Having the data's schema be optional at the level of the 'Schema' did not allow to represent cases where there would be an opaque data at an arbitrary nesting. So I introduced a new variant 'Opaque' on 'Data' to fill that gap.
These functions relied on the same dependency and had the same scope. So insertion was by encounter rather than order determined by dependency handling. Now we switched to dependency order is prioritized to prevent free unique.
-Builitins IR now acts like Record IR in terms of argument consumption
-UnConstrData returns as Pair(Data,Data) to conform with how pairs are treated behind the scenes.
This has been removed from the CIP-0057 specification since validators
are often re-used for multiple purposes (especially validators with
arity 2). It's misleading to assign a validator a purpose since the
purpose distinction actually happens _within_ the validator itself.
This has been bothering me and the more I thought of it the more I
disliked the idea of a warning. The rationale being that in this very
context, there's absolutely no ambiguity. So it is only frustrating
that the parser is even able to make the exact suggestion of what
should be fixed, but still fails.
I can imagine it is going to be very common for people to type:
```
trace "foo"
```
...yet terribly frustrating if they have to remember each time that
this should actually be a string. Because of the `trace`, `todo` and
`error` keywords, we know exactly the surrounding context and what to
expect here. So we can work it nicely.
However, the formatter will re-format it to:
```
trace @"foo"
```
Just for the sake of remaining consistent with the type-system. This
way, we still only manipulate `String` in the AST, but we conveniently
parse a double-quote utf-8 literal when coupled with one of the
specific keywords.
I believe that's the best of both worlds.
This will probably save people minutes/hours of puzzled debugging. This is only a warning because there may be cases where one do actually want to specify an hex-encoded bytearray. In which case, they can get rid of the warning by using the plain bytearray syntax (i.e. as an array of bytes).
The core observation is that **in the context of Aiken** (i.e. on-chain logic)
people do not generally want to use String. Instead, they want
bytearrays.
So, it should be easy to produce bytearrays when needed and it should
be the default. Before this commit, `"foo"` would parse as a `String`.
Now, it parses as a `ByteArray`, whose bytes are the UTF-8 bytes
encoding of "foo".
Now, to make this change really "fool-proof", we now want to:
- [ ] Emit a parse error if we parse a UTF-8 bytearray literal in
place where we would expect a `String`. For example, `trace`,
`error` and `todo` can only be followed by a `String`.
So when we see something like:
```
trace "foo"
```
we know it's a mistake and we can suggest users to use:
```
trace @"foo"
```
instead.
- [ ] Emit a warning if we ever see a bytearray literals UTF-8, which
is either 56 or 64 character long and is a valid hexadecimal string.
For example:
```
let policy_id = "29d222ce763455e3d7a09a665ce554f00ac89d2e99a1a83d267170c6"
```
This is _most certainly_ a mistake, as this generates a ByteArray of
56 bytes, which is effectively the hex-encoding of the provided string.
In this scenario, we want to warn the user and inform them they probably meant to use:
```
let policy_id = #"29d222ce763455e3d7a09a665ce554f00ac89d2e99a1a83d267170c6"
```
This is not supported by the code generation, so it's a bit of a lie
to have them in the language in the first place. There's arguably not
even any use for constant records, list and tuples to begin with. So
this cleans this up everywhere for the sake of moving forward with the
alpha release.
This now reduces constants to:
- Integer
- ByteArray
- String
Anything else can be declared via a function anyway. We can revisit
this choice later.... or not.
Tracing is now turn OFF by default when:
- building project
- building documentation
- building dependencies
It can be turned ON only when building project using `--keep-traces`.
That means it's not possible to build dependencies with traces. The
address `--rebuild` flag will also rebuild without traces.
Tracing is however turn ON by default when:
- checking the project (and running tests).
In this scenario, tracing can be disabled using `--no-traces` (if for
example, one want to analyze the execution units of specific functions
without having to manually remove traces from code).
This caused me some trouble. In my first approach, I ended up having
multiple traces because nested values would be evaluated twice; once
as condition, and once as part of the continuation.
To prevent this, we can simply evaluate the condition once, and return
plain True / False boolean as outcome. So this effectively transforms any
expression:
```
expr
```
as
```
if expr { True } else { trace("...", False) }
```
Interestingly enough, chumsky seems to fail when given a 'choice' with
more than 25 elements. That's why this commit groups together some of
the choices as another nested 'choice'.
