1
1
mirror of https://github.com/github/semantic.git synced 2024-12-26 08:25:19 +03:00

Merge remote-tracking branch 'origin/master' into build-semantic-core-in-travis

This commit is contained in:
Patrick Thomson 2019-06-04 15:28:40 -04:00
commit 43b65ee190
12 changed files with 538 additions and 60 deletions

2
.gitignore vendored
View File

@ -3,6 +3,8 @@
.docsets .docsets
.stack-work .stack-work
.stack-work-profiling .stack-work-profiling
stack.yaml
stack.yaml.lock
profiles profiles
/tags /tags

View File

@ -108,7 +108,7 @@ cabal new-test
cabal new-run semantic -- --help cabal new-run semantic -- --help
``` ```
`semantic` requires GHC 8.6.4. We recommend using [`ghcup`][ghcup] to sandbox GHC versions. Our version bounds are based on [Stackage][stackage] LTS versions. The current LTS version is 13.13; `stack` build should also work if you prefer. `semantic` requires GHC 8.6.4. We recommend using [`ghcup`][ghcup] to sandbox GHC versions. Our version bounds are based on [Stackage][stackage] LTS versions. The current LTS version is 13.13; `stack` build should also work if you prefer, there is an unofficial [`stack.yaml`](https://gist.github.com/jkachmar/f200caee83280f1f25e9cfa2dd2b16bb).
[nix]: https://www.haskell.org/cabal/users-guide/nix-local-build-overview.html [nix]: https://www.haskell.org/cabal/users-guide/nix-local-build-overview.html
[stackage]: https://stackage.org [stackage]: https://stackage.org

30
docs/core-grammar.md Normal file
View File

@ -0,0 +1,30 @@
# Semantic Core grammar
This is an EBNF grammar for the (experimental) core IR language.
```
expr ::= expr '.' expr
| expr ' '+ expr
| '{' expr (';' expr)* ';'? '}'
| 'if' expr 'then' expr 'else' expr
| ('lexical' | 'import' | 'load') expr
| lit
| 'let'? lvalue '=' expr
| '(' expr ')'
lvalue ::= ident
| parens expr
lit ::= '#true'
| '#false'
| 'unit'
| 'frame'
| lambda
| ident
lambda ::= ('λ' | '\') ident ('->' | '→') expr
ident ::= [A-z_] ([A-z0-9_])*
| '#{' [^{}]+ '}'
| '"' [^"]+ '"'
```

View File

@ -26,22 +26,29 @@ library
, Analysis.Typecheck , Analysis.Typecheck
, Control.Effect.Readline , Control.Effect.Readline
, Data.Core , Data.Core
, Data.Core.Parser
, Data.Core.Pretty
, Data.File , Data.File
, Data.Loc , Data.Loc
, Data.Name , Data.Name
, Data.Stack , Data.Stack
-- other-modules: -- other-modules:
-- other-extensions: -- other-extensions:
build-depends: algebraic-graphs ^>= 0.3 build-depends: algebraic-graphs ^>= 0.3
, base >= 4.11 && < 5 , base >= 4.11 && < 5
, containers ^>= 0.6 , containers ^>= 0.6
, directory ^>= 1.3 , directory ^>= 1.3
, filepath ^>= 1.4 , filepath ^>= 1.4
, fused-effects ^>= 0.4 , fused-effects ^>= 0.4
, haskeline ^>= 0.7.5 , haskeline ^>= 0.7.5
, prettyprinter ^>= 1.2.1 , parsers ^>= 0.12.10
, semigroupoids ^>= 5.3 , prettyprinter ^>= 1.2.1
