rebel-parse: use phf for faster keyword parsing

5 files changed, 102 insertions, 36 deletions

diff --git a/Cargo.lock b/Cargo.lock
index a6efe32..6476593 100644
--- a/Cargo.lock
+++ b/Cargo.lock
@@ -695,6 +695,48 @@ dependencies = [
 ]
 
 [[package]]
+name = "phf"
+version = "0.11.2"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "ade2d8b8f33c7333b51bcf0428d37e217e9f32192ae4772156f65063b8ce03dc"
+dependencies = [
+ "phf_macros",
+ "phf_shared",
+]
+
+[[package]]
+name = "phf_generator"
+version = "0.11.2"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "48e4cc64c2ad9ebe670cb8fd69dd50ae301650392e81c05f9bfcb2d5bdbc24b0"
+dependencies = [
+ "phf_shared",
+ "rand",
+]
+
+[[package]]
+name = "phf_macros"
+version = "0.11.2"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "3444646e286606587e49f3bcf1679b8cef1dc2c5ecc29ddacaffc305180d464b"
+dependencies = [
+ "phf_generator",
+ "phf_shared",
+ "proc-macro2",
+ "quote",
+ "syn 2.0.60",
+]
+
+[[package]]
+name = "phf_shared"
+version = "0.11.2"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "90fcb95eef784c2ac79119d1dd819e162b5da872ce6f3c3abe1e8ca1c082f72b"
+dependencies = [
+ "siphasher",
+]
+
+[[package]]
 name = "proc-macro2"
 version = "1.0.81"
 source = "registry+https://github.com/rust-lang/crates.io-index"
@@ -713,6 +755,21 @@ dependencies = [
 ]
 
 [[package]]
+name = "rand"
+version = "0.8.5"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "34af8d1a0e25924bc5b7c43c079c942339d8f0a8b57c39049bef581b46327404"
+dependencies = [
+ "rand_core",
+]
+
+[[package]]
+name = "rand_core"
+version = "0.6.4"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "ec0be4795e2f6a28069bec0b5ff3e2ac9bafc99e6a9a7dc3547996c5c816922c"
+
+[[package]]
 name = "rebel"
 version = "0.1.0"
 dependencies = [
@@ -757,6 +814,7 @@ dependencies = [
  "clap",
  "divan",
  "peg",
+ "phf",
  "rebel-common",
 ]
 
@@ -960,6 +1018,12 @@ dependencies = [
 ]
 
 [[package]]
+name = "siphasher"
+version = "0.3.11"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "38b58827f4464d87d377d175e90bf58eb00fd8716ff0a62f80356b5e61555d0d"
+
+[[package]]
 name = "smallvec"
 version = "1.13.2"
 source = "registry+https://github.com/rust-lang/crates.io-index"
diff --git a/crates/rebel-parse/Cargo.toml b/crates/rebel-parse/Cargo.toml
index fcd802f..7159976 100644
--- a/crates/rebel-parse/Cargo.toml
+++ b/crates/rebel-parse/Cargo.toml
@@ -9,6 +9,7 @@ edition = "2021"
 
 [dependencies]
 peg = "0.8.3"
+phf = { version = "0.11.2", features = ["macros"] }
 rebel-common = { path = "../rebel-common" }
 
 [dev-dependencies]
diff --git a/crates/rebel-parse/src/grammar/recipe.rs b/crates/rebel-parse/src/grammar/recipe.rs
index 4d24dce..61688f5 100644
--- a/crates/rebel-parse/src/grammar/recipe.rs
+++ b/crates/rebel-parse/src/grammar/recipe.rs
@@ -12,10 +12,10 @@ peg::parser! {
 			= recipe:recipe_stmt()* { recipe }
 
 		pub rule recipe_stmt() -> ast::RecipeStmt<'a>
-			= keyword_fetch() name:ident() p('{') body:body() p('}') {
+			= [Token::Keyword(Keyword::Fetch)] name:ident() p('{') body:body() p('}') {
 				ast::RecipeStmt::Fetch { name, body: ast::Body(Vec::new()) }
 			}
-			/ keyword_task() name:ident() p('(') params:func_params() p(')')
+			/ [Token::Keyword(Keyword::Task)] name:ident() p('(') params:func_params() p(')')
 			  p('{') body:body() p('}') {
 				ast::RecipeStmt::Task { name, params, body }
 			}
@@ -99,8 +99,8 @@ peg::parser! {
 			= name:ident() p(':') typ:expr() { ast::FuncParam { name, typ } }
 
