Parser: guess what? Groundwork!

2024-12-25 20:43:21 +00:00 · 2018-01-01 18:58:46 +03:00 · 2018-01-01 18:58:46 +03:00 · cb362626f3
commit cb362626f3
parent 0af33a2587
9 changed files with 199 additions and 58 deletions
--- a/grammar.ron
+++ b/grammar.ron
@ -1,4 +1,12 @@
 Grammar(
+    keywords: [
+        "use",
+        "fn",
+        "struct",
+        "enum",
+        "trait",
+        "impl",
+    ],
    tokens: [
        "ERROR",
        "IDENT",
@ -53,6 +61,6 @@ Grammar(
        "SHEBANG",
    ],
    nodes: [
-        "FILE"
+        "FILE",
    ]
 )
--- a/src/bin/gen.rs
+++ b/src/bin/gen.rs
@ -17,6 +17,7 @@ fn main() {

 #[derive(Deserialize)]
 struct Grammar {
+    keywords: Vec<String>,
    tokens: Vec<String>,
    nodes: Vec<String>,
 }
@ -33,8 +34,10 @@ impl Grammar {
        acc.push_str("use tree::{SyntaxKind, SyntaxInfo};\n");
        acc.push_str("\n");

-        let syntax_kinds: Vec<&String> =
-            self.tokens.iter().chain(self.nodes.iter())
+        let syntax_kinds: Vec<String> =
+            self.keywords.iter().map(|kw| kw_token(kw))
+                .chain(self.tokens.iter().cloned())
+                .chain(self.nodes.iter().cloned())
                .collect();

        for (idx, kind) in syntax_kinds.iter().enumerate() {
@ -60,6 +63,14 @@ impl Grammar {

        acc.push_str("pub(crate) fn syntax_info(kind: SyntaxKind) -> &'static SyntaxInfo {\n");
        acc.push_str("    &INFOS[kind.0 as usize]\n");
+        acc.push_str("}\n\n");
+        acc.push_str("pub(crate) fn ident_to_keyword(ident: &str) -> Option<SyntaxKind> {\n");
+        acc.push_str("   match ident {\n");
+        for kw in self.keywords.iter() {
+            write!(acc, "       {:?} => Some({}),\n", kw, kw_token(kw)).unwrap();
+        }
+        acc.push_str("       _ => None,\n");
+        acc.push_str("   }\n");
        acc.push_str("}\n");
        acc
    }
@ -78,3 +89,7 @@ fn generated_file() -> PathBuf {
 fn scream(word: &str) -> String {
    word.chars().map(|c| c.to_ascii_uppercase()).collect()
 }
+
+fn kw_token(keyword: &str) -> String {
+    format!("{}_KW", scream(keyword))
+}
--- a/src/lexer/mod.rs
+++ b/src/lexer/mod.rs
@ -187,6 +187,9 @@ fn scan_ident(c: char, ptr: &mut Ptr) -> SyntaxKind {
        return if c == '_' { UNDERSCORE } else { IDENT };
    }
    ptr.bump_while(is_ident_continue);
+    if let Some(kind) = ident_to_keyword(ptr.current_token_text()) {
+        return kind;
+    }
    IDENT
 }

--- a/src/lexer/ptr.rs
+++ b/src/lexer/ptr.rs
@ -59,6 +59,11 @@ impl<'s> Ptr<'s> {
        }
    }

+    pub fn current_token_text(&self) -> &str {
+        let len: u32 = self.len.into();
+        &self.text[..len as usize]
+    }
+
    fn chars(&self) -> Chars {
        let len: u32 = self.len.into();
        self.text[len as usize ..].chars()
--- a/src/parser/event_parser/grammar.rs
+++ b/src/parser/event_parser/grammar.rs
@ -3,8 +3,68 @@ use super::parser::Parser;

 use syntax_kinds::*;

