Added parser for $to option in NetworkFilter.

benches/bench_cosmetic_matching.rs

@@ -1,3 +1,4 @@
+use adblock::lists::FilterSet;
 use adblock::Engine;
 use criterion::*;
 
@@ -7,10 +8,12 @@ mod test_utils;
 pub fn make_engine() -> Engine {
     use adblock::resources::Resource;
 
-    let rules = test_utils::rules_from_lists(&["data/brave/brave-main-list.txt"]);
+    let rules = test_utils::rules_from_lists(["data/brave/brave-main-list.txt"]);
     let resource_json = std::fs::read_to_string("data/brave/brave-resources.json").unwrap();
     let resource_list: Vec<Resource> = serde_json::from_str(&resource_json).unwrap();
-    let mut engine = Engine::from_rules_parametrised(rules, Default::default(), true, true);
+    let mut filter_set = FilterSet::new(true);
+    filter_set.add_filter_list(rules, Default::default());
+    let mut engine = Engine::new_with_filter_set(filter_set);
     engine.use_resources(resource_list);
     engine
 }

benches/bench_matching.rs

@@ -4,7 +4,7 @@ use serde::{Deserialize, Serialize};
 
 use adblock::request::Request;
 use adblock::url_parser::parse_url;
-use adblock::{Engine, FilterSet};
+use adblock::Engine;
 
 #[path = "../tests/test_utils.rs"]
 mod test_utils;
@@ -20,30 +20,20 @@ struct TestRequest {
 
 impl From<&TestRequest> for Request {
     fn from(v: &TestRequest) -> Self {
-        Request::new(&v.url, &v.frameUrl, &v.cpt).unwrap()
+        Request::new(&v.url, &v.frameUrl, &v.cpt, "").unwrap()
     }
 }
 
 fn load_requests() -> Vec<TestRequest> {
-    let requests_str = rules_from_lists(&["data/requests.json"]);
+    let requests_str = rules_from_lists(["data/requests.json"]);
     let reqs: Vec<TestRequest> = requests_str
-        .into_iter()
-        .map(|r| serde_json::from_str(&r))
+        .lines()
+        .map(serde_json::from_str)
         .filter_map(Result::ok)
         .collect();
     reqs
 }
 
-fn get_engine(rules: impl IntoIterator<Item = impl AsRef<str>>) -> Engine {
-    let (network_filters, cosmetic_filters) =
-        adblock::lists::parse_filters(rules, false, Default::default());
-
-    Engine::from_filter_set(
-        FilterSet::new_with_rules(network_filters, cosmetic_filters, false),
-        true,
-    )
-}
-
 fn bench_rule_matching(engine: &Engine, requests: &[TestRequest]) -> (u32, u32) {
     let mut matches = 0;
     let mut passes = 0;
@@ -87,6 +77,7 @@ fn bench_rule_matching_browserlike(blocker: &Engine, requests: &[ParsedRequest])
                 source_hostname,
                 request_type,
                 *third_party,
+                "",
             ));
             if check.matched {
                 matches += 1;
@@ -112,21 +103,21 @@ fn rule_match(c: &mut Criterion) {
     group.sample_size(10);
 
     group.bench_function("el+ep", move |b| {
-        let rules = rules_from_lists(&[
+        let rules = rules_from_lists([
             "data/easylist.to/easylist/easylist.txt",
             "data/easylist.to/easylist/easyprivacy.txt",
         ]);
-        let engine = Engine::from_rules(rules, Default::default());
+        let engine = Engine::new_with_list_text(rules);
         b.iter(|| bench_rule_matching(&engine, &elep_req))
     });
     group.bench_function("easylist", move |b| {
-        let rules = rules_from_lists(&["data/easylist.to/easylist/easylist.txt"]);
-        let engine = Engine::from_rules(rules, Default::default());
+        let rules = rules_from_lists(["data/easylist.to/easylist/easylist.txt"]);
+        let engine = Engine::new_with_list_text(rules);
         b.iter(|| bench_rule_matching(&engine, &el_req))
     });
     group.bench_function("slimlist", move |b| {
-        let rules = rules_from_lists(&["data/slim-list.txt"]);
-        let engine = Engine::from_rules(rules, Default::default());
+        let rules = rules_from_lists(["data/slim-list.txt"]);
+        let engine = Engine::new_with_list_text(rules);
         b.iter(|| bench_rule_matching(&engine, &slim_req))
     });
 
