Updating static files never triggers recompilation

[snoyberg]

This seems to be a resurfacing of #94. Using the same code from that issue (slightly updated to include SessionInitParams):

{-# LANGUAGE OverloadedStrings #-}

module Main where

import IdeSession
import Data.Monoid (mconcat)
import qualified Data.ByteString.Lazy.Char8 as L8

main :: IO ()
main = do
    sess <- initSession (SessionInitParams Nothing) defaultSessionConfig
                { configDir             = "."
                }

    let cb = \_ -> return ()
        update = flip (updateSession sess) cb

    let mainContents = L8.pack $ unlines
            [ "{-# LANGUAGE TemplateHaskell #-}"
            , "import Text.Hamlet"
            , "import Text.Blaze.Html.Renderer.String"
            , "main = putStrLn $ renderHtml $(shamletFile \"foo.hamlet\")"
            ]

    update $ mconcat
        [ updateModule "Main.hs" mainContents
        , updateDataFile "foo.hamlet" "<p invalid"
        ]

    -- Error message expected, invalid data file
    getSourceErrors sess >>= print

    update $ updateDataFile "foo.hamlet" "<p>Valid"

    -- No error message expected, but got one anyway
    getSourceErrors sess >>= print

    update $ updateModule "Main.hs" $ mainContents `L8.append` "\n\n-- Trigger a recompile"
    -- No more error message
    getSourceErrors sess >>= print

    update $ updateDataFile "foo.hamlet" "<p invalid"

    -- Error message expected, but recompilation is not triggered.
    getSourceErrors sess >>= print

I would expect the last set of error messages to be non-empty due to the invalid Hamlet syntax. Instead, I'm getting no error messages. This seems to be the cause of a production bug: fpco/fpco#2338.

Pinging @jwiegley, we should make sure to get an isolation-runner test for this.

[Mikolaj]

To make sure and leave a link: I suppose the (ide-backend and perhaps your own?) tests for #94 pass OK on your system? (I mean the following tests: 1169525)

[snoyberg]

When I run locally, all tests pass. I haven't looked at the individual test cases that are run.

[snoyberg]

I've confirmed that this occurs even without Hamlet being involved:

{-# LANGUAGE OverloadedStrings #-}

module Main where

import IdeSession
import Data.Monoid (mconcat)
import qualified Data.ByteString.Lazy.Char8 as L8

main :: IO ()
main = do
    sess <- initSession (SessionInitParams Nothing) defaultSessionConfig
                { configDir             = "."
                }

    let cb = \_ -> return ()
        update = flip (updateSession sess) cb

    let mainContents = L8.pack $ unlines
            [ "{-# LANGUAGE TemplateHaskell #-}"
            , "import Language.Haskell.TH.Syntax"
            , "main = print ($(do"
            , "  qAddDependentFile \"foo.hamlet\""
            , "  s <- qRunIO $ readFile \"foo.hamlet\""
            , "  lift $ (read s :: Int)"
            , "  ) :: Int)"
            ]

    update $ mconcat
        [ updateModule "Main.hs" mainContents
        , updateDataFile "foo.hamlet" "invalid"
        ]

    -- Error message expected, invalid data file
    getSourceErrors sess >>= print . length

    update $ updateDataFile "foo.hamlet" "42"

    -- No error message expected, but got one anyway
    getSourceErrors sess >>= print . length

    update $ updateModule "Main.hs" $ mainContents `L8.append` "\n\n-- Trigger a recompile"
    -- No more error message
    getSourceErrors sess >>= print . length

    update $ updateDataFile "foo.hamlet" "invalid"

    -- Error message expected, but recompilation is not triggered.
    getSourceErrors sess >>= print . length

[edsko]

Just for the record, this is a problem with ghc itself; see also https://github.com/fpco/fpco/issues/2338.

[dcoutts]

We now have a fix for this (ie fix to ghc) in our experimental branch. See also #134 however.

[edsko]

Confirmed that this is resolved by the ghc patch.