internal/conformance API

internal/conformance is the shared result model of the conformance frontends: source locations, per-case results with a disposition, run summaries with fixed counting rules, and deterministic shard selection. frontend/gda_expr, frontend/itl_expr, frontend/mpfr_expr, frontend/testfloat_expr and internal/runner_cli build on it and wrap its types in their own public types. It is an internal package: code outside the Luna-Flow/floating module cannot import it, and its interface may change without notice. Examples on this page are therefore not compiled. See the tutorial and the design page.

Import (inside the module only):

import {
  "Luna-Flow/floating/internal/conformance",
}

Source locations

`SourceLocation`

SourceLocation is a position in a named source.

pub struct SourceLocation {
  // private fields
} derive(Eq, @debug.Debug)

`SourceLocation::new`

SourceLocation::new(source, line, column?) creates a location.

pub fn SourceLocation::new(String, Int, column? : Int) -> Self

column defaults to 1. Line and column are clamped to at least 1, so a location always points at a real character position.

`SourceLocation::source`, `SourceLocation::line`, `SourceLocation::column`

These accessors return the components.

pub fn SourceLocation::source(Self) -> String
pub fn SourceLocation::line(Self) -> Int
pub fn SourceLocation::column(Self) -> Int

///|
test "locations are one-based" {
  let location = @conformance.SourceLocation::new("a.txt", 0, column=-3)
  inspect(location.line(), content="1")
  inspect(location.column(), content="1")
}

Case results

`CaseDisposition`

CaseDisposition classifies a selected case.

pub(all) enum CaseDisposition {
  Executable
  Diagnostic(String)
  Legacy(String)
  Unsupported(String)
} derive(Eq, @debug.Debug)

Executable: the case was run; its passed flag is meaningful.
Diagnostic(reason): the case is not a valid test of this kind (for example a placeholder row).
Legacy(reason): the case belongs to an older version of the corpus format.
Unsupported(reason): the case is valid but the implementation does not provide the feature.

`CaseResult`

CaseResult is the outcome of one case.

pub struct CaseResult {
  // private fields
} derive(Eq, @debug.Debug)

`CaseResult::new`, `CaseResult::executable`

CaseResult::new(id, disposition, passed, message?) builds a result; CaseResult::executable(id, passed, message?) is new with Executable.

pub fn CaseResult::new(String, CaseDisposition, Bool, message? : String) -> Self
pub fn CaseResult::executable(String, Bool, message? : String) -> Self

message defaults to "". No consistency check is made between disposition and passed; summaries only read passed for executable results.

`CaseResult::id`, `disposition`, `passed`, `message`

These accessors return the stored fields.

pub fn CaseResult::id(Self) -> String
pub fn CaseResult::disposition(Self) -> CaseDisposition
pub fn CaseResult::passed(Self) -> Bool
pub fn CaseResult::message(Self) -> String

Sharding

`ShardSpec`

ShardSpec is a validated pair (shard count, shard index).

pub struct ShardSpec {
  // private fields
} derive(Eq, @debug.Debug)

`ShardSpec::try_new`, `ShardSpec::new`

ShardSpec::try_new(count, index) validates and builds a shard; ShardSpec::new aborts with the same message instead of returning Err.

pub fn ShardSpec::try_new(Int, Int) -> Result[Self, String]
pub fn ShardSpec::new(Int, Int) -> Self

Errors: "shard count must be positive" when count <= 0, and "shard index must be within shard count" unless 0 <= index < count.

`ShardSpec::count`, `ShardSpec::index`

These accessors return the count $n$ and the index $i$ .

pub fn ShardSpec::count(Self) -> Int
pub fn ShardSpec::index(Self) -> Int

`ShardSpec::selects`

shard.selects(k) is true when the case with ordinal k belongs to this shard, that is when $k \bmod n = i$ .

pub fn ShardSpec::selects(Self, Int) -> Bool

Ordinals are expected to be non-negative (MoonBit’s % keeps the sign of the dividend).

