Merge remote-tracking branch 'origin/master' into fintan/canonicalize

3 files changed, 440 insertions, 0 deletions

diff --git a/serde_dhall/src/lib.rs b/serde_dhall/src/lib.rs
new file mode 100644
index 0000000..3d07539
--- /dev/null
+++ b/serde_dhall/src/lib.rs
@@ -0,0 +1,279 @@
+#![feature(non_exhaustive)]
+
+//! [Dhall][dhall] is a programmable configuration language that provides a non-repetitive
+//! alternative to JSON and YAML.
+//!
+//! You can think of Dhall as: JSON + types + imports + functions
+//!
+//! For a description of the Dhall language, examples, tutorials, and more, see the [language
+//! website][dhall].
+//!
+//! This crate provides support for consuming Dhall files the same way you would consume JSON or
+//! YAML. It uses the [Serde][serde] serialization library to provide drop-in support for Dhall
+//! for any datatype that supports serde (and that's a lot of them !).
+//!
+//! This library is limited to deserializing (reading) Dhall values; serializing (writing)
+//! values to Dhall is not supported for now.
+//!
+//! # Examples
+//!
+//! ### Custom datatype
+//!
+//! If you have a custom datatype for which you derived [serde::Deserialize], chances are
+//! you will be able to derive [StaticType] for it as well.
+//! This allows easy type-safe deserializing.
+//!
+//! ```edition2018
+//! use serde::Deserialize;
+//! use serde_dhall::{de::Error, StaticType};
+//!
+//! #[derive(Debug, Deserialize, StaticType)]
+//! struct Point {
+//!     x: u64,
+//!     y: u64,
+//! }
+//!
+//! fn main() -> Result<(), Error> {
+//!     // Some Dhall data
+//!     let data = "{ x = 1, y = 1 + 1 }";
+//!
+//!     // Convert the Dhall string to a Point.
+//!     let point: Point = serde_dhall::from_str_auto_type(data)?;
+//!     assert_eq!(point.x, 1);
+//!     assert_eq!(point.y, 2);
+//!
+//!     // Invalid data fails the type validation
+//!     let invalid_data = "{ x = 1, z = 0.3 }";
+//!     assert!(serde_dhall::from_str_auto_type::<Point>(invalid_data).is_err());
+//!
+//!     Ok(())
+//! }
+//! ```
+//!
+//! ### Loosely typed
+//!
+//! If you used to consume JSON or YAML in a loosely typed way, you can continue to do so
+//! with Dhall. You only need to replace [serde_json::from_str] or [serde_yaml::from_str]
+//! with [serde_dhall::from_str][from_str].
+//! More generally, if the [StaticType] derive doesn't suit your
+//! needs, you can still deserialize any valid Dhall file that serde can handle.
+//!
+//! [serde_json::from_str]: https://docs.serde.rs/serde_json/de/fn.from_str.html
+//! [serde_yaml::from_str]: https://docs.serde.rs/serde_yaml/fn.from_str.html
+//!
+//! ```edition2018
+//! # fn main() -> serde_dhall::de::Result<()> {
+//! use std::collections::BTreeMap;
+//!
+//! // Some Dhall data
+//! let data = "{ x = 1, y = 1 + 1 } : { x: Natural, y: Natural }";
+//!
+//! // Deserialize it to a Rust type.
+//! let deserialized_map: BTreeMap<String, usize> = serde_dhall::from_str(data)?;
+//!
+//! let mut expected_map = BTreeMap::new();
+//! expected_map.insert("x".to_string(), 1);
+//! expected_map.insert("y".to_string(), 2);
+//!
+//! assert_eq!(deserialized_map, expected_map);
+//! # Ok(())
+//! # }
+//! ```
+//!
+//! You can alternatively specify a Dhall type that the input should match.
+//!
+//! ```edition2018
+//! # fn main() -> serde_dhall::de::Result<()> {
+//! use std::collections::BTreeMap;
+//!
