Lexer: basic chars & lifetimes

5 files changed, 56 insertions, 2 deletions

diff --git a/grammar.ron b/grammar.ron
index b707248f3..995d71f81 100644
--- a/grammar.ron
+++ b/grammar.ron
@@ -32,5 +32,7 @@ Grammar(
         "FAT_ARROW",
         "NEQ",
         "NOT",
+        "CHAR",
+        "LIFETIME",
     ]
 )
\ No newline at end of file
diff --git a/src/lexer/mod.rs b/src/lexer/mod.rs
index 24c14add0..3f277bd2b 100644
--- a/src/lexer/mod.rs
+++ b/src/lexer/mod.rs
@@ -34,7 +34,9 @@ fn next_token_inner(c: char, ptr: &mut Ptr) -> SyntaxKind {
     }
 
     if is_dec_digit(c) {
-        return scan_number(c, ptr);
+        let kind = scan_number(c, ptr);
+        scan_literal_suffix(ptr);
+        return kind;
     }
 
     // One-byte tokens.
@@ -98,6 +100,8 @@ fn next_token_inner(c: char, ptr: &mut Ptr) -> SyntaxKind {
             }
             _ => NOT,
         },
+
+        // '\'' => scan_char_or_lifetime(ptr),
         _ => (),
     }
     ERROR
@@ -116,6 +120,45 @@ fn scan_ident(c: char, ptr: &mut Ptr) -> SyntaxKind {
     IDENT
 }
 
+fn scan_char_or_lifetime(ptr: &mut Ptr) -> SyntaxKind {
+    // Either a character constant 'a' OR a lifetime name 'abc
+    let c = match ptr.bump() {
+        Some(c) => c,
+        None => return CHAR, // TODO: error reporting is upper in the stack
+    };
+
+    // If the character is an ident start not followed by another single
+    // quote, then this is a lifetime name:
+    if is_ident_start(c) && !ptr.next_is('\'') {
+        while ptr.next_is_p(is_ident_continue) {
+            ptr.bump();
+        }
+
+        // lifetimes shouldn't end with a single quote
+        // if we find one, then this is an invalid character literal
+        if ptr.next_is('\'') {
+            ptr.bump();
+            return CHAR;
+        }
+        return LIFETIME;
+    }
+    scan_char_or_byte(ptr);
+    if !ptr.next_is('\'') {
+        return CHAR; // TODO: error reporting
+    }
+    ptr.bump();
+    scan_literal_suffix(ptr);
+    CHAR
+}
+
+fn scan_literal_suffix(ptr: &mut Ptr) {
+
+}
+
+fn scan_char_or_byte(ptr: &mut Ptr) {
+    ptr.bump();
+}
+
 fn string_literal_start(c: char, c1: Option<char>, c2: Option<char>) -> bool {
     match (c, c1, c2) {
         ('r', Some('"'), _) |
diff --git a/src/lexer/ptr.rs b/src/lexer/ptr.rs
index b380117e6..2f759119a 100644
--- a/src/lexer/ptr.rs
+++ b/src/lexer/ptr.rs
@@ -34,6 +34,10 @@ impl<'s> Ptr<'s> {
         self.nnext() == Some(c)
     }
 
+    pub fn next_is_p<P: Fn(char) -> bool>(&self, p: P) -> bool {
+        self.next().map(p) == Some(true)
+    }
+
     pub fn nnext_is_p<P: Fn(char) -> bool>(&self, p: P) -> bool {
         self.nnext().map(p) == Some(true)
     }
diff --git a/src/syntax_kinds.rs b/src/syntax_kinds.rs
index 040ac1257..4c023757b 100644
--- a/src/syntax_kinds.rs
+++ b/src/syntax_kinds.rs
@@ -33,8 +33,10 @@ 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 CHAR: SyntaxKind = SyntaxKind(32);
+pub const LIFETIME: SyntaxKind = SyntaxKind(33);
 
-static INFOS: [SyntaxInfo; 32] = [
+static INFOS: [SyntaxInfo; 34] = [
     SyntaxInfo { name: "ERROR" },
     SyntaxInfo { name: "IDENT" },
     SyntaxInfo { name: "UNDERSCORE" },
@@ -67,6 +69,8 @@ static INFOS: [SyntaxInfo; 32] = [
     SyntaxInfo { name: "FAT_ARROW" },
     SyntaxInfo { name: "NEQ" },
     SyntaxInfo { name: "NOT" },
+    SyntaxInfo { name: "CHAR" },
+    SyntaxInfo { name: "LIFETIME" },
 ];
 
 pub(crate) fn syntax_info(kind: SyntaxKind) -> &'static SyntaxInfo {
diff --git a/validation.md b/validation.md
index b21ffebd5..a38b4a96e 100644
--- a/validation.md
+++ b/validation.md
@@ -4,4 +4,5 @@ Fixmes:
 * Add more thorough tests for idents for XID_Start & XID_Continue
 * Validate that float and integer literals use digits only of the appropriate
   base, and are in range
+* Validation for unclosed char literal