[perf] use StackVector to save memory & CPU

src/filters/fb_network_builder.rs

@@ -7,6 +7,7 @@ use flatbuffers::WIPOffset;
 use crate::filters::fb_builder::EngineFlatBuilder;
 use crate::filters::network::{FilterTokens, NetworkFilter};
 use crate::filters::token_selector::TokenSelector;
+use crate::utils::TokensBuffer;
 
 use crate::filters::network::NetworkFilterMaskHelper;
 use crate::flatbuffers::containers::flat_multimap::FlatMultiMapBuilder;
@@ -134,6 +135,7 @@ impl<'a> FlatSerialize<'a, EngineFlatBuilder<'a>> for NetworkFilterListBuilder {
         let mut optimizable = HashMap::<ShortHash, Vec<NetworkFilter>>::new();
 
         let mut token_frequencies = TokenSelector::new(rule_list.filters.len());
+        let mut tokens_buffer = TokensBuffer::default();
 
         {
             for network_filter in rule_list.filters {
@@ -157,7 +159,7 @@ impl<'a> FlatSerialize<'a, EngineFlatBuilder<'a>> for NetworkFilterListBuilder {
                     }
                 };
 
-                let multi_tokens = network_filter.get_tokens_optimized();
+                let multi_tokens = network_filter.get_tokens_optimized(&mut tokens_buffer);
                 match multi_tokens {
                     FilterTokens::Empty => {
                         // No tokens, add to fallback bucket (token 0)
@@ -171,7 +173,7 @@ impl<'a> FlatSerialize<'a, EngineFlatBuilder<'a>> for NetworkFilterListBuilder {
                         }
                     }
                     FilterTokens::Other(tokens) => {
-                        let best_token = token_frequencies.select_least_used_token(&tokens);
+                        let best_token = token_frequencies.select_least_used_token(tokens);
                         token_frequencies.record_usage(best_token);
                         store_filter(best_token);
                     }

src/filters/network.rs

@@ -15,9 +15,7 @@ use crate::filters::abstract_network::{
 use crate::lists::ParseOptions;
 use crate::regex_manager::RegexManager;
 use crate::request;
-use crate::utils::{self, Hash};
-
-pub(crate) const TOKENS_BUFFER_SIZE: usize = 200;
+use crate::utils::{self, Hash, TokensBuffer};
 
 /// For now, only support `$removeparam` with simple alphanumeric/dash/underscore patterns.
 static VALID_PARAM: Lazy<Regex> = Lazy::new(|| Regex::new(r"^[a-zA-Z0-9_\-]+$").unwrap());
@@ -312,10 +310,10 @@ pub enum FilterPart {
 }
 
 #[derive(Debug, PartialEq)]
-pub enum FilterTokens {
+pub enum FilterTokens<'a> {
     Empty,
-    OptDomains(Vec<Hash>),
-    Other(Vec<Hash>),
+    OptDomains(&'a [Hash]),
+    Other(&'a [Hash]),
 }
 
 pub struct FilterPartIterator<'a> {
@@ -885,17 +883,21 @@ impl NetworkFilter {
 
     #[deprecated(since = "0.11.1", note = "use get_tokens_optimized instead")]
     pub fn get_tokens(&self) -> Vec<Vec<Hash>> {
-        match self.get_tokens_optimized() {
+        let mut tokens_buffer = TokensBuffer::default();
+        match self.get_tokens_optimized(&mut tokens_buffer) {
             FilterTokens::OptDomains(domains) => {
-                domains.into_iter().map(|domain| vec![domain]).collect()
+                domains.iter().map(|domain| vec![*domain]).collect()
             }
-            FilterTokens::Other(tokens) => vec![tokens],
+            FilterTokens::Other(tokens) => vec![tokens.to_vec()],
             FilterTokens::Empty => vec![],
         }
     }
 
-    pub fn get_tokens_optimized(&self) -> FilterTokens {
-        let mut tokens: Vec<Hash> = Vec::with_capacity(TOKENS_BUFFER_SIZE);
+    pub fn get_tokens_optimized<'a>(
+        &'a self,
+        tokens_buffer: &'a mut TokensBuffer,
+    ) -> FilterTokens<'a> {
+        tokens_buffer.clear();
 
         // If there is only one domain and no domain negation, we also use this
         // domain as a token.
@@ -905,7 +907,7 @@ impl NetworkFilter {
         {
             if let Some(domains) = self.opt_domains.as_ref() {
                 if let Some(domain) = domains.first() {
-                    tokens.push(*domain)
+                    tokens_buffer.push(*domain);
                 }
             }
         }
@@ -918,7 +920,7 @@ impl NetworkFilter {
                         (self.is_plain() || self.is_regex()) && !self.is_right_anchor();
                     let skip_first_token = self.is_right_anchor();
 