+//! // Parse a Dhall type
+//! let point_type_str = "{ x: Natural, y: Natural }";
+//! let point_type = serde_dhall::from_str(point_type_str)?;
+//!
+//! // Some Dhall data
+//! let point_data = "{ x = 1, y = 1 + 1 }";
+//!
+//! // Deserialize the data to a Rust type. This ensures that
+//! // the data matches the point type.
+//! let deserialized_map: BTreeMap<String, usize> =
+//!         serde_dhall::from_str_check_type(point_data, &point_type)?;
+//!
+//! let mut expected_map = BTreeMap::new();
+//! expected_map.insert("x".to_string(), 1);
+//! expected_map.insert("y".to_string(), 2);
+//!
+//! assert_eq!(deserialized_map, expected_map);
+//! # Ok(())
+//! # }
+//! ```
+//!
+//! [dhall]: https://dhall-lang.org/
+//! [serde]: https://docs.serde.rs/serde/
+//! [serde::Deserialize]: https://docs.serde.rs/serde/trait.Deserialize.html
+
+mod serde;
+mod static_type;
+
+#[doc(inline)]
+pub use de::{from_str, from_str_auto_type, from_str_check_type};
+#[doc(hidden)]
+pub use dhall_proc_macros::StaticType;
+pub use static_type::StaticType;
+#[doc(inline)]
+pub use value::Value;
+
+// A Dhall value.
+pub mod value {
+    use dhall::phase::{NormalizedExpr, Parsed, Typed};
+    use dhall_syntax::Builtin;
+
+    use super::de::{Error, Result};
+
+    /// A Dhall value
+    #[derive(Debug, Clone, PartialEq, Eq)]
+    pub struct Value(Typed);
+
+    impl Value {
+        pub fn from_str(s: &str, ty: Option<&Value>) -> Result<Self> {
+            Value::from_str_using_dhall_error_type(s, ty).map_err(Error::Dhall)
+        }
+        fn from_str_using_dhall_error_type(
+            s: &str,
+            ty: Option<&Value>,
+        ) -> dhall::error::Result<Self> {
+            let resolved = Parsed::parse_str(s)?.resolve()?;
+            let typed = match ty {
+                None => resolved.typecheck()?,
+                Some(t) => resolved.typecheck_with(t.as_typed())?,
+            };
+            Ok(Value(typed))
+        }
+        pub(crate) fn to_expr(&self) -> NormalizedExpr {
+            self.0.normalize_to_expr()
+        }
+        pub(crate) fn as_typed(&self) -> &Typed {
+            &self.0
+        }
+
+        pub(crate) fn make_builtin_type(b: Builtin) -> Self {
+            Value(Typed::make_builtin_type(b))
+        }
+        pub(crate) fn make_optional_type(t: Value) -> Self {
+            Value(Typed::make_optional_type(t.0))
+        }
+        pub(crate) fn make_list_type(t: Value) -> Self {
+            Value(Typed::make_list_type(t.0))
+        }
+        // Made public for the StaticType derive macro
+        #[doc(hidden)]
+        pub fn make_record_type(
+            kts: impl Iterator<Item = (String, Value)>,
+        ) -> Self {
+            Value(Typed::make_record_type(kts.map(|(k, t)| (k, t.0))))
+        }
+        #[doc(hidden)]
+        pub fn make_union_type(
+            kts: impl Iterator<Item = (String, Option<Value>)>,
+        ) -> Self {
+            Value(Typed::make_union_type(
+                kts.map(|(k, t)| (k, t.map(|t| t.0))),
+            ))
+        }
+    }
+
+    impl super::de::Deserialize for Value {
+        fn from_dhall(v: &Value) -> Result<Self> {
+            Ok(v.clone())
+        }
+    }
+}
+
+/// Deserialize Dhall data to a Rust data structure.