, transformers ^>= 0.5.6 , prettyprinter-ansi-terminal ^>= 1.1.1
, recursion-schemes ^>= 5.1
, semigroupoids ^>= 5.3
, transformers ^>= 0.5.6
, trifecta ^>= 2
, unordered-containers ^>= 0.2.10
hs-source-dirs: src hs-source-dirs: src
default-language: Haskell2010 default-language: Haskell2010
ghc-options: -Weverything -Wno-missing-local-signatures -Wno-missing-import-lists -Wno-implicit-prelude -Wno-safe -Wno-unsafe -Wno-name-shadowing -Wno-monomorphism-restriction -Wno-missed-specialisations -Wno-all-missed-specialisations ghc-options: -Weverything -Wno-missing-local-signatures -Wno-missing-import-lists -Wno-implicit-prelude -Wno-safe -Wno-unsafe -Wno-name-shadowing -Wno-monomorphism-restriction -Wno-missed-specialisations -Wno-all-missed-specialisations
@ -57,3 +64,17 @@ test-suite doctest
, semantic-core , semantic-core
hs-source-dirs: test hs-source-dirs: test
default-language: Haskell2010 default-language: Haskell2010
test-suite spec
type: exitcode-stdio-1.0
main-is: Spec.hs
other-modules: Generators
build-depends: base
, semantic-core
, hedgehog >= 0.6 && <1
, tasty >= 1.2 && <2
, tasty-hedgehog >= 0.2 && <1
, tasty-hunit >= 0.10 && <1
, trifecta
hs-source-dirs: test
default-language: Haskell2010

View File

@ -119,7 +119,7 @@ prog6 =
] ]
] ]
, File (Loc "main" (locSpan (fromJust here))) $ block , File (Loc "main" (locSpan (fromJust here))) $ block
[ Load (String "dep") [ Load (Var (Path "dep"))
, Let (User "thing") := Var (Path "dep") :. Var (User "var") , Let (User "thing") := Var (Path "dep") :. Var (User "var")
] ]
] ]

View File

@ -1,8 +1,8 @@
{-# LANGUAGE DeriveTraversable, FlexibleContexts, LambdaCase, OverloadedStrings, RecordWildCards #-} {-# LANGUAGE DeriveTraversable, FlexibleContexts, LambdaCase, OverloadedStrings, RecordWildCards, TemplateHaskell, TypeFamilies #-}
module Data.Core module Data.Core
( Core(..) ( Core(..)
, CoreF(..)
, Edge(..) , Edge(..)
, showCore
, lams , lams
, ($$*) , ($$*)
, unapply , unapply
@ -10,18 +10,21 @@ module Data.Core
, block , block
, ann , ann
, annWith , annWith
, stripAnnotations
) where ) where
import Control.Applicative (Alternative (..)) import Control.Applicative (Alternative (..))
import Data.Functor.Foldable hiding (ListF(..))
import Data.Functor.Foldable.TH
import Data.Foldable (foldl') import Data.Foldable (foldl')
import Data.Loc import Data.Loc
import Data.Name import Data.Name
import Data.Stack import Data.Stack
import Data.Text.Prettyprint.Doc (Pretty (..), (<+>), vsep)
import qualified Data.Text.Prettyprint.Doc as Pretty
import qualified Data.Text.Prettyprint.Doc.Render.String as Pretty
import GHC.Stack import GHC.Stack
data Edge = Lexical | Import
deriving (Eq, Ord, Show)
data Core data Core
= Var Name = Var Name
| Let Name | Let Name
@ -50,52 +53,16 @@ infixr 1 :>>
infix 3 := infix 3 :=
infixl 4 :. infixl 4 :.
