Merge pull request #148 from multiformats/feat/const-generic

implement const generics

Author: Stebalien

Cargo.lock

@@ -23,7 +23,7 @@ derive = ["multihash-derive"]
 arb = ["quickcheck", "rand"]
 secure-hashes = ["blake2b", "blake2s", "blake3", "sha2", "sha3"]
 scale-codec = ["parity-scale-codec"]
-serde-codec = ["serde", "generic-array/serde"]
+serde-codec = ["serde", "serde-big-array"]
 
 blake2b = ["blake2b_simd"]
 blake2s = ["blake2s_simd"]
@@ -34,11 +34,11 @@ sha3 = ["digest", "sha-3"]
 strobe = ["strobe-rs"]
 
 [dependencies]
-generic-array = "0.14.4"
 parity-scale-codec = { version = "2.1.1", default-features = false, features = ["derive"], optional = true }
 quickcheck = { version = "0.9.2", optional = true }
 rand = { version = "0.7.3", optional = true }
-serde = { version = "1.0.116", default-features = false, features = ["derive"], optional = true }
+serde = { version = "1.0.116", optional = true, default-features = false, features = ["derive"] }
+serde-big-array = { version = "0.3.2", optional = true, features = ["const-generics"] }
 multihash-derive = { version = "0.7.2", path = "derive", default-features = false, optional = true }
 unsigned-varint = { version = "0.7.0", default-features = false }
 

Cargo.toml

@@ -33,6 +33,8 @@ multihash = "*"
 
 Then run `cargo build`.
 
+MSRV 1.51.0 due to use of const generics
+
 ## Usage
 
 ```rust
@@ -50,14 +52,14 @@ You can derive your own application specific code table:
 
 ```rust
 use multihash::derive::Multihash;
-use multihash::{MultihashDigest, U32, U64};
+use multihash::MultihashCode;
 
 #[derive(Clone, Copy, Debug, Eq, Multihash, PartialEq)]
-#[mh(alloc_size = U64)]
+#[mh(alloc_size = 64)]
 pub enum Code {
-    #[mh(code = 0x01, hasher = multihash::Sha2_256, digest = multihash::Sha2Digest<U32>)]
+    #[mh(code = 0x01, hasher = multihash::Sha2_256, digest = multihash::Sha2Digest<32>)]
     Foo,
-    #[mh(code = 0x02, hasher = multihash::Sha2_512, digest = multihash::Sha2Digest<U64>)]
+    #[mh(code = 0x02, hasher = multihash::Sha2_512, digest = multihash::Sha2Digest<64>)]
     Bar,
 }
 

README.md

@@ -8,8 +8,8 @@
 //!
 //! If you set `#mh(alloc_size = …)` to a too low value, you will get compiler errors. Please note
 //! the the sizes are checked only on a syntactic level and *not* on the type level. This means
-//! that digest need to have a size generic, which is a valid `typenum`, for example `U32` or
-//! `generic_array::typenum::U64`.
+//! that digest need to have a size const generic, which is a valid `usize`, for example `32` or
+//! `64`.
 //!
 //! You can disable those compiler errors with setting the `no_alloc_size_errors` attribute. This
 //! can be useful if you e.g. have specified type aliases for your hash digests and you are sure
@@ -19,14 +19,14 @@
 //!
 //! ```
 //! use multihash::derive::Multihash;
-//! use multihash::{MultihashDigest, U32, U64};
+//! use multihash::MultihashDigest;
 //!
 //! #[derive(Clone, Copy, Debug, Eq, Multihash, PartialEq)]
-//! #[mh(alloc_size = U64)]
+//! #[mh(alloc_size = 64)]
 //! pub enum Code {
-//!     #[mh(code = 0x01, hasher = multihash::Sha2_256, digest = multihash::Sha2Digest<U32>)]
+//!     #[mh(code = 0x01, hasher = multihash::Sha2_256, digest = multihash::Sha2Digest<32>)]
 //!     Foo,
-//!     #[mh(code = 0x02, hasher = multihash::Sha2_512, digest = multihash::Sha2Digest<U64>)]
+//!     #[mh(code = 0x02, hasher = multihash::Sha2_512, digest = multihash::Sha2Digest<64>)]
 //!     Bar,
 //! }
 //!