arg needle
+
+	if [ "$#" -eq 0 ]; then
+		warn "is_anyof: too few arguments (got $#, expected at least 1)"
+	else
+		needle=$1
+		shift
+		for arg; do
+			if [ "${arg}" = "${needle}" ]; then
+				return
+			fi
+		done
+	fi
+	false
+}
+
 #
 # Takes the first parameter as a reference file/directory then determines
 # whether any of the following parameters refer to newer files/directories.
@@ -468,6 +490,46 @@ is_older_than()
 	| read -r _
 }
 
+#
+# Collects the intersection of the parameters up to - but not including - a
+# sentinel value then determines whether the resulting set is a subset of the
+# interection of the remaining parameters. If the SENTINEL variable is set and
+# non-empty, it shall be taken as the value of the sentinel. Otherwise, the
+# value of the sentinel shall be defined as <hyphen-dash><hyphen-dash>. If the
+# sentinel value is not encountered or if either set is empty then the returm
+# value shall be greater than 1.
+#
+is_subset()
+{
+	SENTINEL=${SENTINEL:-'--'} awk -f - -- "$@" <<-'EOF'
+		BEGIN {
+			argc = ARGC
+			ARGC = 1
+			for (i = 1; i < argc; i++) {
+				word = ARGV[i]
+				if (word == ENVIRON["SENTINEL"]) {
+					break
+				} else {
+					set1[word] = ""
+				}
+			}
+			if (i == 1 || argc - i < 2) {
+				exit 1
+			}
+			for (i++; i < argc; i++) {
+				word = ARGV[i]
+				set2[word] = ""
+			}
+			for (word in set2) {
+				delete set1[word]
+			}
+			for (word in set1) {
+				exit 1
+			}
+		}
+	EOF
+}
+
 #
 # Considers one or more pathnames and prints the one having the newest
 # modification time. If at least one parameter is provided, all parameters shall
diff --git a/test-functions b/test-functions
index 1a2eb81..0b987ab 100755
--- a/test-functions
+++ b/test-functions
@@ -575,6 +575,53 @@ test_parallel_run() {
 	iterate_tests 4 "$@"
 }
 
+test_is_anyof() {
+	set -- \
+		ge  1  N/A  N/A  N/A  \
+		ge  1  x    N/A  N/A  \
+		ge  1  x    y    N/A  \
+		ge  1  x    y    z    \
+		eq  0  x    x    N/A  \
+		eq  0  x    x    y    \
+		eq  0  x    y    x
+
+	callback() {
+		shift
+		test_description="is_anyof $(quote_args "$@")"
+		is_anyof "$@"
+	}
+
+	iterate_tests 5 "$@"
+}
+
+test_is_subset() {
+	set -- \
+		ge  1  N/A  N/A  N/A  N/A  N/A  \
+		ge  1  --   N/A  N/A  N/A  N/A  \
+		ge  1  --   --   N/A  N/A  N/A  \
+		ge  1  --   x    N/A  N/A  N/A  \
+		ge  1  x    --   N/A  N/A  N/A  \
+		ge  1  x    y    N/A  N/A  N/A  \
+		ge  1  x    y    x    N/A  N/A  \
+		eq  0  x    --   x    N/A  N/A  \
+		eq  0  x    --   x    y    N/A  \
+		eq  0  x    --   y    x    N/A  \
+		eq  0  x    y    --   x    y    \
+		eq  0  x    y    --   y    x    \
+		ge  1  x    y    --   x    z    \
+		ge  1  y    x    --   z    x    \
+		ge  1  x    z    --   x    y    \
+		ge  1  z    x    --   y    x
+
+	callback() {
+		shift
+		test_description="is_subset $(quote_args "$@")"
+		is_subset "$@"
+	}
+
+	iterate_tests 7 "$@"
+}
+
 iterate_tests() {
 	slice_width=$1
 	shift
@@ -641,6 +688,8 @@ test_trim || rc=1
 test_hr || rc=1
 test_whenceforth || rc=1
 test_parallel_run || rc=1
+test_is_anyof || rc=1
+test_is_subset || rc=1
 
 cleanup_tmpdir
 
-- 
cgit v1.2.3-65-gdbad