///|
test "round-robin selection" {
  let shard = @conformance.ShardSpec::new(3, 1)
  let picked = [0, 1, 2, 3, 4, 5, 6].filter(k => shard.selects(k))
  debug_inspect(picked, content="[1, 4]")
  inspect(@conformance.ShardSpec::try_new(0, 0) is Err(_), content="true")
}

Run summaries

`RunSummary`

RunSummary counts the results of one run or shard.

pub struct RunSummary {
  // private fields
} derive(Eq, @debug.Debug)

`RunSummary::from_results`

RunSummary::from_results(total, results) counts results and records total as the number of cases before shard selection.

pub fn RunSummary::from_results(Int, Array[CaseResult]) -> Self

Each result is counted once: an Executable result as passed or failed according to passed(), any other result as skipped and in its own disposition counter. The results are copied.

`RunSummary::merge`

RunSummary::merge(parts) combines shard summaries.

pub fn RunSummary::merge(Array[Self]) -> Self

All counters are summed except total_cases, which is the maximum over the parts; results are concatenated in the order of parts. An empty array gives an empty summary.

Counters, `results` and `success`

These accessors return the counters, a copy of the results, and the verdict.

pub fn RunSummary::total_cases(Self) -> Int
pub fn RunSummary::selected_cases(Self) -> Int
pub fn RunSummary::executable_cases(Self) -> Int
pub fn RunSummary::passed_cases(Self) -> Int
pub fn RunSummary::failed_cases(Self) -> Int
pub fn RunSummary::skipped_cases(Self) -> Int
pub fn RunSummary::diagnostic_cases(Self) -> Int
pub fn RunSummary::legacy_cases(Self) -> Int
pub fn RunSummary::unsupported_cases(Self) -> Int
pub fn RunSummary::results(Self) -> Array[CaseResult]
pub fn RunSummary::success(Self) -> Bool

selected_cases is the number of results. The identities

\begin{aligned} \text{selected} &= \text{executable} + \text{skipped}, \\ \text{executable} &= \text{passed} + \text{failed}, \\ \text{skipped} &= \text{diagnostic} + \text{legacy} + \text{unsupported} \end{aligned}

hold for every summary built by from_results or merge. success() is failed_cases() == 0; skipped cases never make it false. Frontends that need a stricter verdict (itl_expr also fails on diagnostics) add their own rule.

///|
test "summary counting" {
  let summary = @conformance.RunSummary::from_results(10, [
    @conformance.CaseResult::executable("a", true),
    @conformance.CaseResult::executable("b", false, message="mismatch"),
    @conformance.CaseResult::new("c", @conformance.Unsupported("op"), false),
  ])
  inspect(summary.total_cases(), content="10")
  inspect(summary.selected_cases(), content="3")
  inspect(summary.skipped_cases(), content="1")
  inspect(summary.success(), content="false")
}

Trait implementations

Equality and `Debug`

All five types derive Eq and Debug; these promoted methods compare all fields and render values. Use ==, != and debug_inspect in new code.

pub fn SourceLocation::equal(Self, Self) -> Bool
pub fn SourceLocation::not_equal(Self, Self) -> Bool
pub fn SourceLocation::to_repr(Self) -> @debug.Repr
pub fn CaseDisposition::equal(Self, Self) -> Bool
pub fn CaseDisposition::not_equal(Self, Self) -> Bool
pub fn CaseDisposition::to_repr(Self) -> @debug.Repr
pub fn CaseResult::equal(Self, Self) -> Bool
pub fn CaseResult::not_equal(Self, Self) -> Bool
pub fn CaseResult::to_repr(Self) -> @debug.Repr
pub fn ShardSpec::equal(Self, Self) -> Bool
pub fn ShardSpec::not_equal(Self, Self) -> Bool
pub fn ShardSpec::to_repr(Self) -> @debug.Repr
pub fn RunSummary::equal(Self, Self) -> Bool
pub fn RunSummary::not_equal(Self, Self) -> Bool
pub fn RunSummary::to_repr(Self) -> @debug.Repr

Complete public interface

This snapshot is the generated pkg.generated.mbti of the package. It is the authority when prose and interface disagree.

