A release or so ago we added `non_exhaustive` to the `Network` enum,
this turned out to make usage of the enum un-ergonomic for downstream
users. After much debate we decided that a way forward was to just
minimize the usage of the enum in the public API by instead use
`AsRef<Params>` so that downstream could define their own network enum
based on the networks they support.
Minimize usage of `Network` by using `AsRef<Params>` as a parameter type
instead. "minimize" because the `Network` still appears in some places.
`require_network` is typically called as part of parsing, often in the
same line of code. Counter to our normal errors, it makes
`require_network` more ergonomic to use if we just return a `ParseError`
variant.
Close: #2507
We are currently using the `base58::Error` type to create errors in
`bitcoin`, these are bitcoin errors not `base58` errors.
Note that we add what looks like duplicate
`InvalidBase58PayloadLengthError` types but they are different because
of the expected length. This could have been a field but I elected not
to do so for two reasons:
1. We will need to do so anyways if we crate smash more
2. The `crypto::key` one can have one of two values 33 or 34.
With this applied we can remove the now unused error variants from
`base58::Error`.
The `address` module is currently publicly re-exporting all error types
that appear as return values for any pubic function, except for the
`P2shError` - we should be uniform.
This re-export of error thing has not been discussed/agreed upon as a
policy but I have been doing it for the last few months anytime I
introduced an `error` module - there has been no push back so I assumed
it was acceptable. Before 1.0 we should probably have a policy on this.
7e2a81d03b Remove unused address::Error type (Tobin C. Harding)
a92dc9c35c Add NetworkValidationError (Tobin C. Harding)
Pull request description:
Replaces #2502 because there is going to be way too much arguing on this to bother a newer contributor with.
This PR takes into consideration #2507 but does not improve the issue, it also does not make it worse. I propose to do this and then consider #2507 since this is a step forwards IMO.
Remove the `address::Error` because its not good. Add a `NetworkValidationError` and return it from `require_network` - leaving the door open for resolving Kix's issue in 2507.
ACKs for top commit:
Kixunil:
ACK 7e2a81d03b
apoelstra:
ACK 7e2a81d03b
Tree-SHA512: 0a220dbec1457f35e3d95f32399f258e65c0477e8b4d14dbe2dfad0a44966ab0339d4c2d9828da318bf131db45cecb2447c0e32d5669a5f4c4739b90c83d3c0d
9d688396c9 base58: Use pub extern crate instead of module (Tobin C. Harding)
Pull request description:
We don't add any implementations to the `base58` types so we can just `pub extern` the crate instead of using a module and re-exporting.
ACKs for top commit:
Kixunil:
ACK 9d688396c9
apoelstra:
ACK 9d688396c9
Tree-SHA512: 521af0fd1ae365a6d060dd3c38fc18c1fb3a6c46adb7cba64e53128c7a898c766bcc603078b4cb8a3672df96559633495b23203a33147b5b4959b0c690ab7451
This commit adds the `FromScriptError` struct to handle the errors
while generating address from any script. It includes :
- Unrecognized script error.
- Witness Program error.
- Witness Version error.
We have a bunch of functions that take `Network` when what they really
want is something that can be converted to a `KnownHrp`.
Make `KnownHrp` public and accept `impl Into<KnownHrp>`.
03bfe1d433 Impove rustdoc on assume_checked_ref (Tobin C. Harding)
769809f1f2 Improve the docs on as_unchecked function (Tobin C. Harding)
Pull request description:
In #1765 we added a couple of new functions.
- Patch 1: Fix mis-documented function.
- Patch 2: Do trivial rustdocs fix.
ACKs for top commit:
apoelstra:
ACK 03bfe1d433
Kixunil:
ACK 03bfe1d433
Tree-SHA512: 5be6b2d288c1f4e9096014acd8618dc84a3ec6f45ae38b5d44ef7f95eccc268021bc8e8435152166606d893d4238b03e59e8f9d4fc67ba9a6c33194f3f54fc40
The `as_unchecked` method is never dangerous to call because an
`Address<UncheckedNetwork>` provides a subset of functionality that is
always ok to use. It is only dangerous to go the other way unchecked to
checked.
