Convert audio names to md5 digest

owickstrom · Jan 14, 2019 · baa2bc0 · baa2bc0
1 parent f4db3f0
commit baa2bc0
Show file tree

Hide file tree

Showing 2 changed files with 9 additions and 4 deletions.
diff --git a/komposition.cabal b/komposition.cabal
@@ -89,6 +89,7 @@ library
   build-depends:        base >=4.10 && <5
                       , async
                       , binary
+                      , bytestring
                       , deepseq
                       , directory
                       , exceptions
@@ -119,6 +120,7 @@ library
                       , pipes-safe
                       , primitive
                       , protolude
+                      , pureMD5
                       , typed-process >= 0.2 && < 0.3
                       , row-types
                       , safe-exceptions

diff --git a/src/Komposition/Import/Audio/Sox.hs b/src/Komposition/Import/Audio/Sox.hs
@@ -19,6 +19,8 @@ import           Control.Effect.Carrier
 import           Control.Effect.Sum
 import           Control.Monad.Catch        hiding (bracket)
 import qualified Data.Char                  as Char
+import qualified Data.ByteString.Lazy.Char8 as Lazy
+import           Data.Digest.Pure.MD5
 import qualified Data.Text                  as Text
 import           Data.Time.Clock
 import           Pipes
@@ -167,8 +169,7 @@ transcodeAudioFileToWav
   -> FilePath
   -> Producer ProgressUpdate m FilePath
 transcodeAudioFileToWav tempDir fullLength inPath = do
-  -- TODO: use md5 digest to avoid collisions
-  let outPath = tempDir </> takeBaseName inPath <> ".wav"
+  let outPath = tempDir </> createMD5BaseName inPath <> ".wav"
       cmd =
         Command
         { output = Command.FileOutput outPath
@@ -200,12 +201,11 @@ instance (MonadIO m, Carrier sig m) => Carrier (AudioImport :+: sig) (SoxAudioIm
             (liftIO . removeDirectoryRecursive)
             $ \tempDir -> do
               fullLength <- getAudioFileDuration srcFile
-              -- TODO: use file md5 digest in filename (or for a subdirectory) to avoid collisions
               chunks <-
                 divideProgress4
                   (transcodeAudioFileToWav tempDir fullLength srcFile)
                   (normalizeAudio tempDir)
-                  (splitAudioBySilence (outDir </> "audio-chunks") (takeBaseName srcFile <> "-%5n.wav"))
+                  (splitAudioBySilence (outDir </> "audio-chunks") (createMD5BaseName srcFile <> "-%5n.wav"))
                   dropSilentChunks
               lift (mapM (filePathToAudioAsset outDir) chunks)
         Unclassified -> k $ do
@@ -225,6 +225,9 @@ instance (MonadIO m, Carrier sig m) => Carrier (AudioImport :+: sig) (SoxAudioIm
         -- TODO: Check that it can be processed, not just checking the extension
         k (takeExtension p `elem` [".wav", ".mp3", ".m4a", ".aiff", ".aac"])
 
+createMD5BaseName :: [Char] -> [Char]
+createMD5BaseName = show .md5 . Lazy.pack . takeBaseName
+
 runSoxAudioImport :: (MonadIO m, Carrier sig m) => Eff (SoxAudioImportC m) a -> m a
 runSoxAudioImport = runSoxAudioImportC . interpret