The goal is to handle this without bothering the code generation down the line. That is, we can handle it when transforming from the untyped AST to the typed one. That's why there's no 'TraceIfFalse' constructor in the typed AST. It has disappeared during type-check.
We want the lookup to yield a result when there's only a single
validator; and no title is provided. So that users can simply do
'aiken address' in their project if it's unambiguous. The validator's
name is only required to disambiguate between multiple validators.
I also noticed that the order of arguments in with_validator was
wrong. Somehow.
Todo is fundamentally just a trace and an error. The only reason we kept it as a separate element in the AST is for the formatter to work out whether it should format something back to a todo or something else.
However, this introduces redundancy in the code internally and makes the AIR more complicated than it needs to be. Both todo and errors can actually be represented as trace + errors, and we only need to record their preferred shape when parsing so that we can format them back to what's expected.
We now parse errors as a combination of a trace plus and error term. This is a baby step in order to simplify the code generation down the line and the internal representation of todo / errors.
This however enforces that the argument unifies to a `String`. So this
is more flexible than the previous form, but does fundamentally the
same thing.
Fixes#378.
Not sure what this special case was trying to achieve, but it's not right. There's no need to handle function call with a single argument differently than the others.
List Clauses patterns handle var cases
Fixed Tuple Clauses issue with last clause not being a tuple
Redid how zero arg functions and dependencies are handled. Tough one lol
And also return a structured output as JSON, so it's more easily used
by other tools.
```
Parsing script context
Simulating 78ec148ea647cf9969446891af31939c5d57b275a2455706782c6183ef0b62f1
Redeemer Spend → 0
{"mem":151993,"cpu":58180696}
```
The current implementation assumed that ALL withdrawals present in a
transaction had to be locked by a script and failed otherwise. But a
transaction can actually be composed of both. So instead of failing,
we should rather just ignore withdrawals that can't be referenced by
redeemers.
There's arguably no use case ever for that in the context of on-chain
Plutus. Strings are really just meant to be used for tracing. They
aren't meant to be manipulated as heavily as in classic programming
languages.
Before that commit, the type-checker would allow unsafe list patterns
such as:
```
let [x] = xs
when xs is {
[x] -> ...
[x, ..] -> ...
}
```
This is quite unsafe and can lead to confusing situations. Now at
least the compiler warns about this. It isn't perfect though,
especially in the presence of clause guards. But that's a start.
Whoopsie... || and && were treated with the same precedence, causing very surprising behavior down the line.
I noticed this because of the auto-formatter adding parenthesis where it really shouldn't. The problem came actually from the parser and how it constructed the AST.
fix conversion from inner opaque type for when and assignment
This fixes Clause being used in cases where ListClause or TupleClause should be used
Reset defined and zero arg functions between each code gen
Fixes for optimizations when encountering shadowed variables
This is still a bit clunky as the interface is expecting parameters in UPLC form and we don't do any kind of verification. So it is easy to shoot oneself in the foot at the moment (for example, to apply an integer into something that should have received a data). To be improved later.
Without that, we have no way to distinguish between fully applied
validators and those that still require some hard-coded parameters.
Next steps is to make it easier to apply parameters to those, as well
as forbid the creation of addresses of validators that aren't fully
qualified.
* fix assert on pattern Var
* fix tuple index unwrapping closes#334
* allow wrapping when casting with let
* allow wrapping when casting via function call
I decided to invert how I'm doing it. I'm passing
in a new argument to unify in environment called
allow_cast: bool and essentially at various
unification sites I can control whether or not I
want to allow casting to even occur. So we can
assume it's false by default always and then we
turn it on in a few places vs. just opening the
flood gates and locking it down at various sites
as they come up# Please enter the commit message
for your changes. Lines starting
* you cannot cast FROM Data with a `let`
* you cannot cast FROM Data by passing
Data to none Data when calling a function
* you MUST use `assert` to cast from data
* you can cast INTO Data with a `let`
* you can cast INTO Data by passing none Data
to Data when calling a function
* You cannot assert cast Data without an
annotation
Weirdly enough, we got the parsing wrong for byte literals in expressions (but did okay in constants). But got the formatting wrong in constants (yet did okay for formatting expressions). I've factored out the code in both cases to avoid the duplication that led to this in the first place. Plus added test coverage to make sure this doesn't happen in the future.
