Updated user report

Now uses an AWK one-liner, improving performance due to making fewer calls to getent. This
might be have been anissue with certain remote-user passwd backends that do not cache a list of users.

Additionally, added fallback to simply scan /etc/passwd if getent is not available on the system.

user-report/report.sh

@@ -1,14 +1,26 @@
 #!/bin/bash
 
 # Returns a list of all users on the system with UID >= 1000 and their home directory and shell
-# Sorted lexicographically by uid
+# Sorted lexicographically by username
 # Uses getent passwd so as to include remote users (e.g. LDAP) in the list
 
 # getent passwd returns a list of all users on the system, local and remote
 # it may return duplicates, since a user might be listed identically in multiple nsswitch databases (e.g. files and systemd)
-for uid in $(getent passwd | cut -d: -f3 | sort | uniq); do
-	if [ $uid -ge 1000 ]; then
-		# Use AWK to parse the output of the passwd entry
-		getent passwd $uid | awk -F: '{print "User: " $1 ", UID: " $3 ", Home: " $6 ", Shell: " $7}'
-	fi
-done
+# as such, we sort the list and remove duplicates with uniq before processing it
+
+# I'm also removing the "nobody" user, since it's a special case and not a real user
+
+# passwd entries have the following format:
+#  $username:x:$uid:$gid:$displayname:$home:$shell
+#  e.g. root:x:0:0::/root:/bin/bash
+# Awk is 1-indexed, so $1 is the username, $2 is the x, $3 is the uid, etc.
+# the fields we care about are:
+#  $1: username
+#  $3: uid
+#  $7: shell
+if which getent &>/dev/null; then
+	getent passwd | sort | uniq | awk -F: '{if ($3 >= 1000 && $1 != "nobody") print "User: " $1 ", UID: " $3 ", Shell: " $7}'
+else
+	echo "getent not found, falling back to /etc/passwd" >&2
+	awk -F: '{if ($3 >= 1000 && $1 != "nobody") print "User: " $1 ", UID: " $3 ", Shell: " $7}' /etc/passwd
+fi