+pub mod de {
+    use super::StaticType;
+    use super::Value;
+    pub use error::{Error, Result};
+
+    mod error {
+        use dhall::error::Error as DhallError;
+
+        pub type Result<T> = std::result::Result<T, Error>;
+
+        #[derive(Debug)]
+        #[non_exhaustive]
+        pub enum Error {
+            Dhall(DhallError),
+            Deserialize(String),
+        }
+
+        impl std::fmt::Display for Error {
+            fn fmt(&self, f: &mut std::fmt::Formatter) -> std::fmt::Result {
+                match self {
+                    Error::Dhall(err) => write!(f, "{}", err),
+                    Error::Deserialize(err) => write!(f, "{}", err),
+                }
+            }
+        }
+
+        impl std::error::Error for Error {}
+
+        impl serde::de::Error for Error {
+            fn custom<T>(msg: T) -> Self
+            where
+                T: std::fmt::Display,
+            {
+                Error::Deserialize(msg.to_string())
+            }
+        }
+    }
+
+    /// A data structure that can be deserialized from a Dhall expression
+    ///
+    /// This is automatically implemented for any type that [serde][serde]
+    /// can deserialize.
+    ///
+    /// This trait cannot be implemented manually.
+    // TODO: seal trait
+    pub trait Deserialize: Sized {
+        /// See [serde_dhall::from_str][crate::from_str]
+        fn from_dhall(v: &Value) -> Result<Self>;
+    }
+
+    /// Deserialize an instance of type `T` from a string of Dhall text.
+    ///
+    /// This will recursively resolve all imports in the expression, and
+    /// typecheck it before deserialization. Relative imports will be resolved relative to the
+    /// provided file. More control over this process is not yet available
+    /// but will be in a coming version of this crate.
+    pub fn from_str<T>(s: &str) -> Result<T>
+    where
+        T: Deserialize,
+    {
+        T::from_dhall(&Value::from_str(s, None)?)
+    }
+
+    /// Deserialize an instance of type `T` from a string of Dhall text,
+    /// additionally checking that it matches the supplied type.
+    ///
+    /// Like [from_str], but this additionally checks that
+    /// the type of the provided expression matches the supplied type.
+    pub fn from_str_check_type<T>(s: &str, ty: &Value) -> Result<T>
+    where
+        T: Deserialize,
+    {
+        T::from_dhall(&Value::from_str(s, Some(ty))?)
+    }
+
+    /// Deserialize an instance of type `T` from a string of Dhall text,
+    /// additionally checking that it matches the type of `T`.
+    ///
+    /// Like [from_str], but this additionally checks that
+    /// the type of the provided expression matches the output type `T`. The [StaticType] trait
+    /// captures Rust types that are valid Dhall types.
+    pub fn from_str_auto_type<T>(s: &str) -> Result<T>
+    where
+        T: Deserialize + StaticType,
+    {
+        from_str_check_type(s, &<T as StaticType>::static_type())
+    }
+}
diff --git a/serde_dhall/src/serde.rs b/serde_dhall/src/serde.rs
new file mode 100644
index 0000000..26708c1
--- /dev/null
+++ b/serde_dhall/src/serde.rs
@@ -0,0 +1,68 @@
+use std::borrow::Cow;
+
+use dhall::phase::NormalizedExpr;
+use dhall_syntax::ExprF;
+
+use crate::de::{Deserialize, Error, Result};
+use crate::Value;
+
+impl<'a, T> Deserialize for T
+where
+    T: serde::Deserialize<'a>,
+{
+    fn from_dhall(v: &Value) -> Result<Self> {
+        T::deserialize(Deserializer(Cow::Owned(v.to_expr())))
+    }
+}
+
+struct Deserializer<'a>(Cow<'a, NormalizedExpr>);
+
+impl<'de: 'a, 'a> serde::de::IntoDeserializer<'de, Error> for Deserializer<'a> {
+    type Deserializer = Deserializer<'a>;
+    fn into_deserializer(self) -> Self::Deserializer {
+        self
+    }
+}
+
+impl<'de: 'a, 'a> serde::Deserializer<'de> for Deserializer<'a> {