+// Items //
+
 pub fn file(p: &mut Parser) {
    p.start(FILE);
-    //TODO: parse_shebang
+    shebang(p);
+    inner_attributes(p);
+    mod_items(p);
+    p.finish();
+}
+
+type Result = ::std::result::Result<(), ()>;
+const OK: Result = Ok(());
+const ERR: Result = Err(());
+
+fn shebang(_: &mut Parser) {
+    //TODO
+}
+
+fn inner_attributes(_: &mut Parser) {
+    //TODO
+}
+
+fn mod_items(p: &mut Parser) {
+    loop {
+        skip_until_item(p);
+        if p.is_eof() {
+            return;
+        }
+        if item(p).is_err() {
+            skip_one_token(p);
+        }
+    }
+}
+
+fn item(p: &mut Parser) -> Result {
+    outer_attributes(p)?;
+    visibility(p)?;
+    ERR
+}
+
+
+
+// Paths, types, attributes, and stuff //
+
+fn outer_attributes(_: &mut Parser) -> Result {
+    OK
+}
+
+fn visibility(_: &mut Parser) -> Result {
+    OK
+}
+
+// Expressions //
+
+// Error recovery and high-order utils //
+
+fn skip_until_item(_: &mut Parser) {
+    //TODO
+}
+
+fn skip_one_token(p: &mut Parser) {
+    p.start(ERROR);
+    p.bump().unwrap();
    p.finish();
 }
--- a/src/parser/event_parser/parser.rs
+++ b/src/parser/event_parser/parser.rs
@ -34,10 +34,14 @@ impl<'t> Parser<'t> {
    }

    pub(crate) fn into_events(self) -> Vec<Event> {
-        assert!(self.pos == self.non_ws_tokens.len());
+        assert!(self.is_eof());
        self.events
    }

+    pub(crate) fn is_eof(&self) -> bool {
+        self.pos == self.non_ws_tokens.len()
+    }
+
    pub(crate) fn start(&mut self, kind: SyntaxKind) {
        self.event(Event::Start { kind });
    }
@ -46,6 +50,15 @@ impl<'t> Parser<'t> {
        self.event(Event::Finish);
    }

+    pub(crate) fn bump(&mut self) -> Option<SyntaxKind> {
+        if self.is_eof() {
+            return None;
+        }
+        let idx = self.non_ws_tokens[self.pos].0;
+        self.pos += 1;
+        Some(self.raw_tokens[idx].kind)
+    }
+
    fn event(&mut self, event: Event) {
        self.events.push(event)
    }
--- a/src/syntax_kinds.rs
+++ b/src/syntax_kinds.rs
@ -1,60 +1,72 @@
 // Generated from grammar.ron
 use tree::{SyntaxKind, SyntaxInfo};

