From 5cc7255168bf0bf63b1e50cb8b2d79cc58a35148 Mon Sep 17 00:00:00 2001
From: Felix Martin <mail@felixm.de>
Date: Sun, 16 May 2021 13:04:55 -0400
Subject: [PATCH] Implement basic lexer

---
 .gitignore    |  1 +
 src/lexer.rs  | 16 +++++++++++-----
 src/main.rs   |  7 +++++--
 src/parser.rs | 39 +++++++++++++++++++++++++++++++++++++++
 4 files changed, 56 insertions(+), 7 deletions(-)
 create mode 100644 src/parser.rs

diff --git a/.gitignore b/.gitignore
index fb1121f..765cf3b 100644
--- a/.gitignore
+++ b/.gitignore
@@ -6,3 +6,4 @@
 # These are backup files generated by rustfmt
 **/*.rs.bk
 
+.vscode
diff --git a/src/lexer.rs b/src/lexer.rs
index 50842ff..ad2a3d0 100644
--- a/src/lexer.rs
+++ b/src/lexer.rs
@@ -1,4 +1,4 @@
-#[derive(Debug)]
+#[derive(Debug, PartialEq, Eq)]
 pub enum Token {
     Identifier(String),
     Boolean(bool),
@@ -6,13 +6,19 @@ pub enum Token {
     LeftRoundBracket,
     RightRoundBracket,
     Quote,
+    None,
 }
 
-type Tokens = Vec<Token>;
+impl Default for Token {
+    fn default() -> Token {
+        Token::None
+    }
+}
 
-pub fn read(code: &str) -> () {
-    let tokens = scan(code, 0, vec![]);
-    print!("{:?}", tokens);
+pub type Tokens = Vec<Token>;
+
+pub fn read(code: &str) -> Tokens {
+    scan(code, 0, vec![])
 }
 
 fn scan(code: &str, mut ix: usize, mut tokens: Tokens) -> Tokens {
diff --git a/src/main.rs b/src/main.rs
index fe16819..dce5c80 100644
--- a/src/main.rs
+++ b/src/main.rs
@@ -1,6 +1,9 @@
 mod lexer;
+mod parser;
 
 fn main() {
-    let scm_code = "(+ a 32)";
-    lexer::read(scm_code);
+    let scm_code = "(+ a (* 32 b) c #t #f)";
+    let tokens = lexer::read(scm_code);
+    let datum = parser::parse(tokens);
+    println!("{:?}", datum);
 }
diff --git a/src/parser.rs b/src/parser.rs
new file mode 100644
index 0000000..a73c59b
--- /dev/null
+++ b/src/parser.rs
@@ -0,0 +1,39 @@
+use crate::lexer::Token;
+use crate::lexer::Tokens;
+
+
+#[derive(Debug)]
+pub enum Datum {
+    Boolean(bool),
+    Number(i64),
+    Symbol(String),
+    List(Vec<Datum>),
+}
+
+pub fn parse(tokens: Tokens) -> Datum {
+    let (datum, _) = parse_datum(&tokens, 0);
+    datum
+}
+
+fn parse_datum(tokens: &Tokens, ix: usize) -> (Datum, usize) {
+    match &tokens[ix] {
+        Token::Identifier(s) => (Datum::Symbol(s.to_string()), ix + 1),
+        Token::Boolean(b) => (Datum::Boolean(*b), ix + 1),
+        Token::Number(n) => (Datum::Number(*n), ix + 1),
+        Token::LeftRoundBracket => parse_list(tokens, ix + 1),
+        _ => panic!("Unexpected token {:?}", tokens[ix])
+    }
+}
+
+pub fn parse_list(tokens: &Tokens, mut ix: usize) -> (Datum, usize) {
+    let mut datums = vec![];
+
+    // FIXME: will crash when RightRoundBracket is missing 
+    while tokens[ix] != Token::RightRoundBracket {
+        let (datum, new_ix) = parse_datum(tokens, ix);
+        datums.push(datum);
+        ix = new_ix;
+    }
+
+    (Datum::List(datums), ix + 1)
+}