data Edge = Lexical | Import makeBaseFunctor ''Core
deriving (Eq, Ord, Show)
instance Pretty Edge where infixl 2 :$$
pretty = pretty . show infixr 1 :>>$
infix 3 :=$
infixl 4 :.$
instance Semigroup Core where instance Semigroup Core where
(<>) = (:>>) (<>) = (:>>)
softsemi :: Pretty.Doc a
softsemi = Pretty.flatAlt mempty ";"
showCore :: Core -> String
showCore = Pretty.renderString . Pretty.layoutPretty Pretty.defaultLayoutOptions . pretty
instance Pretty Core where
pretty = \case
Var a -> pretty a
Let a -> "let" <+> pretty a
a :>> b -> vsep [pretty a <> softsemi, pretty b]
Lam x f -> vsep [ Pretty.nest 2 $ vsep [ "λ" <> pretty x <+> "-> {"
, pretty f
]
, "}"
]
f :$ x -> pretty f <> "." <> pretty x
Unit -> Pretty.parens mempty
Bool b -> pretty b
If c x y -> Pretty.sep [ "if" <+> pretty c
, "then" <+> pretty x
, "else" <+> pretty y
]
String s -> pretty (show s)
Frame -> Pretty.braces mempty
Load p -> "load" <+> pretty p
Edge e n -> pretty e <+> pretty n
a :. b -> "push" <+> Pretty.parens (pretty a) <+> Pretty.brackets (pretty b)
var := x -> pretty var <+> "=" <+> pretty x
Ann (Loc p s) c -> pretty c <> Pretty.brackets (pretty p <> ":" <> pretty s)
lams :: Foldable t => t Name -> Core -> Core lams :: Foldable t => t Name -> Core -> Core
lams names body = foldr Lam body names lams names body = foldr Lam body names
@ -124,3 +91,8 @@ ann = annWith callStack
annWith :: CallStack -> Core -> Core annWith :: CallStack -> Core -> Core
annWith callStack c = maybe c (flip Ann c) (stackLoc callStack) annWith callStack c = maybe c (flip Ann c) (stackLoc callStack)
stripAnnotations :: Core -> Core
stripAnnotations = cata go where
go (AnnF _ item) = item
go item = embed item

View File

@ -0,0 +1,115 @@
module Data.Core.Parser
( module Text.Trifecta
, core
, lit
, expr
, lvalue
) where
-- Consult @doc/grammar.md@ for an EBNF grammar.
import Control.Applicative
import qualified Data.Char as Char
import Data.Core
import Data.Name
import Data.Semigroup
import Data.String
import qualified Text.Parser.Token as Token
import qualified Text.Parser.Token.Highlight as Highlight
import Text.Trifecta hiding (ident)
-- * Identifier styles and derived parsers
validIdentifierStart :: Char -> Bool
validIdentifierStart c = not (Char.isDigit c) && isSimpleCharacter c
coreIdents :: TokenParsing m => IdentifierStyle m
coreIdents = Token.IdentifierStyle
{ _styleName = "core"
, _styleStart = satisfy validIdentifierStart
, _styleLetter = satisfy isSimpleCharacter
, _styleReserved = reservedNames
, _styleHighlight = Highlight.Identifier
, _styleReservedHighlight = Highlight.ReservedIdentifier
}
reserved :: (TokenParsing m, Monad m) => String -> m ()
reserved = Token.reserve coreIdents
identifier :: (TokenParsing m, Monad m, IsString s) => m s
identifier = choice [quote, plain] <?> "identifier" where
plain = Token.ident coreIdents
quote = between (string "#{") (symbol "}") (fromString <$> some (noneOf "{}"))
-- * Parsers (corresponding to EBNF)
core :: (TokenParsing m, Monad m) => m Core
core = expr
expr :: (TokenParsing m, Monad m) => m Core
expr = atom `chainl1` go where
go = choice [ (:.) <$ dot
, (:$) <$ notFollowedBy dot
]
atom :: (TokenParsing m, Monad m) => m Core
atom = choice
[ comp
, ifthenelse
, edge
, lit
, ident
, assign
, parens expr
]
comp :: (TokenParsing m, Monad m) => m Core
