derive_more

API docs
Changelog

Breaking changes

• use derive_more::SomeTrait now imports macro only. Importing macro with its trait along is possible now via use derive_more::with_trait::SomeTrait. (#406)
• Top-level #[display("...")] attribute on an enum now has defaulting behavior instead of replacing when no wrapping is possible (no _variant placeholder). (#395)

Fixed

• Associated types of type parameters not being treated as generics in Debug and Display expansions. (#399)
• unreachable_code warnings on generated code when ! (never type) is used. (#404)
• Ambiguous associated item error when deriving TryFrom, TryInto or FromStr with an associated item called Error or Err respectively. (#410)
• Top-level #[display("...")] attribute on an enum being incorrectly treated
  as transparent or wrapping. (#395)
• Omitted raw identifiers in Debug and Display expansions. (#431)
• Incorrect rendering of raw identifiers as field names in Debug expansions. (#431)
• Top-level #[display("...")] attribute on an enum not working transparently for directly specified fields. (#438)
• Incorrect dereferencing of unsized fields in Debug and Display expansions. (#440)

New Contributors

• @bluurryy made their first contribution in #410
• @minh-dng made their first contribution in #423

Full Changelog: v1.0.0...v2.0.0

derive_more is a library that adds derives for many of the standard library traits. By using this library the following code just works:

use derive_more::{Add, Display, From, Into};

#[derive(PartialEq, From, Add)]
struct MyInt(i32);

#[derive(PartialEq, From, Into)]
struct Point2D {
    x: i32,
    y: i32,
}

#[derive(PartialEq, From, Add, Display)]
enum MyEnum {
    #[display("int: {_0}")]
    Int(i32),
    Uint(u32),
    #[display("nothing")]
    Nothing,
}

assert!(MyInt(11) == MyInt(5) + 6.into());
assert!((5, 6) == Point2D { x: 5, y: 6 }.into());
assert!(MyEnum::Int(15) == (MyEnum::Int(8) + 7.into()).unwrap());
assert!(MyEnum::Int(15).to_string() == "int: 15");
assert!(MyEnum::Uint(42).to_string() == "42");
assert!(MyEnum::Nothing.to_string() == "nothing");

Now, more than 8 years after the first commit and almost 5 years after the 0.99.0 release, derive_more has finally reached its 1.0.0 release. This release contains a lot of changes (including some breaking ones) to make it easier to use the derives and make it possible to extend them without having to break backwards compatibility again. There are five major changes that I would like to call out, but there are many more changes that are documented below:

1. There is a new Debug derive that can be used to easily customize Debug formatting.
2. A greatly improved Display derive, which allows you to do anything that thiserror provides, but it works for any type not just errors. And by combining the Display derive with the Error and From derives, there shouldn't really be any need to use thiserror anymore (if you are missing a feature/behaviour from thiserror please report an issue).
3. Traits that can return errors now return a type that implements Error when an error occurs instead of a &'static str.
4. When using use derive_more::SomeTrait the actual trait is also imported not just the derive macro. This is especially useful for Error and
   Display
5. The docs are now rendered on docs.rs and are much better overall.

Breaking changes

• The minimum supported Rust version (MSRV) is now Rust 1.75.
• Add the std feature which should be disabled in no_std environments.
• All Cargo features, except std, are now disabled by default. The full feature can be used to get the old behavior of supporting all possible derives.
• The TryFrom, Add, Sub, BitAnd, BitOr, BitXor, Not and Neg derives now return a dedicated error type instead of a &'static str on error.
• The FromStr derive now uses a dedicated FromStrError error type instead of generating unique one each time.
• The Display derive (and other fmt-like ones) now uses #[display("...", (<expr>),*)] syntax instead of #[display(fmt = "...", ("<expr>"),*)], and #[display(bound(<bound>))] instead of #[display(bound = "<bound>")]. So without the double quotes around the expressions and bounds.
• The Debug and Display derives (and other fmt-like ones) now transparently delegate to the inner type when #[display("...", (<expr>),*)] attribute is trivially substitutable with a transparent call. (#322)
• The DebugCustom derive is renamed to just Debug (gated now under a separate debug feature), and its semantics were changed to be a superset of std variant of Debug.
• The From derive doesn't derive From<()> for enum variants without any fields anymore. This feature was removed because it was considered useless in practice.
• The From derive now uses #[from(<types>)] instead of #[from(types(<types>))] and ignores field type itself.
• The Into derive now uses #[into(<types>)] instead of #[into(types(<types>))] and ignores field type itself.
• The Into derive now generates separate impls for each field whenever the #[into(...)] attribute is applied to it. (#291)
• Importing a derive macro now also imports its corresponding trait.
• The Error derive is updated with changes to the error_generic_member_access unstable feature for nightly users. (#200, #294)
• The as_mut feature is removed, and the AsMut derive is now gated by the as_ref feature. (#295)
• A top level #[display("...")] attribute on an enum now requires the usage of {_variant} to include the variant instead of including it at {}. The reason is that {} now references the first argument to the format string, just like in all other format strings. (#377)