@@ -136,15 +127,15 @@ fn rule_match(c: &mut Criterion) {
 fn rule_match_parsed_el(c: &mut Criterion) {
     let mut group = c.benchmark_group("rule-match-parsed");
 
-    let rules = rules_from_lists(&["data/easylist.to/easylist/easylist.txt"]);
+    let rules = rules_from_lists(["data/easylist.to/easylist/easylist.txt"]);
     let requests = load_requests();
     let requests_parsed: Vec<_> = requests
         .into_iter()
-        .map(|r| Request::new(&r.url, &r.frameUrl, &r.cpt))
+        .map(|r| Request::new(&r.url, &r.frameUrl, &r.cpt, ""))
         .filter_map(Result::ok)
         .collect();
     let requests_len = requests_parsed.len() as u64;
-    let engine = get_engine(rules);
+    let engine = Engine::new_with_list_text(rules);
 
     group.throughput(Throughput::Elements(requests_len));
     group.sample_size(10);
@@ -159,27 +150,27 @@ fn rule_match_parsed_el(c: &mut Criterion) {
 fn rule_match_parsed_elep_slimlist(c: &mut Criterion) {
     let mut group = c.benchmark_group("rule-match-parsed");
 
-    let full_rules = rules_from_lists(&[
+    let full_rules = rules_from_lists([
         "data/easylist.to/easylist/easylist.txt",
         "data/easylist.to/easylist/easyprivacy.txt",
     ]);
-    let engine = get_engine(full_rules);
+    let engine = Engine::new_with_list_text(full_rules);
 
     let requests = load_requests();
     let requests_parsed: Vec<_> = requests
         .into_iter()
-        .map(|r| Request::new(&r.url, &r.frameUrl, &r.cpt))
+        .map(|r| Request::new(&r.url, &r.frameUrl, &r.cpt, ""))
         .filter_map(Result::ok)
         .collect();
     let requests_len = requests_parsed.len() as u64;
 
-    let slim_rules = rules_from_lists(&["data/slim-list.txt"]);
-    let slim_engine = get_engine(slim_rules);
+    let slim_rules = rules_from_lists(["data/slim-list.txt"]);
+    let slim_engine = Engine::new_with_list_text(slim_rules);
 
     let requests_copy = load_requests();
     let requests_parsed_copy: Vec<_> = requests_copy
         .into_iter()
-        .map(|r| Request::new(&r.url, &r.frameUrl, &r.cpt))
+        .map(|r| Request::new(&r.url, &r.frameUrl, &r.cpt, ""))
         .filter_map(Result::ok)
         .collect();
 
@@ -245,26 +236,26 @@ fn rule_match_browserlike_comparable(c: &mut Criterion) {
     let requests = requests_parsed(&requests);
 
     group.bench_function("el+ep", |b| {
-        let rules = rules_from_lists(&[
+        let rules = rules_from_lists([
             "data/easylist.to/easylist/easylist.txt",
             "data/easylist.to/easylist/easyprivacy.txt",
         ]);
-        let engine = Engine::from_rules_parametrised(rules, Default::default(), false, true);
+        let engine = Engine::new_with_list_text(rules);
         b.iter(|| bench_rule_matching_browserlike(&engine, &requests))
     });
     group.bench_function("el", |b| {
-        let rules = rules_from_lists(&["data/easylist.to/easylist/easylist.txt"]);
-        let engine = Engine::from_rules_parametrised(rules, Default::default(), false, true);
+        let rules = rules_from_lists(["data/easylist.to/easylist/easylist.txt"]);
+        let engine = Engine::new_with_list_text(rules);
         b.iter(|| bench_rule_matching_browserlike(&engine, &requests))
     });
     group.bench_function("slimlist", |b| {
-        let rules = rules_from_lists(&["data/slim-list.txt"]);
-        let engine = Engine::from_rules_parametrised(rules, Default::default(), false, true);
+        let rules = rules_from_lists(["data/slim-list.txt"]);
+        let engine = Engine::new_with_list_text(rules);
         b.iter(|| bench_rule_matching_browserlike(&engine, &requests))
     });
     group.bench_function("brave-list", |b| {
-        let rules = rules_from_lists(&["data/brave/brave-main-list.txt"]);
-        let engine = Engine::from_rules_parametrised(rules, Default::default(), false, true);
+        let rules = rules_from_lists(["data/brave/brave-main-list.txt"]);
+        let engine = Engine::new_with_list_text(rules);
         b.iter(|| bench_rule_matching_browserlike(&engine, &requests))
     });
 