comp = braces (sconcat <$> sepEndByNonEmpty expr semi) <?> "compound statement"
ifthenelse :: (TokenParsing m, Monad m) => m Core
ifthenelse = If
<$ reserved "if" <*> core
<* reserved "then" <*> core
<* reserved "else" <*> core
<?> "if-then-else statement"
assign :: (TokenParsing m, Monad m) => m Core
assign = (:=) <$> try (lvalue <* symbolic '=') <*> core <?> "assignment"
edge :: (TokenParsing m, Monad m) => m Core
edge = kw <*> expr where kw = choice [ Edge Lexical <$ reserved "lexical"
, Edge Import <$ reserved "import"
, Load <$ reserved "load"
]
lvalue :: (TokenParsing m, Monad m) => m Core
lvalue = choice
[ Let <$ reserved "let" <*> name
, ident
, parens expr
]
-- * Literals
name :: (TokenParsing m, Monad m) => m Name
name = choice [regular, strpath] <?> "name" where
regular = User <$> identifier
strpath = Path <$> between (symbolic '"') (symbolic '"') (some $ noneOf "\"")
lit :: (TokenParsing m, Monad m) => m Core
lit = let x `given` n = x <$ reserved n in choice
[ Bool True `given` "#true"
, Bool False `given` "#false"
, Unit `given` "#unit"
, Frame `given` "#frame"
, lambda
] <?> "literal"
lambda :: (TokenParsing m, Monad m) => m Core
lambda = Lam <$ lambduh <*> name <* arrow <*> core <?> "lambda" where
lambduh = symbolic 'λ' <|> symbolic '\\'
arrow = symbol "" <|> symbol "->"
ident :: (Monad m, TokenParsing m) => m Core
ident = Var <$> name <?> "identifier"

View File

@ -0,0 +1,130 @@
{-# LANGUAGE FlexibleContexts, LambdaCase, OverloadedStrings, TypeApplications #-}
module Data.Core.Pretty
( showCore
, printCore
, showFile
, printFile
, prettyCore
) where
import Control.Effect
import Control.Effect.Reader
import Data.Core
import Data.File
import Data.Functor.Foldable
import Data.Name
import Data.Text.Prettyprint.Doc (Pretty (..), annotate, softline, (<+>))
import qualified Data.Text.Prettyprint.Doc as Pretty
import qualified Data.Text.Prettyprint.Doc.Render.String as Pretty
import qualified Data.Text.Prettyprint.Doc.Render.Terminal as Pretty
showCore :: Core -> String
showCore = Pretty.renderString . Pretty.layoutSmart Pretty.defaultLayoutOptions . Pretty.unAnnotate . prettyCore Ascii
printCore :: Core -> IO ()
printCore p = Pretty.putDoc (prettyCore Unicode p) *> putStrLn ""
showFile :: File Core -> String
showFile = showCore . fileBody
printFile :: File Core -> IO ()
printFile = printCore . fileBody
type AnsiDoc = Pretty.Doc Pretty.AnsiStyle
keyword, symbol, strlit, primitive :: AnsiDoc -> AnsiDoc
keyword = annotate (Pretty.colorDull Pretty.Cyan)
symbol = annotate (Pretty.color Pretty.Yellow)
strlit = annotate (Pretty.colorDull Pretty.Green)
primitive = keyword . mappend "#"
type Prec = Int
data Style = Unicode | Ascii
lambda, arrow :: (Member (Reader Style) sig, Carrier sig m) => m AnsiDoc
lambda = ask @Style >>= \case
Unicode -> pure $ symbol "λ"
Ascii -> pure $ symbol "\\"
arrow = ask @Style >>= \case
Unicode -> pure $ symbol ""
Ascii -> pure $ symbol "->"
name :: Name -> AnsiDoc
name = \case
Gen p -> pretty p
Path p -> strlit (Pretty.viaShow p)
User n -> encloseIf (needsQuotation n) (symbol "#{") (symbol "}") (pretty n)
with :: (Member (Reader Prec) sig, Carrier sig m) => Prec -> m a -> m a
with n = local (const n)
inParens :: (Member (Reader Prec) sig, Carrier sig m) => Prec -> m AnsiDoc -> m AnsiDoc
inParens amount go = do
prec <- ask
body <- with amount go