Added

• Add support captured identifiers in Display derives. So now you can use: #[display(fmt = "Prefix: {field}")] instead of needing to use #[display(fmt = "Prefix: {}", field)]
• Add FromStr derive support for enums that contain variants without fields. If you pass the name of the variant to from_str it will create the matching variant.
• Add #[unwrap(owned, ref, ref_mut)] attribute for the Unwrap derive. By using them, it is possible to derive implementations for the reference types as well. (#206)
• Add TryUnwrap derive similar to the Unwrap derive. This one returns a Result and does not panic. (#206)
• Add support for container format in Debug derive with the same syntax as Display derives. (#279)
• derive_more::derive module exporting only macros, without traits. (#290)
• Add support for specifying concrete types to AsRef/AsMut derives. (#298)
• Add TryFrom derive for enums to convert from their discriminant. (#300)
• #[inline] attributes to IsVariant and Debug implementations. (#334
• Add #[track_caller] to Add, Mul, AddAssign and MulAssign derives (#378

Changed

• The Constructor and IsVariant derives now generate const fn functions.
• Static methods derived by IsVariant are now marked #[must_use]. (#350)
• The Unwrap and IsVariant derives now generate doc comments.
• #[automatically_derived] is now emitted from all macro expansions. This should prevent code style linters from attempting to modify the generated code.
• Upgrade to syn 2.0.
• The Error derive now works in nightly no_std environments

Fixed

• Use a deterministic HashSet in all derives, this is needed for rust analyzer to work correctly.
• Use Provider API for backtraces in Error derive.
• Fix Error derive not working with const generics.
• Support trait objects for source in Error, e.g. Box<dyn Error + Send + 'static>
• Fix bounds on derived IntoIterator impls for generic structs. (#284)
• Fix documentation of generated bounds in Display derive. (#297)
• Hygiene of macro expansions in presence of custom core crate. (#327)
• Fix documentation of generated methods in IsVariant derive.
• Make {field:p} do the expected thing in format strings for Display and Debug. Also document weirdness around Pointer formatting when using expressions, due to field variables being references. (#381)

New Contributors

• @Fuuzetsu made their first contribution in #178
• @Robrisi made their first contribution in #183
• @Technohacker made their first contribution in #179
• @aj-bagwell made their first contribution in #176
• @momirza made their first contribution in #187
• @phip1611 made their first contribution in #192
• @tvercruyssen made their first contribution in #196
• @fmiguelgarcia made their first contribution in #218
• @zohnannor made their first contribution in #238
• @rassvet2 made their first contribution in #206
• @dependabot made their first contribution in #251
• @pymongo made their first contribution in #264
• @ModProg made their first contribution in #279
• @MegaBluejay made their first contribution in #284
• @fritzrehde made their first contribution in #310
• @DaniPopes made their first contribution in #313
• @TheLostLambda made their first contribution in #350
• @alpha-tango-kilo made their first contribution in #360
• @wangcundashang made their first contribution in #372
• @DenisGorbachev made their first contribution in #374
• @sornas made their first contribution in #371

Full Changelog: v0.99.18...v1.0.0

API docs
Changelog

chore: Release derive_more version 0.99.20

chore: Release derive_more version 0.99.19

chore: Release derive_more version 0.99.18

Added

• From supports additional types for conversion: #[from(types(u8, u16))].

Changed

• When specifying specific features of the crate to only enable specific derives, the extra-traits feature of syn is not always enabled when those the specified features do not require it. This should speed up compile time of syn when this feature is not needed.