@@ -288,9 +279,8 @@ fn rule_match_first_request(c: &mut Criterion) {
         b.iter_custom(|iters| {
             let mut total_time = std::time::Duration::ZERO;
             for _ in 0..iters {
-                let rules = rules_from_lists(&["data/brave/brave-main-list.txt"]);
-                let engine =
-                    Engine::from_rules_parametrised(rules, Default::default(), false, true);
+                let rules = rules_from_lists(["data/brave/brave-main-list.txt"]);
+                let engine = Engine::new_with_list_text(rules);
 
                 // Measure only the matching time, skip setup and destruction
                 let start_time = std::time::Instant::now();

benches/bench_memory.rs

@@ -128,15 +128,15 @@ struct TestRequest {
 
 impl From<&TestRequest> for Request {
     fn from(v: &TestRequest) -> Self {
-        Request::new(&v.url, &v.frameUrl, &v.cpt).unwrap()
+        Request::new(&v.url, &v.frameUrl, &v.cpt, "").unwrap()
     }
 }
 
 fn load_requests() -> Vec<TestRequest> {
-    let requests_str = rules_from_lists(&["data/requests.json"]);
+    let requests_str = rules_from_lists(["data/requests.json"]);
     let reqs: Vec<TestRequest> = requests_str
-        .into_iter()
-        .map(|r| serde_json::from_str(&r))
+        .lines()
+        .map(serde_json::from_str)
         .filter_map(Result::ok)
         .collect();
     reqs
@@ -243,8 +243,8 @@ fn bench_cb(
 ) {
     let single_run = || {
         ALLOCATOR.reset();
-        let rules = rules_from_lists(&["data/brave/brave-main-list.txt"]);
-        let mut engine = Engine::from_rules(rules, Default::default());
+        let rules = rules_from_lists(["data/brave/brave-main-list.txt"]);
+        let mut engine = Engine::new_with_list_text(rules);
         let resource_json = std::fs::read_to_string("data/brave/brave-resources.json").unwrap();
         let resource_list: Vec<Resource> = serde_json::from_str(&resource_json).unwrap();
         std::mem::drop(resource_json);

benches/bench_redirect_performance.rs

@@ -1,8 +1,11 @@
-use adblock::{Engine, FilterSet};
+use std::borrow::Cow;
+use std::sync::OnceLock;
+
+use adblock::Engine;
 use criterion::*;
 use tokio::runtime::Runtime;
 
-use adblock::filters::network::{NetworkFilter, NetworkFilterMask, NetworkFilterMaskHelper};
+use adblock::filters::network::{NetworkFilter, NetworkFilterMask};
 use adblock::request::Request;
 use adblock::resources::Resource;
 
@@ -55,23 +58,31 @@ async fn get_all_filters() -> Vec<String> {
         .collect()
 }
 
-/// Gets all rules with redirects, and modifies them to apply to resources at `a{0-n}.com/bad.js`
-fn get_redirect_rules() -> Vec<NetworkFilter> {
-    let async_runtime = Runtime::new().expect("Could not start Tokio runtime");
+static ALL_FILTERS: OnceLock<Box<[String]>> = OnceLock::new();
 