pure (encloseIf (amount >= prec) (symbol "(") (symbol ")") body)
encloseIf :: Monoid m => Bool -> m -> m -> m -> m
encloseIf True l r x = l <> x <> r
encloseIf False _ _ x = x
prettify :: (Member (Reader Prec) sig, Member (Reader Style) sig, Carrier sig m)
=> CoreF (m AnsiDoc)
-> m AnsiDoc
prettify = \case
VarF a -> pure $ name a
LetF a -> pure $ keyword "let" <+> name a
a :>>$ b -> do
prec <- ask @Prec
fore <- with 12 a
aft <- with 12 b
let open = symbol ("{" <> softline)
close = symbol (softline <> "}")
separator = ";" <> Pretty.line
body = fore <> separator <> aft
pure . Pretty.align $ encloseIf (12 > prec) open close (Pretty.align body)
LamF x f -> inParens 11 $ do
body <- f
lam <- lambda
arr <- arrow
pure (lam <> name x <+> arr <+> body)
FrameF -> pure $ primitive "frame"
UnitF -> pure $ primitive "unit"
BoolF b -> pure $ primitive (if b then "true" else "false")
StringF s -> pure . strlit $ Pretty.viaShow s
f :$$ x -> inParens 11 $ (<+>) <$> f <*> x
IfF con tru fal -> do
con' <- "if" `appending` con
tru' <- "then" `appending` tru
fal' <- "else" `appending` fal
pure $ Pretty.sep [con', tru', fal']
LoadF p -> "load" `appending` p
EdgeF Lexical n -> "lexical" `appending` n
EdgeF Import n -> "import" `appending` n
item :.$ body -> inParens 5 $ do
f <- item
g <- body
pure (f <> symbol "." <> g)
lhs :=$ rhs -> inParens 4 $ do
f <- lhs
g <- rhs
pure (f <+> symbol "=" <+> g)
-- Annotations are not pretty-printed, as it lowers the signal/noise ratio too profoundly.
AnnF _ c -> c
appending :: Functor f => AnsiDoc -> f AnsiDoc -> f AnsiDoc
appending k item = (keyword k <+>) <$> item
prettyCore :: Style -> Core -> AnsiDoc
prettyCore s = run . runReader @Prec 0 . runReader s . cata prettify

View File

@ -1,8 +1,11 @@
{-# LANGUAGE DeriveFunctor, ExistentialQuantification, FlexibleContexts, FlexibleInstances, GeneralizedNewtypeDeriving, LambdaCase, MultiParamTypeClasses, OverloadedStrings, StandaloneDeriving, TypeOperators, UndecidableInstances #-} {-# LANGUAGE DeriveFunctor, ExistentialQuantification, FlexibleContexts, FlexibleInstances, GeneralizedNewtypeDeriving, LambdaCase, MultiParamTypeClasses, OverloadedLists, OverloadedStrings,StandaloneDeriving, TypeOperators, UndecidableInstances #-}
module Data.Name module Data.Name
( User ( User
, Namespaced , Namespaced
, Name(..) , Name(..)
, reservedNames
, isSimpleCharacter
, needsQuotation
, Gensym(..) , Gensym(..)
, (//) , (//)
, gensym , gensym
@ -20,6 +23,9 @@ import Control.Effect.State
import Control.Effect.Sum import Control.Effect.Sum
import Control.Monad.Fail import Control.Monad.Fail
import Control.Monad.IO.Class import Control.Monad.IO.Class
import qualified Data.Char as Char
import Data.HashSet (HashSet)
import qualified Data.HashSet as HashSet
import Data.Text.Prettyprint.Doc (Pretty (..)) import Data.Text.Prettyprint.Doc (Pretty (..))
import qualified Data.Text.Prettyprint.Doc as Pretty import qualified Data.Text.Prettyprint.Doc as Pretty
@ -50,6 +56,25 @@ instance Pretty Name where
User n -> pretty n User n -> pretty n
Path p -> pretty (show p) Path p -> pretty (show p)
reservedNames :: HashSet User
reservedNames = [ "#true", "#false", "let", "#frame", "if", "then", "else"
, "lexical", "import", "#unit", "load"]
-- | Returns true if any character would require quotation or if the
-- name conflicts with a Core primitive.
