Typos, TODOs, etc.

harpocrates · harpocrates · commit f265c97df2f3 · 2017-10-13T00:58:36.000-07:00
Nothing semantically significant.
diff --git a/README.md b/README.md
@@ -1,13 +1,35 @@
-# Parser and pretty printer for Rust [![Build Status][4]][5] [![Windows build status][7]][8]
+# Parser and pretty-printer for Rust [![Build Status][4]][5] [![Windows build status][7]][8]
 
 `language-rust` aspires to efficiently and accurately parse and pretty-print the [Rust language][0].
 The underlying AST structures are also intended to be as similar as possible to the AST `rustc` uses
 itself. When `language-rust` and `rustc` have diverging AST, the divergence should be detailed in
 the documentation.
 
+A typical use looks like:
+
+```haskell
+>>> :set -XTypeApplications +t
+>>> import Language.Rust.Syntax
+>>>
+>>> -- Sample use of the parser
+>>> import Language.Rust.Parser
+>>> let inp = inputStreamFromString "fn main () { println!(\"Hello world!\"); }"
+inp :: InputStream
+>>> Right sourceFile = parse @(SourceFile Span) inp
+sourceFile :: SourceFile Span
+>>>
+>>> -- Sample use of the pretty-printer
+>>> import Language.Rust.Pretty
+>>> pretty sourceFile
+fn main() {
+  println!("Hello world!");
+}
+it :: Doc b
+```
+
 ## Building
 
-## Cabal
+### Cabal
 
 With Cabal and GHC, you should only need to run
 
@@ -16,7 +38,7 @@ With Cabal and GHC, you should only need to run
     cabal configure
     cabal build
 
-## Stack
+### Stack
 
 With the [Stack][1] tool installed, you should only need to run
 
@@ -33,6 +55,8 @@ from nightly as they come out, but in general will only target compatibility wit
 
 ## Bugs
 
+Please report any bugs to the [github issue tracker][9].
+
 ### Parser
 
 Any difference between what is accepted by the `rustc` parser and the `language-rust` parser
@@ -60,3 +84,4 @@ pretty-printing.
 [6]: https://github.com/rust-lang-nursery/fmt-rfcs
 [7]: https://ci.appveyor.com/api/projects/status/um8dxklqmubvn091/branch/master?svg=true
 [8]: https://ci.appveyor.com/project/harpocrates/language-rust/branch/master
+[9]: https://github.com/harpocrates/language-rust/issues
diff --git a/Setup.hs b/Setup.hs
@@ -0,0 +1,2 @@
+import Distribution.Simple
+main = defaultMain
diff --git a/language-rust.cabal b/language-rust.cabal
@@ -15,7 +15,7 @@ stability:           alpha
 bug-reports:         https://github.com/harpocrates/language-rust/issues
 category:            Language
 build-type:          Simple
-extra-source-files:  ChangeLog.md
+extra-source-files:  ChangeLog.md README.md
 cabal-version:       >=1.10
 
 source-repository head
@@ -28,7 +28,7 @@ flag useByteStrings
 
 flag enableQuasiquotes
     description:     Provide the experimental 'Language.Rust.Quote' module
-    default:         True
+    default:         False
 
 library
   hs-source-dirs:      src
diff --git a/src/Language/Rust/Data/InputStream.hs b/src/Language/Rust/Data/InputStream.hs
@@ -41,8 +41,6 @@ import qualified Data.ByteString.UTF8 as BE
 import qualified Data.Char as Char
 #endif
 
--- TODO: backpack this when GHC 8.2 is released
-
 -- | Read a file into an 'InputStream'
 readInputStream :: FilePath -> IO InputStream
 
diff --git a/src/Language/Rust/Parser/Lexer.x b/src/Language/Rust/Parser/Lexer.x
@@ -18,7 +18,7 @@ bitwise and, and unary reference), @&&&x&&&y@ lexes into 'AmpersandAmpersand', '
 @'IdentTok' "x"@, 'AmpersandAmpersand', 'Ampersand', @'IdentTok' "y"@. Although the parser sometimes
 needs to "break apart" tokens, it never has to think about putting them together. That means it can
 easily figure out that @&&&x&&&y@ parses as @&(&(&x)) && (&y)@ and not @&(&(&x)) & (&(&y))@ even if
-bitwise conjunctions binds more tightly that logical conjunctions. 
+bitwise conjunctions bind more tightly that logical conjunctions. 
 