This calculates a validator's address from validators found in a blueprint. It also provides a convenient way to attach a delegation part to the validator if needs be. The command is meant to provide a nice user experience and works 'out of the box' for projects that have only a single validator. Just call 'aiken address' to get the validator's address.
Note that the command-line doesn't provide any option to configure the target network. This automatically assumes testnet, and will until we deem the project ready for mainnet. Those brave enough to run an Aiken's program on mainnet will find a way anyway.
Here's a trick though: I got lazy (a bit) and did not write a full deserializer for Schema because this is busywork and not at all necessary at this stage. Instead, I've made the blueprint parameterized by a generic type <T>; which represents the type of the underlying blueprint's schema. When deserializing from JSON, we can default to 'Value' to get a free deserializer. Since all we're interested about is the program and the metadata (purpose and title) of a validator, it works nicely.
Serialization however expects a Blueprint<Schema>, and most of the functions operates over a Blueprint<Schema> anyway.
This will be useful to re-use this behavior in other structure that contains a Program<DeBruijn> without having to manually serialize or deserialize the entire structure.
In an ideal world, I should have handlded that directly at the conflicting commit in the rebase, but this would have bubbled up through all commits... which I wasn't really quite keen on going through. So here's an extra ugly commit that comes and 'fix the rebase'.
This is quite something, because now we have a testing pipeline that
can also be used for testing other compiler-related stuff such as the
type-checker or the code generator.
This also now introduce two levels of representable types (because it's needed at least for tuples):
Plutus Data (a.k.a Data) and UPLC primitives / constants (a.k.a Schema).
In practice, we don't want to specify blueprints that use direct UPLC primitives because there's little support for producing those in the ecosystem. So we should aim for producing only Data whenever we can. Yet we don't want to forbid it either in case people know what they're doing. Which means that we need to capture that difference well in the type modelling (in Rust and in the CIP-0057 specification).
I've also simplified the error type for now, just to provide some degree of feedback while working on this. I'll refine it later with proper errors.
The blueprint is generated at the root of the repository and is
intended to be versioned with the rest. It acts as a business card
that contains many practical information. There's a variety of tools
we can then build on top of open-source contracts. And, quite
importantly, the blueprint is language-agnostic; it isn't specific to
Aiken. So it is really meant as an interop format within the
ecosystem.
With pretty parse errors on failures. The type-checker was already
implemented for those, so it now only requires some work in the code
generation.
Fixes#297.
This is a bit annoying as we are forced to use #[related] here which isn't quite what we want.
Ideally, this would use #[diagnostic_source] but, there's a bug upstream. See: zkat/miette#172.
In the similar spirit to what we did for sequences. Yet, we need to handle the case of body being just an assignment -- or a trace of an assignment which is basically the same thing.
Far less verbose than defining classes by hand, plus, it allows to have everything about a single error be co-located. And finally, it allows to use 'related', 'label' and so on more easily.
While Gleam originally allowed various kinds of expressions to be discarded in a sequence, we simply do not allow expressions to be discarded implicitly. So any non-final expression in a sequence must be a let-binding. This prevents silly mistakes.
- Display function's signature next to the function name
(instead of being repeated below the function documentation).
- Same for module constants
- Display record constructors in a more concise manner, with
constructors fields next to constructors.
- Display generic parameters, if any, next to the type
- Plus some minor color and icon rework.
Somehow missed it when reworking tuples. We need to allow the new
'NewLineLeftParen' token in this situation as well. Especially because
this is what the formatter outputs.
Due to how PlutusData works it doesn't make sense
to allow user defined types to contain
functions.
```
type Foo {
bar: fn(Int) -> Int
}
```
The above definition will now return an error.
There are restrictions regarding how modules are called, but given that packages are tight to repositories anyway; there's no way someone can publish and use an aiken package on 'aiken-lang' without being part of the organization. So the restriction on the command-line is pointless. Plus, it prevents us from using 'aiken-lang' as a placeholder name for tutorials.