needsQuotation :: User -> Bool
needsQuotation u = HashSet.member u reservedNames || any (not . isSimpleCharacter) u
-- | A simple character is, loosely defined, a character that is compatible
-- with identifiers in most ASCII-oriented programming languages. This is defined
-- as the alphanumeric set plus @$@ and @_@.
isSimpleCharacter :: Char -> Bool
isSimpleCharacter = \case
'$' -> True -- common in JS
'_' -> True
'?' -> True -- common in Ruby
c -> Char.isAlphaNum c
data Gensym data Gensym
= Root String = Root String
| Gensym :/ (String, Int) | Gensym :/ (String, Int)

View File

@ -0,0 +1,56 @@
{-# LANGUAGE ScopedTypeVariables #-}
module Generators
( literal
, name
, variable
, boolean
, lambda
, apply
, ifthenelse
) where
import Prelude hiding (span)
import Hedgehog hiding (Var)
import qualified Hedgehog.Gen as Gen
import qualified Hedgehog.Range as Range
import Data.Core
import Data.Name
-- The 'prune' call here ensures that we don't spend all our time just generating
-- fresh names for variables, since the length of variable names is not an
-- interesting property as they parse regardless.
name :: MonadGen m => m Name
name = Gen.prune (User <$> names) where
names = Gen.string (Range.linear 1 10) Gen.lower
boolean :: MonadGen m => m Core
boolean = Bool <$> Gen.bool
variable :: MonadGen m => m Core
variable = Var <$> name
ifthenelse :: MonadGen m => m Core -> m Core
ifthenelse bod = Gen.subterm3 boolean bod bod If
apply :: MonadGen m => m Core -> m Core
apply gen = go where
go = Gen.recursive
Gen.choice
[ Gen.subterm2 gen gen (:$)]
[ Gen.subterm2 go go (:$) -- balanced
, Gen.subtermM go (\x -> Lam <$> name <*> pure x)
]
lambda :: MonadGen m => m Core -> m Core
lambda bod = do
arg <- name
Gen.subterm bod (Lam arg)
atoms :: MonadGen m => [m Core]
atoms = [boolean, variable, pure Unit, pure Frame]
literal :: MonadGen m => m Core
literal = Gen.recursive Gen.choice atoms [lambda literal]

127
semantic-core/test/Spec.hs Normal file
View File

@ -0,0 +1,127 @@
{-# LANGUAGE OverloadedStrings, TypeApplications #-}
{-# OPTIONS_GHC -fno-warn-orphans #-}
module Main (main) where
import Data.String
import qualified Text.Trifecta as Trifecta
import Hedgehog hiding (Var)
import Test.Tasty
import Test.Tasty.Hedgehog
import Test.Tasty.HUnit
import Data.File
import qualified Generators as Gen
import qualified Analysis.Eval as Eval
import Data.Core
import Data.Core.Pretty
import Data.Core.Parser as Parse
import Data.Name
-- * Helpers
true, false :: Core
true = Bool True
false = Bool False
instance IsString Name where fromString = User
parseEither :: Trifecta.Parser a -> String -> Either String a
parseEither p = Trifecta.foldResult (Left . show . Trifecta._errDoc) Right . Trifecta.parseString (p <* Trifecta.eof) mempty
-- * Parser roundtripping properties. Note that parsing and prettyprinting is generally
-- not a roundtrip, because the parser inserts 'Ann' nodes itself.