 This sort of amguity where one tokens need to be broken up occurs for
 
diff --git a/src/Language/Rust/Parser/ParseMonad.hs b/src/Language/Rust/Parser/ParseMonad.hs
@@ -8,7 +8,7 @@ Stability   : experimental
 Portability : portable
 
 Both the lexer and the parser run inside of the 'P' monad. As detailed in the section on 
-[threaded-lexers](https://www.haskell.org/happy/doc/html/sec-monads.html#sec-lexers) in Happy's
+on [threaded-lexers](https://www.haskell.org/happy/doc/html/sec-monads.html#sec-lexers) in Happy's
 instruction manual, the benefits of this are that:
 
   * Lexical errors can be treated in the same way as parse errors
@@ -101,7 +101,7 @@ getPState = P $ \ !s pOk _ -> pOk s s
 
 -- | Update the state stored in the parser.
 setPState :: PState -> P ()
-setPState s = P $ \_ pOk _ -> pOk () s 
+setPState s = P $ \ _ pOk _ -> pOk () s 
 
 -- | Modify the state stored in the parser.
 modifyPState :: (PState -> PState) -> P ()
@@ -113,7 +113,7 @@ getPosition = curPos <$> getPState
 
 -- | Update the current position of the parser.
 setPosition :: Position -> P ()
-setPosition pos = modifyPState $ \s -> s{ curPos = pos }
+setPosition pos = modifyPState $ \ s -> s{ curPos = pos }
 
 -- | Retrieve the current 'InputStream' of the parser
 getInput :: P InputStream
diff --git a/src/Language/Rust/Pretty/Internal.hs b/src/Language/Rust/Pretty/Internal.hs
@@ -101,12 +101,12 @@ import Language.Rust.Syntax.AST
 import Language.Rust.Syntax.Token
 import Language.Rust.Syntax.Ident
 
-import Data.Text.Prettyprint.Doc hiding ((<+>), hsep, indent, vsep)
+import Data.Text.Prettyprint.Doc hiding ( (<+>), hsep, indent, vsep )
 
-import Data.Maybe (maybeToList, fromMaybe)
-import Data.Foldable (toList)
-import Data.List (unfoldr)
-import Data.List.NonEmpty (NonEmpty(..))
+import Data.Maybe               ( maybeToList, fromMaybe )
+import Data.Foldable            ( toList )
+import Data.List                ( unfoldr )
+import Data.List.NonEmpty       ( NonEmpty(..) )
 import qualified Data.List.NonEmpty as N
 
 -- | Print a source file
diff --git a/src/Language/Rust/Pretty/Resolve.hs b/src/Language/Rust/Pretty/Resolve.hs
@@ -97,8 +97,8 @@ data Severity
 
 -- | Localized information about an issue in a syntax tree
 data Issue = Issue
-  { description :: String  -- ^ Description of the issue
-  , severity :: Severity   -- ^ Severity of the issue
+  { description :: String    -- ^ Description of the issue
+  , severity :: !Severity    -- ^ Severity of the issue
   -- | The first element in this list is the syntax tree where the issue occurs. The next elements
   -- are increasingly zoomed out syntax trees centered on the first element. In lieu of positional
   -- information, this provides a next best way of debugging exactly where the problem is.
diff --git a/src/Language/Rust/QuasiQuote.hs b/src/Language/Rust/QuasiQuote.hs
diff --git a/src/Language/Rust/QuasiQuoteParser.y b/src/Language/Rust/QuasiQuoteParser.y
diff --git a/src/Language/Rust/Quote.hs b/src/Language/Rust/Quote.hs
@@ -27,7 +27,7 @@ i32
 >>> pretty body
 { (1i32) + x }
 
-For now, however, you cannot use '$x' or '$x:ty' meta variables.
+For now, however, you cannot use @$x@ or @$x:ty@ meta variables.
 -}
 module Language.Rust.Quote (
   lit, attr, ty, pat, stmt, expr, item, sourceFile, implItem, traitItem, tokenTree, block
diff --git a/src/Language/Rust/Syntax/AST.hs b/src/Language/Rust/Syntax/AST.hs
@@ -425,7 +425,7 @@ instance Located a => Located (Expr a) where
 
 -- | Field in a struct literal expression (@syntax::ast::Field@).
 --
--- Example: @x: 1@ in @Point{ x: 1, y: 2 }@
+-- Example: @x: 1@ in @Point { x: 1, y: 2 }@
 data Field a = Field Ident (Maybe (Expr a)) a
   deriving (Eq, Ord, Functor, Show, Typeable, Data, Generic, Generic1, NFData)
 
@@ -434,7 +434,7 @@ instance Located a => Located (Field a) where spanOf (Field _ _ s) = spanOf s
 -- | Field in a struct literal pattern (@syntax::ast::FieldPat@). The field name 'Ident' is optional
 -- but, when it is 'Nothing', the pattern the field is destructured to must be 'IdentP'.
 --
--- Example: @x@ in @Point{ x, y }@
+-- Example: @x@ in @Point { x, y }@
 data FieldPat a = FieldPat (Maybe Ident) (Pat a) a
   deriving (Eq, Ord, Functor, Show, Typeable, Data, Generic, Generic1, NFData)
 
