Util.hs

{-# LANGUAGE CPP #-}
-- Copyright (c) 2019 The DAML Authors. All rights reserved.
-- SPDX-License-Identifier: Apache-2.0

-- | General utility functions, mostly focused around GHC operations.
module Development.IDE.GHC.Util(
    modifyDynFlags,
    evalGhcEnv,
    -- * GHC wrappers
    printRdrName,
    Development.IDE.GHC.Util.printName,
    ParseResult(..), runParser,
    lookupPackageConfig,
    textToStringBuffer,
    bytestringToStringBuffer,
    stringBufferToByteString,
    moduleImportPath,
    cgGutsToCoreModule,
    fingerprintToBS,
    fingerprintFromByteString,
    fingerprintFromStringBuffer,
    fingerprintFromPut,
    -- * General utilities
    readFileUtf8,
    hDuplicateTo',
    setHieDir,
    dontWriteHieFiles,
    disableWarningsAsErrors,
    printOutputable,
    getExtensions,
    textInRange
    ) where

import           Control.Concurrent
import           Control.Exception                 as E
import           Data.Binary.Put                   (Put, runPut)
import qualified Data.ByteString                   as BS
import           Data.ByteString.Internal          (ByteString (..))
import qualified Data.ByteString.Internal          as BS
import qualified Data.ByteString.Lazy              as LBS
import           Data.IORef
import           Data.List.Extra
import           Data.Maybe
import qualified Data.Text                         as T
import qualified Data.Text.Encoding                as T
import qualified Data.Text.Encoding.Error          as T
import           Data.Typeable
import           Development.IDE.GHC.Compat        as GHC hiding (unitState)
import qualified Development.IDE.GHC.Compat.Parser as Compat
import qualified Development.IDE.GHC.Compat.Units  as Compat
import           Development.IDE.Types.Location
import           Foreign.ForeignPtr
import           Foreign.Ptr
import           Foreign.Storable
import           GHC                               hiding (ParsedModule (..),
                                                    parser)
import           GHC.IO.BufferedIO                 (BufferedIO)
import           GHC.IO.Device                     as IODevice
import           GHC.IO.Encoding
import           GHC.IO.Exception
import           GHC.IO.Handle.Internals
import           GHC.IO.Handle.Types
import           Ide.PluginUtils                   (unescape)
import           System.FilePath

import           GHC.Data.EnumSet
import           GHC.Data.FastString
import           GHC.Data.StringBuffer
import           GHC.Utils.Fingerprint
----------------------------------------------------------------------
-- GHC setup

-- | Used to modify dyn flags in preference to calling 'setSessionDynFlags',
--   since that function also reloads packages (which is very slow).
modifyDynFlags :: GhcMonad m => (DynFlags -> DynFlags) -> m ()
modifyDynFlags f = do
  newFlags <- f <$> getSessionDynFlags
  -- We do not use setSessionDynFlags here since we handle package
  -- initialization separately.
  modifySession $ \h ->
    hscSetFlags newFlags h { hsc_IC = (hsc_IC h) {ic_dflags = newFlags} }

-- | Given a 'Unit' try and find the associated 'PackageConfig' in the environment.
lookupPackageConfig :: Unit -> HscEnv -> Maybe GHC.UnitInfo
lookupPackageConfig unit env =
    Compat.lookupUnit' False unitState prClsre unit
    where
        unitState = Compat.getUnitInfoMap env
        prClsre = preloadClosureUs env


-- | Convert from the @text@ package to the @GHC@ 'StringBuffer'.
--   Currently implemented somewhat inefficiently (if it ever comes up in a profile).
textToStringBuffer :: T.Text -> StringBuffer
textToStringBuffer = stringToStringBuffer . T.unpack

runParser :: DynFlags -> String -> P a -> ParseResult a
runParser flags str parser = unP parser parseState
    where
      filename = "<interactive>"
      location = mkRealSrcLoc (mkFastString filename) 1 1
      buffer = stringToStringBuffer str
      parseState = Compat.initParserState (Compat.initParserOpts flags) buffer location

stringBufferToByteString :: StringBuffer -> ByteString
stringBufferToByteString StringBuffer{..} = PS buf cur len

bytestringToStringBuffer :: ByteString -> StringBuffer
bytestringToStringBuffer (PS buf cur len) = StringBuffer{..}

-- | Pretty print a 'RdrName' wrapping operators in parens
printRdrName :: RdrName -> String
printRdrName name = T.unpack $ printOutputable $ parenSymOcc rn (ppr rn)
  where
    rn = rdrNameOcc name

-- | Pretty print a 'Name' wrapping operators in parens
printName :: Name -> String
printName = printRdrName . nameRdrName