-                    utils::tokenize_filter_to(f, skip_first_token, skip_last_token, &mut tokens);
+                    utils::tokenize_filter_to(f, skip_first_token, skip_last_token, tokens_buffer);
                 }
             }
             FilterPart::AnyOf(_) => (), // across AnyOf set of filters no single token is guaranteed to match to a request
@@ -928,45 +930,43 @@ impl NetworkFilter {
         // Append tokens from hostname, if any
         if !self.mask.contains(NetworkFilterMask::IS_HOSTNAME_REGEX) {
             if let Some(hostname) = self.hostname.as_ref() {
-                utils::tokenize_to(hostname, &mut tokens);
+                utils::tokenize_to(hostname, tokens_buffer);
             }
         } else if let Some(hostname) = self.hostname.as_ref() {
             // Find last dot to tokenize the prefix
             let last_dot_pos = hostname.rfind('.');
             if let Some(last_dot_pos) = last_dot_pos {
-                utils::tokenize_to(&hostname[..last_dot_pos], &mut tokens);
+                utils::tokenize_to(&hostname[..last_dot_pos], tokens_buffer);
             }
         }
 
-        if tokens.is_empty() && self.mask.contains(NetworkFilterMask::IS_REMOVEPARAM) {
+        if tokens_buffer.is_empty() && self.mask.contains(NetworkFilterMask::IS_REMOVEPARAM) {
             if let Some(removeparam) = &self.modifier_option {
                 if VALID_PARAM.is_match(removeparam) {
-                    utils::tokenize_to(&removeparam.to_ascii_lowercase(), &mut tokens);
+                    utils::tokenize_to(&removeparam.to_ascii_lowercase(), tokens_buffer);
                 }
             }
         }
 
         // If we got no tokens for the filter/hostname part, then we will dispatch
         // this filter in multiple buckets based on the domains option.
-        if tokens.is_empty() && self.opt_domains.is_some() && self.opt_not_domains.is_none() {
+        if tokens_buffer.is_empty() && self.opt_domains.is_some() && self.opt_not_domains.is_none()
+        {
             if let Some(opt_domains) = self.opt_domains.as_ref() {
                 if !opt_domains.is_empty() {
-                    return FilterTokens::OptDomains(opt_domains.clone());
+                    return FilterTokens::OptDomains(opt_domains);
                 }
             }
             FilterTokens::Empty
         } else {
             // Add optional token for protocol
             if self.for_http() && !self.for_https() {
-                tokens.push(utils::fast_hash("http"));
+                tokens_buffer.push(utils::fast_hash("http"));
             } else if self.for_https() && !self.for_http() {
-                tokens.push(utils::fast_hash("https"));
+                tokens_buffer.push(utils::fast_hash("https"));
             }
 
-            // Remake a vector to drop extra capacity.
-            let mut t = Vec::with_capacity(tokens.len());
-            t.extend(tokens);
-            FilterTokens::Other(t)
+            FilterTokens::Other(tokens_buffer.as_slice())
         }
     }
 }

src/flatbuffers/unsafe_tools.rs

@@ -97,3 +97,57 @@ impl VerifiedFlatbufferMemory {
         &self.raw_data[self.start..]
     }
 }
+
+/// A stack-allocated vector that uses [T; MAX_SIZE] with Default initialization.
+/// All elements are initialized to T::default(), and we track the logical size separately.
+/// Note: a future impl can switch to using MaybeUninit with unsafe code for better efficiency.
+pub struct StackVector<T, const MAX_SIZE: usize> {
+    data: [T; MAX_SIZE],
+    size: usize,
+}
+
+impl<T: Default + Copy, const MAX_SIZE: usize> Default for StackVector<T, MAX_SIZE> {
+    fn default() -> Self {
+        Self {
+            data: [T::default(); MAX_SIZE],
+            size: 0,
+        }
+    }
+}
+
+impl<T: Default, const MAX_SIZE: usize> StackVector<T, MAX_SIZE> {
+    pub fn push(&mut self, value: T) -> bool {
+        if self.size < MAX_SIZE {
+            self.data[self.size] = value;
+            self.size += 1;
+            true
+        } else {
+            false
+        }
+    }
+
+    pub fn is_empty(&self) -> bool {
+        self.size == 0
+    }
+
+    pub fn get_free_capacity(&self) -> usize {
+        MAX_SIZE - self.size
+    }
+
+    pub fn clear(&mut self) {
+        self.size = 0;
+    }
+
+    pub fn as_slice(&self) -> &[T] {
+        &self.data[..self.size]
+    }
+
+    pub fn into_vec(mut self) -> Vec<T> {
+        let mut v = Vec::with_capacity(self.size);
+        for i in 0..self.size {
+            v.push(std::mem::take(&mut self.data[i]));
+        }
+        self.size = 0;
+        v
+    }
+}

src/request.rs

@@ -239,11 +239,11 @@ impl Request {
 }
 
 fn calculate_tokens(url_lower_cased: &str) -> Vec<utils::Hash> {
-    let mut tokens = vec![];
+    let mut tokens = utils::TokensBuffer::default();
     utils::tokenize_pooled(url_lower_cased, &mut tokens);
     // Add zero token as a fallback to wildcard rule bucket
     tokens.push(0);
-    tokens
+    tokens.into_vec()
 }
 
