Add support for allowlists

Adds comment support for domain lists.

allowlists/.gitignore

@@ -0,0 +1,6 @@
+# Ignore everything in this directory
+*
+# Except this file
+!.gitignore
+# ... And some others
+!/allowlist.domains.example

allowlists/allowlist.domains.example

@@ -0,0 +1,8 @@
+# Omit these domains from the blocklist.
+#
+# Must be in a file called *.domains in this directory.
+# Comments are allowed. One domain per line.
+#
+# Examples:
+# domain1.example.com
+# domain2.example.com

make-block.pl

@@ -23,15 +23,10 @@ use FindBin qw//;
 my %domains;
 my $dupes = 0;
 my $skip = 0;
+my $removed_allowed = 0;
 sub add_domain_list {
     my $file = shift;
-
-    open my $fni, '<', $file
-        or die "Failed to open file $file for reading: $!";
-
-    foreach my $line (<$fni>) {
-        chomp $line;
-        $line =~ s/^\s+|\s+$//;
+    foreach my $line (&read_stripped($file)) {
         my $domain = lc $line;
         ++$dupes if defined $domains{$domain};
         $domains{$domain} = 1;
@@ -40,15 +35,7 @@ sub add_domain_list {
 
 sub add_host_file {
     my $file = shift;
-
-    open my $fni, '<', $file
-        or die "Failed to open file $file for reading: $!";
-
-    foreach my $line (<$fni>) {
-        chomp $line;
-        # strip whitespace and comments
-        $line =~ s/^\s+|\s+$|\s*#.*$//;
-        next unless $line;
+    foreach my $line (&read_stripped($file)) {
         my @parts = split /\s+/, $line;
         die "Malformed line in $file: $line; @parts"
             unless @parts > 1;
@@ -66,6 +53,20 @@ sub add_host_file {
     }
 }
 
+sub read_stripped {
+    my $file = shift;
+
+    open my $fni, '<', $file
+        or die "Failed to open file $file for reading: $!";
+
+    map {
+        chomp;
+        # Strip whitespace and comments
+        s/^\s+|\s+$|\s*#.*$//;
+        $_ || ()
+    } <$fni>;
+}
+
 MAIN: {
     my $out;
     my $block_ip = '0.0.0.0 ::1';
@@ -82,6 +83,7 @@ MAIN: {
 
     my @domain_lists = glob "$workdir/lists/*.domains";
     my @hosts_lists = glob "$workdir/lists/*.hosts";
+    my @allow_lists = glob "$workdir/allowlists/*.domains";
 
     foreach my $listfile (@domain_lists) {
         &add_domain_list($listfile);
@@ -90,6 +92,16 @@ MAIN: {
         &add_host_file($hostfile);
     }
 
+    # Apply allowlists
+    my @allow_domains;
+    foreach my $allowlist (@allow_lists) {
+        push @allow_domains, &read_stripped($allowlist);
+    }
+    my $before = %domains;
+    delete %domains{@allow_domains};
+    # Count number removed
+    $removed_allowed = $before - %domains;
+
     my $written = 0;
     my $fho = \*STDOUT;
     if (defined $out && length $out) {
@@ -109,5 +121,6 @@ MAIN: {
     printf STDERR " - %d .domains files\n", (scalar @domain_lists);
     printf STDERR " - %d .hosts files\n", (scalar @hosts_lists);
     printf STDERR "(%d duplicates)\n", $dupes if $dupes;
+    printf STDERR "(%d domains removed via allowlist)\n", $removed_allowed if $removed_allowed;
     printf STDERR "(%d skipped)\n", $skip if $skip;
 }

update.sh

@@ -36,6 +36,34 @@ cleanup() {
 }
 trap 'cleanup' EXIT
 
+copy_perms() {
+    local -r FROM=$1 TO=$2
+    chmod --reference="$FROM" -- "$TO"
+    if [[ $UID -eq 0 ]]; then
+        chown --reference="$FROM" -- "$TO"
+    fi
+}
+
+replace_with() {
+    local -r ORIG=$1 NEW=$2
+    if ! diff -q -- "$ORIG" "$NEW" &>/dev/null; then
+        # There's a change
+        if [[ -f $ORIG ]]; then
+            copy_perms "$ORIG" "$NEW"
+            if [[ -n $BACKUPSUFFIX ]]; then
+                mv -- "$ORIG" "$ORIG$BACKUPSUFFIX"
+            fi
+        else
+            mkdir -p -- "${ORIG%/*}"
+        fi
+        mv -- "$NEW" "$ORIG"
+    else
+        printf 'File "%s" not modified\n' \
+            "$ORIG" \
+            >&2
+    fi
+}
+
 (cd "$WORKDIR" && git submodule update --init --remote)
 
 for (( I = 0 ; I < ${#OUT[@]} ; ++I )); do
@@ -52,44 +80,16 @@ for (( I = 0 ; I < ${#OUT[@]} ; ++I )); do
 
     TEMP_OUT=$(mktemp -p "$TEMP_DIR")
 
-    if [[ -f $MY_OUT ]]; then
-        cp -a -- "$MY_OUT" "$TEMP_OUT"
-    fi
-
     wget \
         -O "$TEMP_OUT" \
         "$MY_URL"
 
-    if [[ -f $MY_OUT ]]; then
-        chmod --reference="$MY_OUT" "$TEMP_OUT"
-        if [[ -n $BACKUPSUFFIX ]]; then
-            mv -- "$MY_OUT" "$MY_OUT$BACKUPSUFFIX"
-        fi
-    fi
-
-    mkdir -p -- "${MY_OUT%/*}"
-    mv -- "$TEMP_OUT" "$MY_OUT"
-
-    # If the old one is the same, don't keep it around
-    if [[ -n $BACKUPSUFFIX && -f $MY_OUT$BACKUPSUFFIX ]]; then
-        if diff -q "$MY_OUT" "$MY_OUT$BACKUPSUFFIX"; then
-            rm -f -- "$MY_OUT$BACKUPSUFFIX"
-        fi
-    fi
-
+    replace_with "$MY_OUT" "$TEMP_OUT"
 done
 
-if [[ -n $BACKUPSUFFIX && -f $BLOCKLIST ]]; then
-    mv -- "$BLOCKLIST" "$BLOCKLIST$BACKUPSUFFIX"
-fi
-"$WORKDIR/make-block.pl" --out="$BLOCKLIST"
-
-# If the old one is the same the same, don't keep it around
-if [[ -n $BACKUPSUFFIX && -f $BLOCKLIST$BACKUPSUFFIX ]]; then
-    if diff -q "$BLOCKLIST" "$BLOCKLIST$BACKUPSUFFIX"; then
-        rm -f -- "$BLOCKLIST$BACKUPSUFFIX"
-    fi
-fi
+TEMP_BLOCKLIST=$(mktemp -p "$TEMP_DIR")
+"$WORKDIR/make-block.pl" --out="$TEMP_BLOCKLIST"
+replace_with "$BLOCKLIST" "$TEMP_BLOCKLIST"
 
 if [[ ${#DNSMASQ_RESTART_COMMAND[@]} -gt 0 ]]; then
     "${DNSMASQ_RESTART_COMMAND[@]}"