 		rule literal() -> ast::Literal<'a>
-			= keyword_true() { ast::Literal::Bool(true) }
-			/ keyword_false() { ast::Literal::Bool(false) }
+			= [Token::Keyword(Keyword::True)] { ast::Literal::Bool(true) }
+			/ [Token::Keyword(Keyword::False)] { ast::Literal::Bool(false) }
 			/ [Token::Number(content)] { ?
 				ast::Literal::number(content)
 			}
@@ -146,34 +146,7 @@ peg::parser! {
 			= p_(ch1) p(ch2)
 
 		rule ident() -> ast::Ident<'a>
-			= !keyword() [Token::Ident(name)] { ast::Ident { name } }
-
-		rule keyword()
-			= keyword_true()
-			/ keyword_false()
-			/ keyword_fetch()
-			/ keyword_task()
-			/ keyword_struct()
-			/ keyword_set()
-			/ keyword_map()
-
-		rule keyword_true()
-			= const_ident("true")
-		rule keyword_false()
-			= const_ident("false")
-		rule keyword_fetch()
-			= const_ident("fetch")
-		rule keyword_task()
-			= const_ident("task")
-		rule keyword_struct()
-			= const_ident("struct")
-		rule keyword_set()
-			= const_ident("set")
-		rule keyword_map()
-			= const_ident("map")
-
-		rule const_ident(keyword: &str)
-			= [Token::Ident(name) if *name == keyword]
+			= [Token::Ident(name)] { ast::Ident { name } }
 
 		rule delimited<T>(expr: rule<T>, delim: rule<()>) -> Vec<T>
 			= values:(expr() ++ delim()) delim()? { values }
diff --git a/crates/rebel-parse/src/grammar/tokenize.rs b/crates/rebel-parse/src/grammar/tokenize.rs
index 4758e9e..a696852 100644
--- a/crates/rebel-parse/src/grammar/tokenize.rs
+++ b/crates/rebel-parse/src/grammar/tokenize.rs
@@ -2,6 +2,16 @@ use crate::token::*;
 
 pub use rules::*;
 
+static KEYWORDS: phf::Map<&'static str, Keyword> = phf::phf_map! {
+	"true" => Keyword::True,
+	"false" => Keyword::False,
+	"fetch" => Keyword::Fetch,
+	"task" => Keyword::Task,
+	"struct" => Keyword::Struct,
+	"set" =>  Keyword::Set,
+	"map" => Keyword::Map,
+};
+
 peg::parser! {
 	pub grammar rules() for str {
 		pub rule token_stream() -> TokenStream<'input>
@@ -10,14 +20,20 @@ peg::parser! {
 		pub rule token() -> Token<'input>
 			= number:number() { Token::Number(number) }
 			/ string:string() { Token::Str(string) }
-			/ ident:ident() { Token::Ident(ident) }
+			/ token:ident_or_keyword() { token }
 			/ punct:punct() { Token::Punct(punct) }
 
-		rule ident() -> &'input str
-			= $(
+		rule ident_or_keyword() -> Token<'input>
+			= s:$(
 				['a'..='z' | 'A' ..='Z' | '_' ]
 				['a'..='z' | 'A' ..='Z' | '_' | '0'..='9']*
-			)
+			) {
+				if let Some(kw) = KEYWORDS.get(s) {
+					Token::Keyword(*kw)
+				} else {
+					Token::Ident(s)
+				}
+			}
 
 		rule punct() -> Punct
 			= ch:punct_char() spacing:spacing() { Punct(ch, spacing) }
diff --git a/crates/rebel-parse/src/token.rs b/crates/rebel-parse/src/token.rs
index 80d758d..bfa9c42 100644
--- a/crates/rebel-parse/src/token.rs
+++ b/crates/rebel-parse/src/token.rs
@@ -1,5 +1,6 @@
 #[derive(Clone, Debug, PartialEq, Eq)]
 pub enum Token<'a> {
+	Keyword(Keyword),
 	Ident(&'a str),
 	Punct(Punct),
 	Str(Str<'a>),
@@ -7,6 +8,17 @@ pub enum Token<'a> {
 }
 
 #[derive(Clone, Copy, Debug, PartialEq, Eq)]
+pub enum Keyword {
+	True,
+	False,
+	Fetch,
+	Task,
+	Struct,
+	Set,
+	Map,
+}
+
+#[derive(Clone, Copy, Debug, PartialEq, Eq)]
 pub struct Punct(pub char, pub Spacing);
 
 #[derive(Clone, Copy, Debug, PartialEq, Eq)]