From 40c747619e257e83e12bbdf4d74c9c5bf9bad299 Mon Sep 17 00:00:00 2001
From: Caleb Jones <code@calebjones.net>
Date: Wed, 14 Dec 2016 01:38:53 -0500
Subject: [PATCH] Add an extremely basic s-expression parser using nom

---
 .gitignore |   2 ++
 Cargo.toml |   7 ++++
 src/lib.rs | 100 +++++++++++++++++++++++++++++++++++++++++++++++++++++
 3 files changed, 109 insertions(+)
 create mode 100644 .gitignore
 create mode 100644 Cargo.toml
 create mode 100644 src/lib.rs

diff --git a/.gitignore b/.gitignore
new file mode 100644
index 0000000..a9d37c5
--- /dev/null
+++ b/.gitignore
@@ -0,0 +1,2 @@
+target
+Cargo.lock
diff --git a/Cargo.toml b/Cargo.toml
new file mode 100644
index 0000000..6cfd35a
--- /dev/null
+++ b/Cargo.toml
@@ -0,0 +1,7 @@
+[package]
+name = "sexp2"
+version = "0.1.0"
+authors = ["Caleb Jones <code@calebjones.net>"]
+
+[dependencies]
+nom = "^2.0"
diff --git a/src/lib.rs b/src/lib.rs
new file mode 100644
index 0000000..b8982f4
--- /dev/null
+++ b/src/lib.rs
@@ -0,0 +1,100 @@
+//! A lightweight S-expression parser intended for language implementation.
+
+#![warn(missing_docs)]
+#![deny(unsafe_code)]
+
+#[macro_use]
+extern crate nom;
+
+use nom::{digit, multispace, IResult};
+
+#[derive(Debug, PartialEq, Clone, PartialOrd)]
+pub enum Atom {
+    /// A value representing a symbol. A symbol is an atomic unit
+    Sym(String),
+    /// A value representing a string literal.
+    Str(String),
+    /// A value representing an integer. Any number containing no decimal point
+    /// will be parsed as an `Int`.
+    Int(i64),
+    /// A value representing a float. Any number containing a decimal point will
+    /// be parsed as a `Float`.
+    Float(f64),
+}
+
+#[derive(Debug, PartialEq, Clone, PartialOrd)]
+pub enum Sexp {
+    /// A wrapper around the atom type
+    Atom {
+        atom: Atom,
+    },
+    /// A list of subexpressions
+    List {
+        list: Vec<Sexp>,
+    }
+}
+
+pub fn parse(input: &str) -> Result<Sexp, ()> {
+    match sexp(input) {
+        IResult::Done(_, res) => Ok(res),
+        _ => Err(()),
+    }
+}
+
+named!(sexp<&str, Sexp>,
+  alt!(
+      list => { |list| Sexp::List { list: list } }
+    | atom => { |atom| Sexp::Atom { atom: atom } }
+  )
+);
+
+named!(list<&str, Vec<Sexp> >,
+  do_parse!(
+    opt!(multispace) >>
+    tag_s!("(") >>
+    entries: many0!(sexp) >>
+    opt!(multispace) >>
+    tag_s!(")") >>
+    (entries)
+  )
+);
+
+named!(atom<&str, Atom>, alt!(string | symbol | number));
+
+named!(string<&str, Atom>,
+  do_parse!(
+    opt!(multispace) >>
+    tag_s!("\"") >>
+    contents: take_until_s!("\"") >>
+    tag_s!("\"") >>
+    opt!(multispace) >>
+    (Atom::Str(contents.into()))
+  )
+);
+
+named!(symbol<&str, Atom>,
+  do_parse!(
+    peek!(not!(digit)) >>
+    name: take_while1_s!(valid_ident_char) >>
+    (Atom::Sym(name.into()))
+  )
+);
+
+fn valid_ident_char(c: char) -> bool {
+    !c.is_whitespace() && c != '"' && c != '(' && c != ')'
+}
+
+named!(number<&str, Atom>,
+  do_parse!(
+    opt!(multispace) >>
+    integral: digit >>
+    (Atom::Int(integral.chars().fold(0, |i, c| i * 10 + c as i64 - '0' as i64)))
+  )
+);
+
+#[cfg(test)]
+mod tests {
+    #[test]
+    fn it_works() {
+    }
+}
-- 
2.47.0