* move MaxEncodedLen trait from Substrate

* Move derive macro and tests from Substrate

* only run ui tests when derive feature enabled

* reduce note-taking documentation

Co-authored-by: Bastian Köcher <bkchr@users.noreply.github.com>

* Bless trybuild `tests/max_encoded_len_ui/union.rs` test

* Update docs (these files are not part of substrate)

Co-authored-by: Andronik Ordian <write@reusable.software>

* Mention new `MaxEncodedLen` trait in the CHANGELOG.md

* Prepare a 2.2.0-rc.1 release

This is a pre-release rather than a full release in order to help shape
the new `MaxEncodedLen` trait used in Substrate in case some more
involved changes are found out to be required.
The API did not change since its introduction until now so chances are
slim but it's good to leave some leeway.

* incorporate changes made to Substrate version of MaxEncodedLen

Note: doesn't include the MaxEncodedLen impl for H160, H256, H512.
A substrate companion will be necessary to re-add those.

* remove redundant no_std

* Rewrite fn max_encoded_len_trait for clarity

* simplify logic checking for invalid attr

Co-authored-by: Bastian Köcher <bkchr@users.noreply.github.com>

* remove bogus whitespace

Co-authored-by: Bastian Köcher <bkchr@users.noreply.github.com>

* use Path::is_ident() helper

Co-authored-by: Bastian Köcher <bkchr@users.noreply.github.com>

* rm unused import

Co-authored-by: Bastian Köcher <bkchr@users.noreply.github.com>
Co-authored-by: Igor Matuszewski <Xanewok@gmail.com>
Co-authored-by: Andronik Ordian <write@reusable.software>

* Bump proc-macro-crate to v1.0.0

Signed-off-by: koushiro <koushiro.cqx@gmail.com>

* Bump honggfuzz and arbitrary

Signed-off-by: koushiro <koushiro.cqx@gmail.com>

* Bump quickcheck to v1.0

Signed-off-by: koushiro <koushiro.cqx@gmail.com>

* Impl Decode/Encode for Range/RangeInclusive

* Typo

* Adds support for custom where bounds

The user can now specify a custom where bound when using the derive
macros:

- `#[codec(encode_bound(T: Encode))]` for `Encode`
- `#[codec(decode_bound(T: Encode))]` for `Decode`

If nothing is specified (`encode_bound()`) the where bounds will be empty.

* Apply suggestions from code review

* Apply suggestions from code review

Co-authored-by: Gavin Wood <gavin@parity.io>

* Switch array impls to const generics

* addr comments

Co-authored-by: Bastian Köcher <bkchr@users.noreply.github.com>

Co-authored-by: David <dvdplm@gmail.com>

Co-authored-by: Guillaume Thiolliere <gui.thiolliere@gmail.com>
Co-authored-by: Bastian Köcher <bkchr@users.noreply.github.com>

Co-authored-by: Andronik Ordian <write@reusable.software>
Co-authored-by: Bastian Köcher <bkchr@users.noreply.github.com>

Co-authored-by: Bastian Köcher <bkchr@users.noreply.github.com>
Co-authored-by: David <dvdplm@gmail.com>

* CI: add miri test; new phase; remoe CARGO_HOME cache

* add miri and falgs to test

* Update .gitlab-ci.yml

Co-authored-by: Guillaume Thiolliere <gui.thiolliere@gmail.com>

Co-authored-by: Cheme
Co-authored-by: Brian Anderson <andersrb@gmail.com>
Co-authored-by: Bastian Köcher <bkchr@users.noreply.github.com>
Co-authored-by: David <dvdplm@gmail.com>

* change (CI): cache pre-population script; use of cargo home and cargo target caches; rules instead only

* fix (CI): wrong pasta

* fix (CI): persimmons to the pre_pop script

This makes it easier to write code that uses `Error` with or without `std`.

Seek trait doesn't ensure that the length is fixed, thus our usage in IoReader is wrong. Now fixed.

* fix

* better impl

Normally, the span of a field has the same hygiene context as
`Span::call_site`. However, it's possible for a struct definition
to be 'constructed' via a `macro_rules` macro such that the field
has a different hygiene context. This will cause the expanded code
to be unable to resolve any references to `self`, resulting in a
compilation error.

This pull request uses `quote!` instead of `quote_spanned!` when
emitting a 'self' identifier. `quote_spanned!` is still used for
everything else in the emitted method, meaning that error messages
will still point to the proper field.

I've included a test case which triggers this issue on
Rust 1.43.1. It's current difficult to hit this issue
other than in this artificial case, but that will change
once rust-lang/rust#72622 is re-landed.

* decode_length for tuple'

* add test

* refactor

* Verify bytes after encoding an object that was decoded from original bytes

* Add BinaryHeapWrapper to allow PartialEq trait impl

* Add Arbitrary crate and derives

* Add prototype for encoding testing

* Add partial encode fuzz testing

* Derive arbitrary to compact

* Add types for encode fuzzing

* Simplify macro

* Remove BitVec from Encoding macro

* Add comments about sorting

* Add comment for BinaryHeapWrapper

* Update bytes sorting and move type display to the end

* Make raw1 immutable

* Hide arbitrary crate behind feature 'fuzz'

* Indent macro

* Impl Arbitrary for BitVecWrapper

* Differentiate flows depending on fuzzing input

* Update comments and add TODO

* Change minimal to only_decode

* Use more straightforward code for defining eq trait

Co-Authored-By: Bastian Köcher <bkchr@users.noreply.github.com>

* Put comments before derive

Co-Authored-By: Bastian Köcher <bkchr@users.noreply.github.com>

* Put comments before derive

Co-Authored-By: Bastian Köcher <bkchr@users.noreply.github.com>

* Use match in macros; Remove comments on into_iter_sorted and fix compile for paritaleq trait

* Use match in fuzz_encode

* Fix typo in comments

* Apply suggestions from code review

Co-Authored-By: Bastian Köcher <bkchr@users.noreply.github.com>

* Remove useless mut

Co-authored-by: Bastian Köcher <bkchr@users.noreply.github.com>