Merge #42

42: Extract parser input into a separate struct r=matklad a=matklad

bors r+

src/parser/input.rs

@@ -0,0 +1,77 @@
+use {SyntaxKind, TextRange, TextUnit, Token};
+use syntax_kinds::EOF;
+use super::is_insignificant;
+
+use std::ops::{Add, AddAssign};
+
+pub(crate) struct ParserInput<'t> {
+    #[allow(unused)]
+    text: &'t str,
+    #[allow(unused)]
+    start_offsets: Vec<TextUnit>,
+    tokens: Vec<Token>, // non-whitespace tokens
+}
+
+impl<'t> ParserInput<'t> {
+    pub fn new(text: &'t str, raw_tokens: &'t [Token]) -> ParserInput<'t> {
+        let mut tokens = Vec::new();
+        let mut start_offsets = Vec::new();
+        let mut len = TextUnit::new(0);
+        for &token in raw_tokens.iter() {
+            if !is_insignificant(token.kind) {
+                tokens.push(token);
+                start_offsets.push(len);
+            }
+            len += token.len;
+        }
+
+        ParserInput {
+            text,
+            start_offsets,
+            tokens,
+        }
+    }
+
+    pub fn kind(&self, pos: InputPosition) -> SyntaxKind {
+        let idx = pos.0 as usize;
+        if !(idx < self.tokens.len()) {
+            return EOF;
+        }
+        self.tokens[idx].kind
+    }
+
+    #[allow(unused)]
+    pub fn text(&self, pos: InputPosition) -> &'t str {
+        let idx = pos.0 as usize;
+        if !(idx < self.tokens.len()) {
+            return "";
+        }
+        let start_offset = self.start_offsets[idx];
+        let end_offset = self.tokens[idx].len;
+        let range = TextRange::from_to(start_offset, end_offset);
+        &self.text[range]
+    }
+}
+
+#[derive(Copy, Clone, Ord, PartialOrd, Eq, PartialEq)]
+pub(crate) struct InputPosition(u32);
+
+impl InputPosition {
+    pub fn new() -> Self {
+        InputPosition(0)
+    }
+}
+
+impl Add<u32> for InputPosition {
+    type Output = InputPosition;
+
+    fn add(self, rhs: u32) -> InputPosition {
+        InputPosition(self.0 + rhs)
+    }
+}
+
+impl AddAssign<u32> for InputPosition {
+    fn add_assign(&mut self, rhs: u32) {
+        self.0 += rhs
+    }
+}

src/parser/mod.rs

@@ -4,6 +4,7 @@ use syntax_kinds::*;
 
 #[macro_use]
 mod parser;
+mod input;
 mod event;
 mod grammar;
 use self::event::Event;
@@ -11,7 +12,8 @@ use self::event::Event;
 /// Parse a sequence of tokens into the representative node tree
 pub fn parse(text: String, tokens: &[Token]) -> File {
     let events = {
-        let mut parser = parser::Parser::new(&text, tokens);
+        let input = input::ParserInput::new(&text, tokens);
+        let mut parser = parser::Parser::new(&input);
         grammar::file(&mut parser);
         parser.into_events()
     };

src/parser/parser.rs

@@ -1,7 +1,6 @@
-use {SyntaxKind, TextUnit, Token};
 use super::Event;
-use super::is_insignificant;
-use SyntaxKind::{EOF, TOMBSTONE};
+use super::input::{InputPosition, ParserInput};
+use SyntaxKind::{self, EOF, TOMBSTONE};
 
 pub(crate) struct Marker {
     pos: u32,
@@ -98,35 +97,18 @@ macro_rules! token_set {
 }
 
 pub(crate) struct Parser<'t> {
-    #[allow(unused)]
-    text: &'t str,
-    #[allow(unused)]
-    start_offsets: Vec<TextUnit>,
-    tokens: Vec<Token>, // non-whitespace tokens
+    inp: &'t ParserInput<'t>,
 
-    pos: usize,
+    pos: InputPosition,
     events: Vec<Event>,
 }
 
 impl<'t> Parser<'t> {
-    pub(crate) fn new(text: &'t str, raw_tokens: &'t [Token]) -> Parser<'t> {
-        let mut tokens = Vec::new();
-        let mut start_offsets = Vec::new();
-        let mut len = TextUnit::new(0);
-        for &token in raw_tokens.iter() {
-            if !is_insignificant(token.kind) {
-                tokens.push(token);
-                start_offsets.push(len);
-            }
-            len += token.len;
-        }
-
+    pub(crate) fn new(inp: &'t ParserInput<'t>) -> Parser<'t> {
         Parser {
-            text,
-            start_offsets,
-            tokens,
+            inp,
 
-            pos: 0,
+            pos: InputPosition::new(),
             events: Vec::new(),
         }
     }
@@ -163,8 +145,8 @@ impl<'t> Parser<'t> {
         });
     }
 
-    pub(crate) fn nth(&self, n: usize) -> SyntaxKind {
-        self.tokens.get(self.pos + n).map(|t| t.kind).unwrap_or(EOF)
+    pub(crate) fn nth(&self, n: u32) -> SyntaxKind {
+        self.inp.kind(self.pos + n)
     }
 
     pub(crate) fn current(&self) -> SyntaxKind {