@@ -609,7 +609,7 @@ instance Located a => Located (Lifetime a) where spanOf (Lifetime _ s) = spanOf
 -- | A lifetime definition, introducing a lifetime and the other lifetimes that bound it
 -- (@syntax::ast::LifetimeDef@).
 --
--- Example: @\'a: \'b+\'c+\'d@
+-- Example: @\'a: \'b + \'c + \'d@
 data LifetimeDef a = LifetimeDef [Attribute a] (Lifetime a) [Lifetime a] a
   deriving (Eq, Ord, Functor, Show, Typeable, Data, Generic, Generic1, NFData)
 
@@ -1069,7 +1069,7 @@ instance Located a => Located (Variant a) where spanOf (Variant _ _ _ _ s) = spa
 -- | Main payload in a 'Variant' (@syntax::ast::VariantData@).
 data VariantData a
   = StructD [StructField a] a -- ^ struct variant (example: @Bar { .. }@ as in @enum Foo { Bar { .. } }@)
-  | TupleD [StructField a] a  -- ^ tuple variant (exmaple: @Bar(..)@ as in enum Foo { Bar(..) }@)
+  | TupleD [StructField a] a  -- ^ tuple variant (exmaple: @Bar(..)@ as in enum @Foo { Bar(..) }@)
   | UnitD a                   -- ^ unit variant (example @Bar@ as in @enum Foo { Bar = .. }@)
   deriving (Eq, Ord, Functor, Show, Typeable, Data, Generic, Generic1, NFData)
 
diff --git a/src/Language/Rust/Syntax/Ident.hs b/src/Language/Rust/Syntax/Ident.hs
@@ -11,7 +11,7 @@ Data structure behind identifiers.
 -}
 {-# LANGUAGE DeriveDataTypeable, DeriveGeneric, DeriveAnyClass #-}
 
-module Language.Rust.Syntax.Ident (Ident(..), mkIdent, invalidIdent, Name) where
+module Language.Rust.Syntax.Ident (Ident(..), mkIdent, Name) where
 
 import GHC.Generics (Generic)
 import Data.Data (Data)
@@ -25,8 +25,8 @@ import Data.Semigroup
 
 -- | An identifier
 data Ident
-  = Ident { name :: Name  -- ^ payload of the identifier
-          , hash :: !Int  -- ^ hash for quick comparision
+  = Ident { name :: Name                 -- ^ payload of the identifier
+          , hash :: {-# UNPACK #-} !Int  -- ^ hash for quick comparision
           } deriving (Data, Typeable, Generic, NFData)
 
 instance Show Ident where
@@ -46,10 +46,12 @@ instance Ord Ident where
                     rt -> rt
 
 instance Monoid Ident where
-  Ident n1 _ `mappend` Ident n2 _ = mkIdent (n1 `mappend` n2)
-  mempty = invalidIdent
+  mappend = (<>)
+  mempty = mkIdent ""
+
+instance Semigroup Ident where
+  Ident n1 _ <> Ident n2 _ = mkIdent (n1 <> n2)
 
-instance Semigroup Ident
 
 -- | Smart constructor for making an 'Ident'.
 mkIdent :: String -> Ident
@@ -62,10 +64,6 @@ hashString = foldl' f golden
          magic = 0xdeadbeef
          golden = 1013904242
 
--- | The empty identifier is invalid
-invalidIdent :: Ident
-invalidIdent = mkIdent ""
-
--- | TODO: backpack
+-- | The payload of an identifier
 type Name = String
 
diff --git a/test/README.md b/test/README.md
@@ -26,9 +26,9 @@ Whenever a bug is fixed or a feature introduced, new tests covering that case sh
 ## `rustc-tests`
 
 These require `rustup` to be installed. The idea is to feed in as input many rust source files, run
-them through the Rust compiler with `rustc -Z ast-json` (now only available on nightly) to get a
-JSON of the AST, then parse the same file on our side, and finally find the differences between the
-two. Specifically, these tests check that:
+them through the Rust compiler with `rustc -Z ast-json -Z parse-only` (now only available on
+nightly) to get a JSON of the AST, then parse the same file on our side, and finally find the
+differences between the two. Specifically, these tests check that:
 
   * we parse the same thing as `rustc`
   * pretty-printing the AST and then reparsing it doesn't change anything
diff --git a/test/rustc-tests/Diff.hs b/test/rustc-tests/Diff.hs
@@ -635,6 +635,10 @@ instance Diffable Ident where
   Ident i _ === String s | fromString i == s = pure ()
   ident'    === val = diff "identifiers are different" ident' val
 
+-- | The empty identifier is invalid
+invalidIdent :: Ident
+invalidIdent = mkIdent ""
+
 instance Diffable BindingMode where
   bm === val =
     case (val ! "variant", bm) of

Original file line number	Diff line number	Diff line change
`@@ -0,0 +1,2 @@`
	`1`	`+import Distribution.Simple`
	`2`	`+main = defaultMain`