-    let filters = async_runtime.block_on(get_all_filters());
-    let (network_filters, _) = adblock::lists::parse_filters(&filters, true, Default::default());
+fn all_filters() -> &'static [String] {
+    ALL_FILTERS.get_or_init(|| {
+        let async_runtime = Runtime::new().expect("Could not start Tokio runtime");
+        async_runtime.block_on(get_all_filters()).into_boxed_slice()
+    })
+}
+
+/// Gets all rules with redirects, and modifies them to apply to resources at `a{0-n}.com/bad.js`
+fn get_redirect_rules() -> Vec<NetworkFilter<'static>> {
+    let filters = all_filters();
+    let (network_filters, _) =
+        adblock::lists::parse_filters(filters.iter().map(|s| s.as_str()), true, Default::default());
 
     network_filters
         .into_iter()
         .filter(NetworkFilter::is_redirect)
         .filter(NetworkFilter::also_block_redirect)
-        .filter(|rule| rule.modifier_option.as_ref().unwrap() != "none")
+        .filter(|rule| rule.modifier_option.unwrap() != "none")
         .enumerate()
         .map(|(index, mut rule)| {
             rule.mask.insert(NetworkFilterMask::IS_LEFT_ANCHOR);
             rule.mask.insert(NetworkFilterMask::IS_RIGHT_ANCHOR);
-            rule.hostname = Some(format!("a{index}.com/bad.js"));
+            rule.hostname = Some(Cow::Owned(format!("a{index}.com/bad.js")));
 
             rule.filter = adblock::filters::network::FilterPart::Empty;
             rule.mask.remove(NetworkFilterMask::IS_HOSTNAME_ANCHOR);
@@ -86,8 +97,7 @@ fn get_redirect_rules() -> Vec<NetworkFilter> {
 
 /// Loads the supplied rules, and the test set of resources, into a Engine
 fn get_preloaded_engine(rules: Vec<NetworkFilter>) -> Engine {
-    let filter_set = FilterSet::new_with_rules(rules, vec![], false);
-    Engine::from_filter_set(filter_set, true /* optimize */)
+    Engine::new_with_parsed_rules(rules, vec![])
 }
 
 fn get_resources_for_filters(#[allow(unused)] filters: &[NetworkFilter]) -> Vec<Resource> {
@@ -118,17 +128,17 @@ fn get_resources_for_filters(#[allow(unused)] filters: &[NetworkFilter]) -> Vec<
             .iter()
             .filter(|f| f.is_redirect())
             .map(|f| {
-                let mut redirect = f.modifier_option.as_ref().unwrap().as_str();
+                let mut redirect = f.modifier_option.as_ref().unwrap().to_string();
                 // strip priority, if present
                 if let Some(i) = redirect.rfind(':') {
-                    redirect = &redirect[0..i];
+                    redirect.truncate(i);
                 }
 
                 Resource {
-                    name: redirect.to_owned(),
+                    name: redirect.clone(),
                     aliases: vec![],
-                    kind: ResourceType::Mime(MimeType::from_extension(redirect)),
-                    content: BASE64_STANDARD.encode(redirect),
+                    kind: ResourceType::Mime(MimeType::from_extension(&redirect)),
+                    content: BASE64_STANDARD.encode(&redirect),
                     dependencies: vec![],
                     permission: Default::default(),
                 }
@@ -175,7 +185,7 @@ pub fn build_custom_requests(rules: Vec<NetworkFilter>) -> Vec<Request> {
             let domain = &rule_hostname[..rule_hostname.find('/').unwrap()];
             let hostname = domain;
 
-            let raw_line = rule.raw_line.clone().unwrap();
+            let raw_line = rule.raw_line.as_deref().unwrap().to_string();
             let source_hostname = if rule.opt_domains.is_some() {
                 let domain_start = raw_line.rfind("domain=").unwrap() + "domain=".len();
                 let from_start = &raw_line[domain_start..];
@@ -194,7 +204,7 @@ pub fn build_custom_requests(rules: Vec<NetworkFilter>) -> Vec<Request> {
 
             let source_url = format!("https://{source_hostname}");
 
-            Request::new(&url, &source_url, raw_type).unwrap()
+            Request::new(&url, &source_url, raw_type, "").unwrap()
         })
         .collect::<Vec<_>>()
 }