-- | Run a 'Ghc' monad value using an existing 'HscEnv'. Sets up and tears down all the required
--   pieces, but designed to be more efficient than a standard 'runGhc'.
evalGhcEnv :: HscEnv -> Ghc b -> IO b
evalGhcEnv env act = snd <$> runGhcEnv env act

-- | Run a 'Ghc' monad value using an existing 'HscEnv'. Sets up and tears down all the required
--   pieces, but designed to be more efficient than a standard 'runGhc'.
runGhcEnv :: HscEnv -> Ghc a -> IO (HscEnv, a)
runGhcEnv env act = do
    hsc_env <- initTempFs env
    ref <- newIORef hsc_env
    res <- unGhc (withCleanupSession act) (Session ref)
    (,res) <$> readIORef ref

-- | Given a module location, and its parse tree, figure out what is the include directory implied by it.
--   For example, given the file @\/usr\/\Test\/Foo\/Bar.hs@ with the module name @Foo.Bar@ the directory
--   @\/usr\/Test@ should be on the include path to find sibling modules.
moduleImportPath :: NormalizedFilePath -> GHC.ModuleName -> Maybe FilePath
-- The call to takeDirectory is required since DAML does not require that
-- the file name matches the module name in the last component.
-- Once that has changed we can get rid of this.
moduleImportPath (takeDirectory . fromNormalizedFilePath -> pathDir) mn
    -- This happens for single-component modules since takeDirectory "A" == "."
    | modDir == "." = Just pathDir
    | otherwise = dropTrailingPathSeparator <$> stripSuffix modDir pathDir
  where
    -- A for module A.B
    modDir =
        takeDirectory $
        fromNormalizedFilePath $ toNormalizedFilePath' $
        moduleNameSlashes mn

-- | Read a UTF8 file, with lenient decoding, so it will never raise a decoding error.
readFileUtf8 :: FilePath -> IO T.Text
readFileUtf8 f = T.decodeUtf8With T.lenientDecode <$> BS.readFile f

-- | Convert from a 'CgGuts' to a 'CoreModule'.
cgGutsToCoreModule :: SafeHaskellMode -> CgGuts -> ModDetails -> CoreModule
cgGutsToCoreModule safeMode guts modDetails = CoreModule
    (cg_module guts)
    (md_types modDetails)
    (cg_binds guts)
    safeMode

-- | Convert a 'Fingerprint' to a 'ByteString' by copying the byte across.
--   Will produce an 8 byte unreadable ByteString.
fingerprintToBS :: Fingerprint -> BS.ByteString
fingerprintToBS (Fingerprint a b) = BS.unsafeCreate 8 $ \ptr -> do
    let ptr' = castPtr ptr
    pokeElemOff ptr' 0 a
    pokeElemOff ptr' 1 b

-- | Take the 'Fingerprint' of a 'StringBuffer'.
fingerprintFromStringBuffer :: StringBuffer -> IO Fingerprint
fingerprintFromStringBuffer (StringBuffer buf len cur) =
    withForeignPtr buf $ \ptr -> fingerprintData (ptr `plusPtr` cur) len

fingerprintFromByteString :: ByteString -> IO Fingerprint
fingerprintFromByteString bs = do
    let (fptr, offset, len) = BS.toForeignPtr bs
    withForeignPtr fptr $ \ptr ->
        fingerprintData (ptr `plusPtr` offset) len

fingerprintFromPut :: Put -> IO Fingerprint
fingerprintFromPut = fingerprintFromByteString . LBS.toStrict . runPut

-- | A slightly modified version of 'hDuplicateTo' from GHC.
--   Importantly, it avoids the bug listed in https://gitlab.haskell.org/ghc/ghc/merge_requests/2318.
hDuplicateTo' :: Handle -> Handle -> IO ()
hDuplicateTo' h1@(FileHandle path m1) h2@(FileHandle _ m2)  = do
 withHandle__' "hDuplicateTo" h2 m2 $ \h2_ -> do
   -- The implementation in base has this call to hClose_help.
   -- _ <- hClose_help h2_
   -- hClose_help does two things:
   -- 1. It flushes the buffer, we replicate this here
   _ <- flushWriteBuffer h2_ `E.catch` \(_ :: IOException) -> pure ()
   -- 2. It closes the handle. This is redundant since dup2 takes care of that
   -- but even worse it is actively harmful! Once the handle has been closed
   -- another thread is free to reallocate it. This leads to dup2 failing with EBUSY
   -- if it happens just in the right moment.
   withHandle_' "hDuplicateTo" h1 m1 $ \h1_ -> do
     dupHandleTo path h1 Nothing h2_ h1_ (Just handleFinalizer)
hDuplicateTo' h1@(DuplexHandle path r1 w1) h2@(DuplexHandle _ r2 w2)  = do
 withHandle__' "hDuplicateTo" h2 w2  $ \w2_ -> do
   _ <- hClose_help w2_
   withHandle_' "hDuplicateTo" h1 w1 $ \w1_ -> do
     dupHandleTo path h1 Nothing w2_ w1_ (Just handleFinalizer)
 withHandle__' "hDuplicateTo" h2 r2  $ \r2_ -> do
   _ <- hClose_help r2_
   withHandle_' "hDuplicateTo" h1 r1 $ \r1_ -> do
     dupHandleTo path h1 (Just w1) r2_ r1_ Nothing
hDuplicateTo' h1 _ =
  ioe_dupHandlesNotCompatible h1