-pub const ERROR: SyntaxKind = SyntaxKind(0);
-pub const IDENT: SyntaxKind = SyntaxKind(1);
-pub const UNDERSCORE: SyntaxKind = SyntaxKind(2);
-pub const WHITESPACE: SyntaxKind = SyntaxKind(3);
-pub const INT_NUMBER: SyntaxKind = SyntaxKind(4);
-pub const FLOAT_NUMBER: SyntaxKind = SyntaxKind(5);
-pub const SEMI: SyntaxKind = SyntaxKind(6);
-pub const COMMA: SyntaxKind = SyntaxKind(7);
-pub const DOT: SyntaxKind = SyntaxKind(8);
-pub const DOTDOT: SyntaxKind = SyntaxKind(9);
-pub const DOTDOTDOT: SyntaxKind = SyntaxKind(10);
-pub const DOTDOTEQ: SyntaxKind = SyntaxKind(11);
-pub const L_PAREN: SyntaxKind = SyntaxKind(12);
-pub const R_PAREN: SyntaxKind = SyntaxKind(13);
-pub const L_CURLY: SyntaxKind = SyntaxKind(14);
-pub const R_CURLY: SyntaxKind = SyntaxKind(15);
-pub const L_BRACK: SyntaxKind = SyntaxKind(16);
-pub const R_BRACK: SyntaxKind = SyntaxKind(17);
-pub const L_ANGLE: SyntaxKind = SyntaxKind(18);
-pub const R_ANGLE: SyntaxKind = SyntaxKind(19);
-pub const AT: SyntaxKind = SyntaxKind(20);
-pub const POUND: SyntaxKind = SyntaxKind(21);
-pub const TILDE: SyntaxKind = SyntaxKind(22);
-pub const QUESTION: SyntaxKind = SyntaxKind(23);
-pub const COLON: SyntaxKind = SyntaxKind(24);
-pub const COLONCOLON: SyntaxKind = SyntaxKind(25);
-pub const DOLLAR: SyntaxKind = SyntaxKind(26);
-pub const EQ: SyntaxKind = SyntaxKind(27);
-pub const EQEQ: SyntaxKind = SyntaxKind(28);
-pub const FAT_ARROW: SyntaxKind = SyntaxKind(29);
-pub const NEQ: SyntaxKind = SyntaxKind(30);
-pub const NOT: SyntaxKind = SyntaxKind(31);
-pub const LIFETIME: SyntaxKind = SyntaxKind(32);
-pub const CHAR: SyntaxKind = SyntaxKind(33);
-pub const BYTE: SyntaxKind = SyntaxKind(34);
-pub const STRING: SyntaxKind = SyntaxKind(35);
-pub const RAW_STRING: SyntaxKind = SyntaxKind(36);
-pub const BYTE_STRING: SyntaxKind = SyntaxKind(37);
-pub const RAW_BYTE_STRING: SyntaxKind = SyntaxKind(38);
-pub const PLUS: SyntaxKind = SyntaxKind(39);
-pub const MINUS: SyntaxKind = SyntaxKind(40);
-pub const STAR: SyntaxKind = SyntaxKind(41);
-pub const SLASH: SyntaxKind = SyntaxKind(42);
-pub const CARET: SyntaxKind = SyntaxKind(43);
-pub const PERCENT: SyntaxKind = SyntaxKind(44);
-pub const AMPERSAND: SyntaxKind = SyntaxKind(45);
-pub const PIPE: SyntaxKind = SyntaxKind(46);
-pub const THIN_ARROW: SyntaxKind = SyntaxKind(47);
-pub const COMMENT: SyntaxKind = SyntaxKind(48);
-pub const DOC_COMMENT: SyntaxKind = SyntaxKind(49);
-pub const SHEBANG: SyntaxKind = SyntaxKind(50);
-pub const FILE: SyntaxKind = SyntaxKind(51);
+pub const USE_KW: SyntaxKind = SyntaxKind(0);
+pub const FN_KW: SyntaxKind = SyntaxKind(1);
+pub const STRUCT_KW: SyntaxKind = SyntaxKind(2);
+pub const ENUM_KW: SyntaxKind = SyntaxKind(3);
+pub const TRAIT_KW: SyntaxKind = SyntaxKind(4);
+pub const IMPL_KW: SyntaxKind = SyntaxKind(5);
+pub const ERROR: SyntaxKind = SyntaxKind(6);
+pub const IDENT: SyntaxKind = SyntaxKind(7);
+pub const UNDERSCORE: SyntaxKind = SyntaxKind(8);
+pub const WHITESPACE: SyntaxKind = SyntaxKind(9);
+pub const INT_NUMBER: SyntaxKind = SyntaxKind(10);
+pub const FLOAT_NUMBER: SyntaxKind = SyntaxKind(11);
+pub const SEMI: SyntaxKind = SyntaxKind(12);
+pub const COMMA: SyntaxKind = SyntaxKind(13);
+pub const DOT: SyntaxKind = SyntaxKind(14);
+pub const DOTDOT: SyntaxKind = SyntaxKind(15);
+pub const DOTDOTDOT: SyntaxKind = SyntaxKind(16);
+pub const DOTDOTEQ: SyntaxKind = SyntaxKind(17);
+pub const L_PAREN: SyntaxKind = SyntaxKind(18);
+pub const R_PAREN: SyntaxKind = SyntaxKind(19);
+pub const L_CURLY: SyntaxKind = SyntaxKind(20);
+pub const R_CURLY: SyntaxKind = SyntaxKind(21);
+pub const L_BRACK: SyntaxKind = SyntaxKind(22);
+pub const R_BRACK: SyntaxKind = SyntaxKind(23);
+pub const L_ANGLE: SyntaxKind = SyntaxKind(24);
+pub const R_ANGLE: SyntaxKind = SyntaxKind(25);
+pub const AT: SyntaxKind = SyntaxKind(26);
+pub const POUND: SyntaxKind = SyntaxKind(27);
+pub const TILDE: SyntaxKind = SyntaxKind(28);
+pub const QUESTION: SyntaxKind = SyntaxKind(29);
+pub const COLON: SyntaxKind = SyntaxKind(30);
+pub const COLONCOLON: SyntaxKind = SyntaxKind(31);
+pub const DOLLAR: SyntaxKind = SyntaxKind(32);
+pub const EQ: SyntaxKind = SyntaxKind(33);
+pub const EQEQ: SyntaxKind = SyntaxKind(34);
+pub const FAT_ARROW: SyntaxKind = SyntaxKind(35);
+pub const NEQ: SyntaxKind = SyntaxKind(36);
+pub const NOT: SyntaxKind = SyntaxKind(37);
+pub const LIFETIME: SyntaxKind = SyntaxKind(38);
+pub const CHAR: SyntaxKind = SyntaxKind(39);
+pub const BYTE: SyntaxKind = SyntaxKind(40);
+pub const STRING: SyntaxKind = SyntaxKind(41);
+pub const RAW_STRING: SyntaxKind = SyntaxKind(42);
+pub const BYTE_STRING: SyntaxKind = SyntaxKind(43);
+pub const RAW_BYTE_STRING: SyntaxKind = SyntaxKind(44);
+pub const PLUS: SyntaxKind = SyntaxKind(45);
+pub const MINUS: SyntaxKind = SyntaxKind(46);
+pub const STAR: SyntaxKind = SyntaxKind(47);
+pub const SLASH: SyntaxKind = SyntaxKind(48);
+pub const CARET: SyntaxKind = SyntaxKind(49);
+pub const PERCENT: SyntaxKind = SyntaxKind(50);
+pub const AMPERSAND: SyntaxKind = SyntaxKind(51);
+pub const PIPE: SyntaxKind = SyntaxKind(52);
+pub const THIN_ARROW: SyntaxKind = SyntaxKind(53);
+pub const COMMENT: SyntaxKind = SyntaxKind(54);
+pub const DOC_COMMENT: SyntaxKind = SyntaxKind(55);
+pub const SHEBANG: SyntaxKind = SyntaxKind(56);
+pub const FILE: SyntaxKind = SyntaxKind(57);

