Dmitrii Kovanikov

It would be really good to have `[-v|--version]` option which prints current version and perfectly current used commit of `importify`.

Multiple qualified imports sharing same name are not touched when one is unused. ```haskell import qualified Distribution.ModuleName as Cabal import qualified Language.Haskell.Extension as Cabal (Extension (..), KnownExtension (..)) ```

Consider next imports: ```haskell import Language.Haskell.Names (symbolName) import qualified Language.Haskell.Names as N ``` GHC prints warning: ``` The import of ‘N.Symbol(symbolName)’ from module ‘Language.Haskell.Names’ is redundant ``` if only `N.symbolName`...

See next example: ```haskell {-# LANGUAGE TemplateHaskell #-} import Control.Lens (makeLenses) data Foo = Foo { _foo :: Int } makeLenses ''Foo ``` In this case `makeLenses` is considered unused...

Currently we don't do this because `importify` is parsing-only tool. Our decision about detecting instances provided below: > Options we have regarding detecting unused instainces: > 1. Parse stack build...

Consider following situation. Module `X` from package `P` exports function `foo`. Package `Q` depends on `P`, imports `X (foo)` and exports this `foo`. When this happen `importify` doesn't add `X.foo`...

Support TOML spec v1.0.0

2

The latest spec is released here: * https://toml.io/en/v1.0.0 Need to check what changed since 0.5.0. Probably, will be easier to do together with #372.

[RFC] Different TOML AST

2

An idea I had in mind for a while, but not sure, whether it's better or worse. Currently, the TOML type looks like this: ```haskell data TOML = TOML {...

Verify with the official TOML compliance test-suite

2

https://github.com/toml-lang/compliance

Instead of `megaparsec`. And benchmark the newly written solution. As we are going to rewrite the parser entirely, we can as well fix a few more issues on the way,...