dataframe-csv-th 1.3.0.0 → 1.4.0.0

raw patch · 3 files changed

+42/−21 lines, 3 filesdep +bytestringdep ~dataframe-th

Dependencies added: bytestring

Dependency ranges changed: dataframe-th

Files

• dataframe-csv-th.cabal +3/−2
• src/DataFrame/TH/CSV.hs +38/−12
• src/DataFrame/Typed/TH/CSV.hs +1/−7

@@ -1,6 +1,6 @@ cabal-version:      2.4 name:               dataframe-csv-th-version:            1.3.0.0+version:            1.4.0.0 synopsis:           CSV-file-based Template Haskell splices for the dataframe ecosystem. description:     Splices that read CSV files at compile time and emit per-column@@ -30,8 +30,9 @@                         DataFrame.TH.CSV                         DataFrame.Typed.TH.CSV     build-depends:      base >= 4 && < 5,+                        bytestring >= 0.11 && < 0.14,                         dataframe-csv >= 2.3 && < 2.4,-                        dataframe-th >= 2.2 && < 2.3,+                        dataframe-th >= 2.3 && < 2.4,                         template-haskell >= 2.0 && < 3     hs-source-dirs:     src     default-language:   Haskell2010

@@ -1,27 +1,25 @@ {-# LANGUAGE FlexibleContexts #-}--{- |-Module      : DataFrame.TH.CSV-License     : MIT+{-# LANGUAGE OverloadedStrings #-} -CSV-file-based 'DataFrame.TH' splices. Splits out the CSV ingest path so-@dataframe-th@ stays IO-agnostic.--}+-- | Splices that derive column bindings from a CSV file. module DataFrame.TH.CSV (     declareColumnsFromCsvFile,     declareColumnsFromCsvWithOpts,+    deriveSchemaValuesFromCsvFile,+    deriveSchemaValuesFromCsvWithOpts, ) where +import qualified Data.ByteString as BS+ import Control.Monad.IO.Class (liftIO)+import Data.Maybe (fromMaybe) import Language.Haskell.TH+import System.IO (IOMode (ReadMode), withFile)  import qualified DataFrame.IO.CSV as CSV-import DataFrame.TH.Records (declareColumns)+import DataFrame.TH.Records (declareColumns, declareSchemaValues) -{- | Splice a binding for every column of the 'DataFrame' read from a CSV-file. Each binding has type @Expr T@ where @T@ is the inferred column-type.--}+-- | Splice an @Expr T@ binding for every column of a CSV file. declareColumnsFromCsvFile :: String -> DecsQ declareColumnsFromCsvFile path = do     df <-@@ -37,3 +35,31 @@ declareColumnsFromCsvWithOpts opts path = do     df <- liftIO (CSV.readSeparated opts path)     declareColumns df++deriveSchemaValuesFromCsvFile :: String -> String -> DecsQ+deriveSchemaValuesFromCsvFile =+    deriveSchemaValuesFromCsvWithOpts+        CSV.defaultReadOptions{CSV.numRowsToRead = Just 100}++deriveSchemaValuesFromCsvWithOpts ::+    CSV.ReadOptions -> String -> String -> DecsQ+deriveSchemaValuesFromCsvWithOpts opts prefix path = do+    df <- liftIO $ do+        sample <- readCsvSample sampleWindowBytes path+        CSV.decodeSeparatedStrict opts sample+    declareSchemaValues prefix df++sampleWindowBytes :: Int+sampleWindowBytes = 1024 * 1024++readCsvSample :: Int -> FilePath -> IO BS.ByteString+readCsvSample n path = do+    chunk <- withFile path ReadMode (\h -> BS.hGet h (n + 1))+    let stripBom b = fromMaybe b (BS.stripPrefix "\xEF\xBB\xBF" b)+    pure . stripBom $+        if BS.length chunk <= n+            then chunk+            else+                let window = BS.take n chunk+                 in maybe window (\i -> BS.take (i + 1) window) $+                        BS.elemIndexEnd 0x0a window

@@ -1,10 +1,4 @@-{- |-Module      : DataFrame.Typed.TH.CSV-License     : MIT--CSV-file-based typed-schema derivation. Splits the CSV path out of-@DataFrame.Typed.TH@ so it can live in @dataframe-csv-th@.--}+-- | Derives a typed schema from a CSV file. module DataFrame.Typed.TH.CSV (     deriveSchemaFromCsvFile,     deriveSchemaFromCsvFileWith,