-static INFOS: [SyntaxInfo; 52] = [
+static INFOS: [SyntaxInfo; 58] = [
+    SyntaxInfo { name: "USE_KW" },
+    SyntaxInfo { name: "FN_KW" },
+    SyntaxInfo { name: "STRUCT_KW" },
+    SyntaxInfo { name: "ENUM_KW" },
+    SyntaxInfo { name: "TRAIT_KW" },
+    SyntaxInfo { name: "IMPL_KW" },
    SyntaxInfo { name: "ERROR" },
    SyntaxInfo { name: "IDENT" },
    SyntaxInfo { name: "UNDERSCORE" },
@ -112,3 +124,15 @@ static INFOS: [SyntaxInfo; 52] = [
 pub(crate) fn syntax_info(kind: SyntaxKind) -> &'static SyntaxInfo {
    &INFOS[kind.0 as usize]
 }
+
+pub(crate) fn ident_to_keyword(ident: &str) -> Option<SyntaxKind> {
+   match ident {
+       "use" => Some(USE_KW),
+       "fn" => Some(FN_KW),
+       "struct" => Some(STRUCT_KW),
+       "enum" => Some(ENUM_KW),
+       "trait" => Some(TRAIT_KW),
+       "impl" => Some(IMPL_KW),
+       _ => None,
+   }
+}
--- a/tests/data/lexer/0011_keywords.rs
+++ b/tests/data/lexer/0011_keywords.rs
@ -0,0 +1 @@
+fn use struct trait enum impl
--- a/tests/data/lexer/0011_keywords.txt
+++ b/tests/data/lexer/0011_keywords.txt
@ -0,0 +1,12 @@
+FN_KW 2 "fn"
+WHITESPACE 1 " "
+USE_KW 3 "use"
+WHITESPACE 1 " "
+STRUCT_KW 6 "struct"
+WHITESPACE 1 " "
+TRAIT_KW 5 "trait"
+WHITESPACE 1 " "
+ENUM_KW 4 "enum"
+WHITESPACE 1 " "
+IMPL_KW 4 "impl"
+WHITESPACE 1 "\n"