Implement hosts-only mode

ghostery · Aug 5, 2021 · 8dfb8ca · 8dfb8ca
1 parent dccc7ba
commit 8dfb8ca
Show file tree

Hide file tree

Showing 2 changed files with 22 additions and 1 deletion.
diff --git a/packages/adblocker-benchmarks/Makefile b/packages/adblocker-benchmarks/Makefile
@@ -5,6 +5,10 @@ ifeq ($(SHOW_MEMORY), 1)
 	run_options := $(run_options) --memory
 endif
 
+ifeq ($(HOSTS_ONLY), 1)
+	run_options := $(run_options) --hosts-only
+endif
+
 .PHONY: all clean deps run ubo-core adblockpluscore tsurlfilter-node adblock-rs brave cliqz cliqz-compression ublock adblockplus tsurlfilter tldts url min
 
 all: deps run

diff --git a/packages/adblocker-benchmarks/run.js b/packages/adblocker-benchmarks/run.js
@@ -80,10 +80,27 @@ function isSupportedUrl(url) {
   );
 }
 
+function looksLikeHostFilter(raw) {
+  // https://en.wikipedia.org/wiki/Hostname#Syntax
+  // The colon is included for the port number. e.g. ||localhost:8080^ could
+  // also be a host filter. The brackets are included for IPv6 addresses.
+  // e.g. ||[::1]^ could be a host filter. This also covers filters like
+  // ||localhost/ and ||localhost:
+  return /^(@@)?\|\|[a-z0-9.:[\]-]+[/:^]$/i.test(raw);
+}
+
 function loadLists() {
   return fs.readFileSync(path.resolve(__dirname, './easylist.txt'), { encoding: 'utf-8' });
 }
 
+function filterLists(rawLists) {
+  if (FLAGS.includes('--hosts-only')) {
+    rawLists = rawLists.split(/\n/g).filter(looksLikeHostFilter).join('\n');
+  }
+
+  return rawLists;
+}
+
 function wait(milliseconds) {
   return new Promise(resolve => {
     setTimeout(resolve, milliseconds);
@@ -109,7 +126,7 @@ async function memoryUsage(base = { heapUsed: 0, heapTotal: 0, }) {
 }
 
 async function main() {
-  const rawLists = loadLists();
+  const rawLists = filterLists(loadLists());
 
   let moduleId;
   switch (ENGINE) {