Just use top level benchmarks

semantic-tags/bench/Main.hs

@@ -1,33 +0,0 @@
-module Main (main) where
-
-import Tags.Tagging.Precise
-import qualified Data.ByteString as B
-import Gauge
-import Source.Loc (Loc (..))
-import Source.Range
-import Source.Source as Source
-import Source.Span (Pos (..), Span (..))
-import Data.Char (ord)
-
-main :: IO ()
-main = do
-  bytes <- fmap fromUTF8 $ B.readFile "/Users/tclem/github/github/app/models/user.rb"
-  -- bytes <- fmap fromUTF8 $ B.readFile "/Users/tclem/Downloads/jquery-3.5.1.min.js"
-  -- let lineIndexes = B.elemIndices (toEnum (ord '\n')) (Source.bytes bytes)
-  defaultMain
-    -- [ baselineCalculateLine bytes
-    -- , calculateLinePRVersion bytes
-    [ calculateLineWithMap bytes
-    ]
-
--- baselineCalculateLine :: Source -> Benchmark
--- baselineCalculateLine bytes = bench "baseline calculateLineAndSpans" $ do
---   nf (baselineCalculateLineAndSpans bytes) (Loc (Range 73754 73757) (Span (Pos 1 73669) (Pos 1 73672)))
-
--- calculateLinePRVersion :: Source -> Benchmark
--- calculateLinePRVersion bytes = bench "calculateLineAndSpansPRVersion " $ do
---   nf (calculateLineAndSpansPRVersion bytes) (Loc (Range 73754 73757) (Span (Pos 1 73669) (Pos 1 73672)))
-
-calculateLineWithMap :: Source -> Benchmark
-calculateLineWithMap src = bench "calculateLineAndSpansWithIndexes" $ do
-  nf (calculateLineAndSpans src (LineIndices mempty)) (Loc (Range 73754 73757) (Span (Pos 1 73669) (Pos 1 73672)))

semantic-tags/semantic-tags.cabal

@@ -66,15 +66,3 @@ test-suite test
     , tasty-hedgehog  ^>= 1.0.0.1
     , tasty-hunit      >= 0.10 && <1
     , text
-
-benchmark benchmarks
-  import:              haskell
-  hs-source-dirs:      bench
-  type:                exitcode-stdio-1.0
-  main-is:             Main.hs
-  ghc-options:         -threaded -static -rtsopts
-  build-depends:       base
-                     , bytestring         ^>= 0.10.8.2
-                     , semantic-tags
-                     , semantic-source
-                     , gauge ^>= 0.2.5