Merge pull request #65 from neocturne/peg

Restore peg example, turn into an actual parser

README.md

@@ -9,12 +9,12 @@ This repo tries to assess Rust parsing performance.
 | [lalrpop] | LR(1)       | in grammar  | build script       | `&str`                 | No                  | Yes                 | No              |
 | [logos]   | lexer       | in source   | proc macro         | `&str`, `&[u8]`        | ?                   | ?                   | ?               |
 | [nom]     | combinators | in source   | library            | `&[u8]`, custom        | No                  | Yes                 | Yes             |
+| [peg]     | PEG         | in grammar  | proc macro (block) | `&str`, `&[T]`, custom | Yes                 | Yes                 | No              |
 | [pest]    | PEG         | external    | proc macro (file)  | `&str`                 | Yes                 | No                  | No              |
 | [winnow]  | combinators | in source   | library            | `&str`, `&[T]`, custom | No                  | Yes                 | Yes             |
 | [yap]     | combinators | in source   | library            | `&str`, `&[T]`, custom | No                  | Yes                 | ?               |
 
 Formerly, we compared:
-- [peg]: invalid example
 - [pom]: lack of notoriety
 
 # Results
@@ -27,6 +27,7 @@ logos | 170 KiB | 5s | 17ms | ![Download count](https://img.shields.io/crates/dr
 combine | 204 KiB | 4s | 49ms | ![Download count](https://img.shields.io/crates/dr/combine) | v3.8.1
 lalrpop | 1,615 KiB | 11s | 880ms | ![Download count](https://img.shields.io/crates/dr/lalrpop-util) | v0.20.0
 nom | 99 KiB | 2s | 66ms | ![Download count](https://img.shields.io/crates/dr/nom) | v7.1.3
+peg | 21 KiB | 2s | **invalid** | ![Download count](https://img.shields.io/crates/dr/peg) | v0.8.2
 pest | 104 KiB | 4s | 48ms | ![Download count](https://img.shields.io/crates/dr/pest) | v2.7.6
 serde_json | 41 KiB | 3s | 13ms | ![Download count](https://img.shields.io/crates/dr/serde_json) | v1.0.113
 winnow | 73 KiB | 2s | 22ms | ![Download count](https://img.shields.io/crates/dr/winnow) | v0.6.0

examples/peg-app/Cargo.toml

@@ -0,0 +1,11 @@
+[package]
+name = "peg-app"
+version = "0.1.0"
+edition = "2021"
+
+[[bin]]
+name = "peg-app"
+path = "app.rs"
+
+[dependencies]
+peg = "0.8.2"

examples/peg-app/app.rs

@@ -0,0 +1,25 @@
+mod parser;
+
+use std::{env, fs};
+
+fn main() {
+    let src = fs::read_to_string(env::args().nth(1).expect("Expected file argument"))
+        .expect("Failed to read file");
+
+    match parser::parser::json(&src) {
+        Ok(json) => {
+            #[cfg(debug_assertions)]
+            {
+                println!("{:#?}", json);
+            }
+            #[cfg(not(debug_assertions))]
+            {
+                std::hint::black_box(json);
+            }
+        }
+        Err(err) => {
+            eprintln!("{}", err);
+            std::process::exit(1);
+        }
+    };
+}

examples/peg-app/parser.rs

@@ -0,0 +1,94 @@
+use std::{borrow::Cow, collections::HashMap, str::FromStr};
+
+#[derive(Debug, PartialEq, Clone)]
+pub enum JsonValue {
+    Null,
+    Boolean(bool),
+    Str(String),
+    Num(f64),
+    Array(Vec<JsonValue>),
+    Object(HashMap<String, JsonValue>),
+}
+
+peg::parser!(pub grammar parser() for str {
+
+pub rule json() -> JsonValue
+    = _ value:value() _ { value }
+
+rule _() = [' ' | '\t' | '\r' | '\n']*
+rule value_separator() = _ "," _
+
+rule value() -> JsonValue
+    = boolean() / null() / object() / array() / number() / string()
+
+rule null() -> JsonValue
+    = "null" { JsonValue::Null }
+
+rule boolean() -> JsonValue
+    = "true" { JsonValue::Boolean(true) }
+    / "false" { JsonValue::Boolean(false) }
+
+rule object() -> JsonValue
+    = "{" _ elements:(member() ** value_separator()) _ "}" {
+        JsonValue::Object(elements.into_iter().collect())
+    }
+
+rule member() -> (String, JsonValue)
+    = key:raw_string() _ ":" _ value:value() { (key, value) }
+
+rule array() -> JsonValue
+    = "[" _ elements:(value() ** value_separator()) _ "]" {
+        JsonValue::Array(elements)
+    }
+
+rule string() -> JsonValue
+    = value:raw_string() { JsonValue::Str(value) }
+
+rule raw_string() -> String
+    = "\"" slices:string_slice()* "\"" { slices.concat() }
+
+/// A substring of same-kind (escaped or unescaped) characters
+rule string_slice() -> Cow<'input, str>
+    = value:string_characters() { Cow::Borrowed(value) }
+    / value:string_escapes() { Cow::Owned(value.into_iter().collect()) }
+
+/// A substring of unescaped characters
+rule string_characters() -> &'input str
+    = $([^ '\"' | '\\']+)
+
+/// A substring of escaped characters
+rule string_escapes() -> Vec<char>
+    = ("\\" value:string_escape_char() { value })+
+
+/// Handles a single escape
+rule string_escape_char() -> char
+    = "\"" { '"' }
+    / "\\" { '\\' }
+    / "/"  { '/' }
+    / "b" { '\x08' }
+    / "f" { '\x0C' }
+    / "n" { '\n' }
+    / "r" { '\r' }
+    / "t" { '\t' }
+    / "u" digits:$(hex_digit()*<4>) { ?
+        let value = u16::from_str_radix(digits, 16).unwrap();
+        char::from_u32(value.into()).ok_or("invalid unicode escape")
+    }
+
+rule hex_digit()
+    = ['0'..='9' | 'a'..='f' | 'A'..='F']
+
+rule number() -> JsonValue
+    = "-"? value:$(int() frac()? exp()?) { ?
+        Ok(JsonValue::Num(f64::from_str(value).map_err(|_| "invalid number")?))
+    }
+
+rule int()
+    = ['0'] / ['1'..='9']['0'..='9']*
+
+rule exp()
+    = ("e" / "E") ("-" / "+")? ['0'..='9']*<1,>
+
+rule frac()
+    = "." ['0'..='9']*<1,>
+});