1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130
// Copyright 2014 The Rust Project Developers. See the COPYRIGHT // file at the top-level directory of this distribution and at // http://rust-lang.org/COPYRIGHT. // // Licensed under the Apache License, Version 2.0 <LICENSE-APACHE or // http://www.apache.org/licenses/LICENSE-2.0> or the MIT license // <LICENSE-MIT or http://opensource.org/licenses/MIT>, at your // option. This file may not be copied, modified, or distributed // except according to those terms. //! The AST pointer //! //! Provides `P<T>`, a frozen owned smart pointer, as a replacement for `@T` in //! the AST. //! //! # Motivations and benefits //! //! * **Identity**: sharing AST nodes is problematic for the various analysis //! passes (e.g. one may be able to bypass the borrow checker with a shared //! `ExprAddrOf` node taking a mutable borrow). The only reason `@T` in the //! AST hasn't caused issues is because of inefficient folding passes which //! would always deduplicate any such shared nodes. Even if the AST were to //! switch to an arena, this would still hold, i.e. it couldn't use `&'a T`, //! but rather a wrapper like `P<'a, T>`. //! //! * **Immutability**: `P<T>` disallows mutating its inner `T`, unlike `Box<T>` //! (unless it contains an `Unsafe` interior, but that may be denied later). //! This mainly prevents mistakes, but can also enforces a kind of "purity". //! //! * **Efficiency**: folding can reuse allocation space for `P<T>` and `Vec<T>`, //! the latter even when the input and output types differ (as it would be the //! case with arenas or a GADT AST using type parameters to toggle features). //! //! * **Maintainability**: `P<T>` provides a fixed interface - `Deref`, //! `and_then` and `map` - which can remain fully functional even if the //! implementation changes (using a special thread-local heap, for example). //! Moreover, a switch to, e.g. `P<'a, T>` would be easy and mostly automated. use std::fmt::{self, Display, Debug}; use std::hash::{Hash, Hasher}; use std::ops::Deref; use serialize::{Encodable, Decodable, Encoder, Decoder}; /// An owned smart pointer. pub struct P<T> { ptr: Box<T> } #[allow(non_snake_case)] /// Construct a `P<T>` from a `T` value. pub fn P<T: 'static>(value: T) -> P<T> { P { ptr: Box::new(value) } } impl<T: 'static> P<T> { /// Move out of the pointer. /// Intended for chaining transformations not covered by `map`. pub fn and_then<U, F>(self, f: F) -> U where F: FnOnce(T) -> U, { f(*self.ptr) } /// Transform the inner value, consuming `self` and producing a new `P<T>`. pub fn map<F>(self, f: F) -> P<T> where F: FnOnce(T) -> T, { P(f(*self.ptr)) } } impl<T> Deref for P<T> { type Target = T; fn deref<'a>(&'a self) -> &'a T { &*self.ptr } } impl<T: 'static + Clone> Clone for P<T> { fn clone(&self) -> P<T> { P((**self).clone()) } } impl<T: PartialEq> PartialEq for P<T> { fn eq(&self, other: &P<T>) -> bool { **self == **other } } impl<T: Eq> Eq for P<T> {} impl<T: Debug> Debug for P<T> { fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { Debug::fmt(&**self, f) } } impl<T: Display> Display for P<T> { fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { Display::fmt(&**self, f) } } impl<T> fmt::Pointer for P<T> { fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { fmt::Pointer::fmt(&self.ptr, f) } } impl<T: Hash> Hash for P<T> { fn hash<H: Hasher>(&self, state: &mut H) { (**self).hash(state); } } impl<T: 'static + Decodable> Decodable for P<T> { fn decode<D: Decoder>(d: &mut D) -> Result<P<T>, D::Error> { Decodable::decode(d).map(P) } } impl<T: Encodable> Encodable for P<T> { fn encode<S: Encoder>(&self, s: &mut S) -> Result<(), S::Error> { (**self).encode(s) } }