// Generated using `moon info`, DON'T EDIT IT
package "Luna-Flow/floating/internal/conformance"

import {
  "moonbitlang/core/debug",
}

// Values

// Errors

// Types and methods
pub(all) enum CaseDisposition {
  Executable
  Diagnostic(String)
  Legacy(String)
  Unsupported(String)
} derive(Eq, @debug.Debug)
pub fn CaseDisposition::equal(Self, Self) -> Bool
pub fn CaseDisposition::not_equal(Self, Self) -> Bool
pub fn CaseDisposition::to_repr(Self) -> @debug.Repr

pub struct CaseResult {
  // private fields
} derive(Eq, @debug.Debug)
pub fn CaseResult::disposition(Self) -> CaseDisposition
pub fn CaseResult::equal(Self, Self) -> Bool
pub fn CaseResult::executable(String, Bool, message? : String) -> Self
pub fn CaseResult::id(Self) -> String
pub fn CaseResult::message(Self) -> String
pub fn CaseResult::new(String, CaseDisposition, Bool, message? : String) -> Self
pub fn CaseResult::not_equal(Self, Self) -> Bool
pub fn CaseResult::passed(Self) -> Bool
pub fn CaseResult::to_repr(Self) -> @debug.Repr

pub struct RunSummary {
  // private fields
} derive(Eq, @debug.Debug)
pub fn RunSummary::diagnostic_cases(Self) -> Int
pub fn RunSummary::equal(Self, Self) -> Bool
pub fn RunSummary::executable_cases(Self) -> Int
pub fn RunSummary::failed_cases(Self) -> Int
pub fn RunSummary::from_results(Int, Array[CaseResult]) -> Self
pub fn RunSummary::legacy_cases(Self) -> Int
pub fn RunSummary::merge(Array[Self]) -> Self
pub fn RunSummary::not_equal(Self, Self) -> Bool
pub fn RunSummary::passed_cases(Self) -> Int
pub fn RunSummary::results(Self) -> Array[CaseResult]
pub fn RunSummary::selected_cases(Self) -> Int
pub fn RunSummary::skipped_cases(Self) -> Int
pub fn RunSummary::success(Self) -> Bool
pub fn RunSummary::to_repr(Self) -> @debug.Repr
pub fn RunSummary::total_cases(Self) -> Int
pub fn RunSummary::unsupported_cases(Self) -> Int

pub struct ShardSpec {
  // private fields
} derive(Eq, @debug.Debug)
pub fn ShardSpec::count(Self) -> Int
pub fn ShardSpec::equal(Self, Self) -> Bool
pub fn ShardSpec::index(Self) -> Int
pub fn ShardSpec::new(Int, Int) -> Self
pub fn ShardSpec::not_equal(Self, Self) -> Bool
pub fn ShardSpec::selects(Self, Int) -> Bool
pub fn ShardSpec::to_repr(Self) -> @debug.Repr
pub fn ShardSpec::try_new(Int, Int) -> Result[Self, String]

pub struct SourceLocation {
  // private fields
} derive(Eq, @debug.Debug)
pub fn SourceLocation::column(Self) -> Int
pub fn SourceLocation::equal(Self, Self) -> Bool
pub fn SourceLocation::line(Self) -> Int
pub fn SourceLocation::new(String, Int, column? : Int) -> Self
pub fn SourceLocation::not_equal(Self, Self) -> Bool
pub fn SourceLocation::source(Self) -> String
pub fn SourceLocation::to_repr(Self) -> @debug.Repr

// Type aliases

// Traits

internal/conformance API

Source locations

SourceLocation

SourceLocation::new

SourceLocation::source, SourceLocation::line, SourceLocation::column

Case results

CaseDisposition

CaseResult

CaseResult::new, CaseResult::executable

CaseResult::id, disposition, passed, message

Sharding

ShardSpec

ShardSpec::try_new, ShardSpec::new

ShardSpec::count, ShardSpec::index

ShardSpec::selects