+    type Error = Error;
+    fn deserialize_any<V>(self, visitor: V) -> Result<V::Value>
+    where
+        V: serde::de::Visitor<'de>,
+    {
+        use std::convert::TryInto;
+        use ExprF::*;
+        match self.0.as_ref().as_ref() {
+            NaturalLit(n) => {
+                if let Ok(n64) = (*n).try_into() {
+                    visitor.visit_u64(n64)
+                } else if let Ok(n32) = (*n).try_into() {
+                    visitor.visit_u32(n32)
+                } else {
+                    unimplemented!()
+                }
+            }
+            IntegerLit(n) => {
+                if let Ok(n64) = (*n).try_into() {
+                    visitor.visit_i64(n64)
+                } else if let Ok(n32) = (*n).try_into() {
+                    visitor.visit_i32(n32)
+                } else {
+                    unimplemented!()
+                }
+            }
+            RecordLit(m) => visitor.visit_map(
+                serde::de::value::MapDeserializer::new(m.iter().map(
+                    |(k, v)| (k.as_ref(), Deserializer(Cow::Borrowed(v))),
+                )),
+            ),
+            _ => unimplemented!(),
+        }
+    }
+
+    serde::forward_to_deserialize_any! {
+        bool i8 i16 i32 i64 i128 u8 u16 u32 u64 u128 f32 f64 char str string
+        bytes byte_buf option unit unit_struct newtype_struct seq tuple
+        tuple_struct map struct enum identifier ignored_any
+    }
+}
diff --git a/serde_dhall/src/static_type.rs b/serde_dhall/src/static_type.rs
new file mode 100644
index 0000000..67a7bc4
--- /dev/null
+++ b/serde_dhall/src/static_type.rs
@@ -0,0 +1,93 @@
+use dhall_syntax::{Builtin, Integer, Natural};
+
+use crate::Value;
+
+/// A Rust type that can be represented as a Dhall type.
+///
+/// A typical example is `Option<bool>`,
+/// represented by the dhall expression `Optional Bool`.
+///
+/// This trait can and should be automatically derived.
+///
+/// The representation needs to be independent of the value.
+/// For this reason, something like `HashMap<String, bool>` cannot implement
+/// [StaticType] because each different value would
+/// have a different Dhall record type.
+pub trait StaticType {
+    fn static_type() -> Value;
+}
+
+macro_rules! derive_builtin {
+    ($ty:ty, $builtin:ident) => {
+        impl StaticType for $ty {
+            fn static_type() -> Value {
+                Value::make_builtin_type(Builtin::$builtin)
+            }
+        }
+    };
+}
+
+derive_builtin!(bool, Bool);
+derive_builtin!(Natural, Natural);
+derive_builtin!(u64, Natural);
+derive_builtin!(Integer, Integer);
+derive_builtin!(String, Text);
+
+impl<A, B> StaticType for (A, B)
+where
+    A: StaticType,
+    B: StaticType,
+{
+    fn static_type() -> Value {
+        Value::make_record_type(
+            vec![
+                ("_1".to_owned(), A::static_type()),
+                ("_2".to_owned(), B::static_type()),
+            ]
+            .into_iter(),
+        )
+    }
+}
+
+impl<T, E> StaticType for std::result::Result<T, E>
+where
+    T: StaticType,
+    E: StaticType,
+{
+    fn static_type() -> Value {
+        Value::make_union_type(
+            vec![
+                ("Ok".to_owned(), Some(T::static_type())),
+                ("Err".to_owned(), Some(E::static_type())),
+            ]
+            .into_iter(),
+        )
+    }
+}
+
+impl<T> StaticType for Option<T>
+where
+    T: StaticType,
+{
+    fn static_type() -> Value {
+        Value::make_optional_type(T::static_type())
+    }
+}
+
+impl<T> StaticType for Vec<T>
+where
+    T: StaticType,
+{
+    fn static_type() -> Value {
+        Value::make_list_type(T::static_type())
+    }
+}
+
+impl<'a, T> StaticType for &'a T
+where
+    T: StaticType,
+{
+    fn static_type() -> Value {
+        T::static_type()
+    }
+}