 #[cfg(test)]

src/utils.rs

@@ -6,6 +6,8 @@ use seahash::hash;
 #[cfg(target_pointer_width = "32")]
 use seahash::reference::hash;
 
+use crate::flatbuffers::unsafe_tools::StackVector;
+
 pub type Hash = u64;
 
 // A smaller version of Hash that is used in serialized format.
@@ -27,25 +29,23 @@ fn is_allowed_filter(ch: char) -> bool {
     ch.is_alphanumeric() || ch == '%'
 }
 
-pub(crate) const TOKENS_BUFFER_SIZE: usize = 128;
-pub(crate) const TOKENS_BUFFER_RESERVED: usize = 1;
-const TOKENS_MAX: usize = TOKENS_BUFFER_SIZE - TOKENS_BUFFER_RESERVED;
+pub type TokensBuffer = StackVector<Hash, 256>;
 
 fn fast_tokenizer_no_regex(
     pattern: &str,
     is_allowed_code: &dyn Fn(char) -> bool,
     skip_first_token: bool,
     skip_last_token: bool,
-    tokens_buffer: &mut Vec<Hash>,
+    tokens_buffer: &mut TokensBuffer,
 ) {
     // let mut tokens_buffer_index = 0;
     let mut inside: bool = false;
     let mut start = 0;
     let mut preceding_ch: Option<char> = None; // Used to check if a '*' is not just before a token
 
     for (i, c) in pattern.char_indices() {
-        if tokens_buffer.len() >= TOKENS_MAX {
-            return;
+        if tokens_buffer.get_free_capacity() <= 1 {
+            return; // reserve one free slot for the zero token
         }
         if is_allowed_code(c) {
             if !inside {
@@ -75,17 +75,17 @@ fn fast_tokenizer_no_regex(
     }
 }
 
-pub(crate) fn tokenize_pooled(pattern: &str, tokens_buffer: &mut Vec<Hash>) {
+pub(crate) fn tokenize_pooled(pattern: &str, tokens_buffer: &mut TokensBuffer) {
     fast_tokenizer_no_regex(pattern, &is_allowed_filter, false, false, tokens_buffer);
 }
 
 pub fn tokenize(pattern: &str) -> Vec<Hash> {
-    let mut tokens_buffer: Vec<Hash> = Vec::with_capacity(TOKENS_BUFFER_SIZE);
+    let mut tokens_buffer = TokensBuffer::default();
     tokenize_to(pattern, &mut tokens_buffer);
-    tokens_buffer
+    tokens_buffer.into_vec()
 }
 
-pub(crate) fn tokenize_to(pattern: &str, tokens_buffer: &mut Vec<Hash>) {
+pub(crate) fn tokenize_to(pattern: &str, tokens_buffer: &mut TokensBuffer) {
     fast_tokenizer_no_regex(pattern, &is_allowed_filter, false, false, tokens_buffer);
 }
 
@@ -95,21 +95,21 @@ pub(crate) fn tokenize_filter(
     skip_first_token: bool,
     skip_last_token: bool,
 ) -> Vec<Hash> {
-    let mut tokens_buffer: Vec<Hash> = Vec::with_capacity(TOKENS_BUFFER_SIZE);
+    let mut tokens_buffer = TokensBuffer::default();
     tokenize_filter_to(
         pattern,
         skip_first_token,
         skip_last_token,
         &mut tokens_buffer,
     );
-    tokens_buffer
+    tokens_buffer.into_vec()
 }
 
 pub(crate) fn tokenize_filter_to(
     pattern: &str,
     skip_first_token: bool,
     skip_last_token: bool,
-    tokens_buffer: &mut Vec<Hash>,
+    tokens_buffer: &mut TokensBuffer,
 ) {
     fast_tokenizer_no_regex(
         pattern,

tests/matching.rs

@@ -199,5 +199,5 @@ fn check_rule_matching_browserlike() {
     let (blocked, passes) = bench_rule_matching_browserlike(&engine, &requests);
     let msg = "The number of blocked/passed requests has changed. ".to_string()
         + "If this is expected, update the expected values in the test.";
-    assert_eq!((blocked, passes), (106860, 136085), "{msg}");
+    assert_eq!((blocked, passes), (106861, 136084), "{msg}");
 }

tests/unit/filters/network.rs

@@ -1191,12 +1191,10 @@ mod parse_tests {
     fn test_simple_pattern_tokenization() {
         let rule = "||some.primewire.c*/sw$script,1p";
         let filter = NetworkFilter::parse(rule, true, ParseOptions::default()).unwrap();
+        let mut tokens_buffer = utils::TokensBuffer::default();
         assert_eq!(
-            filter.get_tokens_optimized(),
-            FilterTokens::Other(vec![
-                utils::fast_hash("some"),
-                utils::fast_hash("primewire")
-            ])
+            filter.get_tokens_optimized(&mut tokens_buffer),
+            FilterTokens::Other(&[utils::fast_hash("some"), utils::fast_hash("primewire")])
         );
     }
 }