This makes it easier to add new dependencies, without having to
manually edit the `aiken.toml` file.
The command is accessible via two different paths:
- aiken deps add
or simply
- aiken add
for this is quite common to find at the top-level of the command-line,
and, we still want to keep commands for managing dependencies grouped
under a command sub-group and not all at the top-level. So we're
merely promoting that one for visibility.
This is a bit cleaner, as the 'cmd/new' had many on-the-fly functions
which are better scoped inside this module.
Plus, it plays nicely with the std::str::FromStr trait definition.
This possibly breaks many Aiken programs out there, but it's for the
best. We haven't released the alpha yet so we still have a bit of
freedom when it comes to breaking change.
Plus, the migration path is easy, simply run:
```
find . -name "*.ak" | xargs sed -i "s/#(/(/g"
```
(or `-i ''` on MacOS).
This allows in case of issues with dependencies to at least safely
remove cached packages. Before that, it could be hard to know where
are even located the cached files without looking at the source code.
```
Clearing /Users/ktorz/Library/Caches/aiken/packages
Removing aiken-lang-stdlib-7ca9e659688ea88e1cfdc439b6c20c4c7fae9985.zip
Removing aiken-lang-stdlib-main@04eb45df3c77f6611bbdff842a0e311be2c56390f0fa01f020d69c93ff567fe5.zip
Removing aiken-lang-stdlib-6b482fa00ec37fe936c93155e8c670f32288a686.zip
Removing aiken-lang-stdlib-1cedbe85b7c7e9c4036d63d45cad4ced27b0d50b.zip
Done
```
Aiken's build system uses an internal global cache system to avoid
downloading the same packages over and over across projects. However,
prior to this commit, the cache key would be based of the dependency
version which can be either:
- A commit hash
- A branch or tag name
However, in the latter case, it means that the very first time we end
up fetching a dependency will lock its version forever (or until the
cache is cleared). This was inconvenient.
This commit changes that so that we use not only a branch name as
cache key, but additionally, the etag returned by the GitHub API
server. The etag is part of the HTTP headers, so it can be fetched
quickly using a simple HEAD request. It changes whenever the content
behind the endpoint changes -- which happens to be exactly what we
want. With this, we can quickly check whether an upstream package has
been updated and download the latest version should users have
specified a branch name as a version number.
For example, my current cache now looks as follow:
```
/Users/ktorz/Library/Caches/aiken/packages/
├── aiken-lang-stdlib-1cedbe85b7c7e9c4036d63d45cad4ced27b0d50b.zip
├── aiken-lang-stdlib-6b482fa00ec37fe936c93155e8c670f32288a686.zip
├── aiken-lang-stdlib-7ca9e659688ea88e1cfdc439b6c20c4c7fae9985.zip
└── aiken-lang-stdlib-main@04eb45df3c77f6611bbdff842a0e311be2c56390f0fa01f020d69c93ff567fe5.zip
```
This changes allow to use parenthesis `(` `)` to encapsulate
expressions in addition to braces `{` `}` used to define blocks.
The main use-case is for arithmetic and boolean expressions for which
developers are used to using parenthesis. For example:
```
{ 14 + 42 } * 1337
```
can now be written as:
```
( 14 + 42 ) * 1337
```
This may sound straightforward at first but wasn't necessarily trivial
in Aiken given that (a) everything is an expression, (b) whitespaces
do not generally matter and (c) there's no symbol indicating the end
of a 'statement' (because there's no statement).
Thus, we have to properly disambiguate between:
```
let foo = bar(14 + 42)
```
and
```
let foo = bar
(14 + 42)
```
Before this commit, the latter would be interpreted as a function call
and would lead to a somewhat puzzling error. Now, the newline serves
as a delimiting symbol. The trade-off being that for a function call,
the left parenthesis has to be on the same line as the function name
identifier -- which is a fair trade off. So this is still allowed:
```
let foo = bar(
14 + 42
)
```
As there's very little ambiguity about it.
This fixes#236 and would seemingly allow us to get rid of the leading
`#` in front of tuples.
* add unary op
* parse, typecheck, and code gen it
* express boolean not as unary op as well, previously called negate
Co-authored-by: rvcas <x@rvcas.dev>
KtorZ