-- | This is copied unmodified from GHC since it is not exposed.
dupHandleTo :: FilePath
            -> Handle
            -> Maybe (MVar Handle__)
            -> Handle__
            -> Handle__
            -> Maybe HandleFinalizer
            -> IO Handle__
dupHandleTo filepath h other_side
            _hto_@Handle__{haDevice=devTo}
            h_@Handle__{haDevice=dev} mb_finalizer = do
  flushBuffer h_
  case cast devTo of
    Nothing   -> ioe_dupHandlesNotCompatible h
    Just dev' -> do
      _ <- IODevice.dup2 dev dev'
      FileHandle _ m <- dupHandle_ dev' filepath other_side h_ mb_finalizer
      takeMVar m

-- | This is copied unmodified from GHC since it is not exposed.
-- Note the beautiful inline comment!
dupHandle_ :: (RawIO dev, IODevice dev, BufferedIO dev, Typeable dev) => dev
           -> FilePath
           -> Maybe (MVar Handle__)
           -> Handle__
           -> Maybe HandleFinalizer
           -> IO Handle
dupHandle_ new_dev filepath other_side Handle__{..} mb_finalizer = do
   -- XXX wrong!
  mb_codec <- if isJust haEncoder then fmap Just getLocaleEncoding else return Nothing
  mkHandle new_dev filepath haType True{-buffered-} mb_codec
      NewlineMode { inputNL = haInputNL, outputNL = haOutputNL }
      mb_finalizer other_side

-- | This is copied unmodified from GHC since it is not exposed.
ioe_dupHandlesNotCompatible :: Handle -> IO a
ioe_dupHandlesNotCompatible h =
   ioException (IOError (Just h) IllegalOperation "hDuplicateTo"
                "handles are incompatible" Nothing Nothing)

--------------------------------------------------------------------------------
-- Tracing exactprint terms

-- | Print a GHC value in `defaultUserStyle` without unique symbols.
-- It uses `showSDocUnsafe` with `unsafeGlobalDynFlags` internally.
--
-- This is the most common print utility.
-- It will do something additionally compared to what the 'Outputable' instance does.
--
--   1. print with a user-friendly style: `a_a4ME` as `a`.
--   2. unescape escape sequences of printable unicode characters within a pair of double quotes
printOutputable :: Outputable a => a -> T.Text
printOutputable =
    -- IfaceTyLit from GHC.Iface.Type implements Outputable with 'show'.
    -- Showing a String escapes non-ascii printable characters. We unescape it here.
    -- More discussion at https://github.com/haskell/haskell-language-server/issues/3115.
    unescape . T.pack . printWithoutUniques
{-# INLINE printOutputable #-}

getExtensions :: ParsedModule -> [Extension]
getExtensions = toList . extensionFlags . ms_hspp_opts . pm_mod_summary

-- | Returns [start .. end[
textInRange :: Range -> T.Text -> T.Text
textInRange (Range (Position (fromIntegral -> startRow) (fromIntegral -> startCol)) (Position (fromIntegral -> endRow) (fromIntegral -> endCol))) text =
    case compare startRow endRow of
      LT ->
        let (linesInRangeBeforeEndLine, endLineAndFurtherLines) = splitAt (endRow - startRow) linesBeginningWithStartLine
            (textInRangeInFirstLine, linesBetween) = case linesInRangeBeforeEndLine of
              [] -> ("", [])
              firstLine:linesInBetween -> (T.drop startCol firstLine, linesInBetween)
            maybeTextInRangeInEndLine = T.take endCol <$> listToMaybe endLineAndFurtherLines
        in T.intercalate "\n" (textInRangeInFirstLine : linesBetween ++ maybeToList maybeTextInRangeInEndLine)
      EQ ->
        let line = fromMaybe "" (listToMaybe linesBeginningWithStartLine)
        in T.take (endCol - startCol) (T.drop startCol line)
      GT -> ""
    where
      linesBeginningWithStartLine = drop startRow (T.splitOn "\n" text)