prop_roundtrips :: Gen Core -> Property
prop_roundtrips gen = property $ do
input <- forAll gen
tripping input showCore (parseEither (Parse.core <* Trifecta.eof))
parserProps :: TestTree
parserProps = testGroup "Parsing: roundtripping"
[ testProperty "literals" $ prop_roundtrips Gen.literal
, testProperty "if/then/else" . prop_roundtrips . Gen.ifthenelse $ Gen.variable
, testProperty "lambda" . prop_roundtrips $ Gen.lambda Gen.literal
, testProperty "function application" . prop_roundtrips $ Gen.apply Gen.variable
]
-- * Parser specs
parsesInto :: String -> Core -> Assertion
parsesInto str res = case parseEither Parse.core str of
Right x -> x @?= res
Left m -> assertFailure m
assert_booleans_parse :: Assertion
assert_booleans_parse = do
parseEither Parse.core "#true" @?= Right true
parseEither Parse.core "#false" @?= Right false
a, f, g, h :: Core
(a, f, g, h) = (Var "a", Var "f", Var "g", Var "h")
assert_ifthen_parse :: Assertion
assert_ifthen_parse = "if #true then #true else #false" `parsesInto` (If true true false)
assert_application_parse :: Assertion
assert_application_parse ="f g" `parsesInto` (f :$ g)
assert_application_left_associative :: Assertion
assert_application_left_associative = "f g h" `parsesInto` (f :$ g :$ h)
assert_push_left_associative :: Assertion
assert_push_left_associative = "f.g.h" `parsesInto` (f :. g :. h)
assert_ascii_lambda_parse :: Assertion
assert_ascii_lambda_parse = "\\a -> a" `parsesInto` Lam "a" a
assert_unicode_lambda_parse :: Assertion
assert_unicode_lambda_parse = "λa → a" `parsesInto` Lam "a" a
assert_quoted_name_parse :: Assertion
assert_quoted_name_parse = "#{(NilClass)}" `parsesInto` Var (User "(NilClass)")
assert_let_dot_precedence :: Assertion
assert_let_dot_precedence = "let a = f.g.h" `parsesInto` (Let "a" := (f :. g :. h))
assert_let_in_push_precedence :: Assertion
assert_let_in_push_precedence = "f.let g = h" `parsesInto` (f :. (Let "g" := h))
parserSpecs :: TestTree
parserSpecs = testGroup "Parsing: simple specs"
[ testCase "true/false" assert_booleans_parse
, testCase "if/then/else" assert_ifthen_parse
, testCase "function application" assert_application_parse
, testCase "application is left-associative" assert_application_left_associative
, testCase "dotted push is left-associative" assert_push_left_associative
, testCase "lambda with ASCII syntax" assert_ascii_lambda_parse
, testCase "lambda with unicode syntax" assert_unicode_lambda_parse
, testCase "quoted names" assert_quoted_name_parse
, testCase "let + dot precedence" assert_let_dot_precedence
, testCase "let in push" assert_let_in_push_precedence
]
assert_roundtrips :: File Core -> Assertion
assert_roundtrips (File _ core) = parseEither Parse.core (showCore core) @?= Right (stripAnnotations core)
parserExamples :: TestTree
parserExamples = testGroup "Parsing: Eval.hs examples"
[ testCase "prog1" (assert_roundtrips Eval.prog1)
, testCase "prog2" (assert_roundtrips Eval.prog2)
, testCase "prog3" (assert_roundtrips Eval.prog3)
, testCase "prog4" (assert_roundtrips Eval.prog4)
, testCase "prog6.1" (assert_roundtrips (head Eval.prog6))
, testCase "prog6.2" (assert_roundtrips (last Eval.prog6))
, testCase "ruby" (assert_roundtrips Eval.ruby)
]
tests :: TestTree
tests = testGroup "semantic-core"
[ parserSpecs
, parserExamples
, parserProps
]
main :: IO ()
main = defaultMain tests

View File

@ -166,7 +166,7 @@ graphCommand = command "graph" (info graphArgumentsParser (progDesc "Compute a g
shaReader :: ReadM Git.OID shaReader :: ReadM Git.OID
shaReader = eitherReader parseSha shaReader = eitherReader parseSha
where parseSha arg = if length arg == 40 where parseSha arg = if length arg == 40 || arg == "HEAD"
then Right (Git.OID (T.pack arg)) then Right (Git.OID (T.pack arg))
else Left (arg <> " is not a valid sha1") else Left (arg <> " is not a valid sha1")