P2WPKH requires keys to be compressed which introduces error handling
even in cases when it's statically known that a key is compressed. To
avoid it, this change introduces `CompressedPublicKey` which is similar
to `PublicKey` except it's statically known to be compressed.
This also changes relevant code to use `CompressedPublicKey` instead of
`PublicKey`.
An `AddressInner` struct is created when parsing address strings however
address strings do not map 1:1 to `Network` because signet and testnet
use the same bech32 prefix "tb".
We can fix this by inlining the `Payload` variants into `AddressInner`
and adding prefix enums for legacy addresses and a `KnownHrp` for bech32
addresses.
Also enables removing the `AddressEncoding` struct as we can display the
`AddressInner` struct directly. (The `Display` impl is on `AddressInner`
and not directly on address to ignore the `NetworkValidation` wrapper,
may be able to be simplified still further.)
Calling `wpubkey_hash` on a key that is uncompressed is flat out an
error, really it is a programmer error at build time because a segwit
key should never be compressed, however, for historical reasons we do
not enforce this in the type system. As a step towards clarity make it
an error to call `wpubkey_hash` on a an uncompressed pubkey. This adds
documentation and potentially might assist debugging for newer devs.
7d695f6b41 Improve public re-exports (Tobin C. Harding)
33774122e0 Remove public re-exports from private module (Tobin C. Harding)
Pull request description:
Improve the public exports in two ways:
1. Inline re-exports into the docs of the module that re-exports them.
2. Separate public and private use statements
Recently we discussed a way to separate the public and private import statements to make the code more clear and prevent `rustfmt` joining them all together.
Separate public exports using a code block and `#[rustfmt::skip]`. Has the nice advantage of reducing the number of `#[doc(inline)]` attributes also.
1. Modules first, as they are part of the project's structure.
2. Private imports
3. Public re-exports (using `rustfmt::skip` to prevent merge)
Use the format
```rust
mod xyz;
mod abc;
use ...;
pub use {
...,
};
```
This patch introduces changes to the rendered HTML docs.
ACKs for top commit:
apoelstra:
ACK 7d695f6b41
Tree-SHA512: dc9121c0fe282e3035d862beadb89e2d5a374a7dab6b1c3147a9b5960f8bc2f5af49892f0f713f55c645c46f53464c32daf390c11d85c75553b3ea7e0efc8246
Currently we have functions on `Address` that call through to a public
`Payload` type. The `Payload` type is an implementation detail and
should never have been public. In preparation for modifying the
`AddressInner` and removing `Payload` altogether lets move all the
functionality from `Payload` into `Address` - this is basically just
code moves so it is feasible to review with some confidence.
This is an API breaking change because it makes `Payload` private and
also removes from the pubic `Address` API functions that accept and
return `Payload`. Apart from that the changes can be seen as
refactoring.
The `AddressEncoding` type exists solely to assist us in implementing
`Display` on `Address`, it may have been used in the past by alt-coins
back when we had a more tolerant outlook on supporting them. Nowadays
we explicitly do not support alts.
Improve the public exports in two ways:
1. Inline re-exports into the docs of the module that re-exports them.
2. Separate public and private use statements
Recently we discussed a way to separate the public and private import
statements to make the code more clear and prevent `rustfmt` joining
them all together.
Separate public exports using a code block and `#[rustfmt::skip]`. Has
the nice advantage of reducing the number of `#[doc(inline)]` attributes
also.
1. Modules first, as they are part of the project's structure.
2. Private imports
3. Public re-exports (using `rustfmt::skip` to prevent merge)
Use the format
```rust
mod xyz;
mod abc;
use ...;
pub use {
...,
};
```
This patch introduces changes to the rendered HTML docs.
Update the `bech32` dependency to use the newly release beta version.
The main fix here is silent, a bug fix in `bech32` that was being hit by
our fuzzing suite.
The `address::Error` is module level general, we can make the code
easier to maintain and easier to stabalize by splitting the parse error
out of the general error.
Create a `ParseError` that is returned by `FromStr for Address`. Remove
the now unused variants from the general `address::Error`.
Split the error code out of `address/mod.rs` and into
`address/error.rs`. Code move only, no changes other than to
imports/exports etc. to make it build.
Fmt Bot
Tobin C. Harding
Andrew Poelstra
harshit933
conduition
Martin Habovstiak