Fixed

• Fix generic derives for MulAssign

Changed

• Make sure output of derives is deterministic, for better support in rust-analyzer

Added

• Support for deriving Error!!! (many thanks to @ffuugoo and @tyranron)

Fixed

• Fix generic bounds for Deref and DerefMut with forward, i.e. put Deref bound on whole type, so on where Box<T>: Deref instead of on T: Deref. (#107)

• The tests directory is now correctly included in the crate (requested by Debian package maintainers)

Note: This version is yanked, because quickly after release it was found out tests did not run in CI.

Fixed

• Fix generic bounds for Deref and DerefMut with no forward, i.e. no bounds are necessary. (#107)

Fixed

• Hotfix for a regression in allowed Display derives using # flag, such as {:#b} (#107)

Fixed

• Hotfix for a regression in allowed From derives (#105)

This release is a huge milestone for this library. Lot's of new derives are implemented and a ton of attributes are added for configuration purposes. These attributes will allow future releases to add features/options without breaking backwards compatibility. This is why the next release with breaking changes is planned to be 1.0.0.

Breaking changes

• The minimum supported rust version (MSRV) is now Rust 1.36.
• When using in a Rust 2015 crate, you should add extern crate core to your code.
• no_std feature is removed, the library now supports no_std without having to configure any features.

Added

• Deref derives now dereference to the type in the newtype. So if you have MyBox(Box<i32>), dereferencing it will result in a Box<i32> not an i32. To get the old behaviour of forwarding the dereference you can add the #[deref(forward)] attribute on the struct or field.
• Derives for AsRef, AsMut, Sum, Product, IntoIterator.
• Choosing the field of a struct for which to derive the newtype derive.
• Ignoring variants of enums when deriving From, by using #[from(ignore)].
• Add #[from(forward)] attribute for From derives. This forwards the from calls to the fields themselves. So if your field is an i64 you can call from on an i32 and it will work.
• Add #[mul(forward)] and #[mul_assign(forward)], which implement Mul and MulAssign with the semantics as if they were Add/AddAssign.
• You can use features to cut down compile time of the crate by only compiling the code needed for the derives that you use. (see Cargo.toml for the features, by default they are all on)
• Add #[into(owned, ref, ref_mut)] and #[try_into(owned, ref, ref_mut)] attributes. These cause the Into and TryInto derives to also implement derives that return references to the inner fields.
• Allow #[display(fmt="some shared display text for all enum variants {}")] attribute on enum.
• Better bounds inference of Display trait.

Changed

• Remove dependency on regex to cut down compile time.
• Use syn 1.0

Fixed

• Automatic detection of traits needed for Display format strings

Added

• Added no_std support

Changed

• Suppress unused_variables warnings in derives

Added

• Extended Display-like derives to support custom formats

Changed

• Updated to syn v0.15

Changed

• Updated to syn v0.14, quote v0.6 and proc-macro2 v0.4

Changed

• Updated to latest version of syn and quote

Fixed

• Changed some URLs in the docs so they were correct on crates.io and docs.rs
• The Result type is now referenced in the derives using its absolute path (::std::result::Result) to make sure that the derives don't accidentally use another Result type that is in scope.

Added

• Allow deriving of TryInto
• Allow deriving of Deref
• Allow deriving of DerefMut

Added

• Allow deriving of Display, Binary, Octal, LowerHex, UpperHex, LowerExp, UpperExp, Pointer
• Allow deriving of Index
• Allow deriving of IndexMut

Fixed

• Allow cross crate inlining of derived methods

Added

• Allow deriving of FromStr

Changed

• Updated to latest version of syn and quote

Fixed

• Add #[allow(missing_docs)] to the Constructor definition

Changed

• Changed code to work with newer version of the syn library.

Changed

• Deriving From, Into and Constructor now works for empty structs.

Changed

• The new() method that is created when deriving Constructor is now public. This makes it a lot more useful.

Added

• Derives for Into, Constructor and MulAssign-like

Changed

• From is now derived for enum variants with multiple fields.

Fixed

• Derivations now support generics.

Added

• Lots of docs.
• Derives for Neg-like and AddAssign-like.

Changed

• From can now be derived for structs with multiple fields.