on master help sign in
Narrow the set of whitespace characters considered valid
8ef622cBryan O'Sullivan
Data/Aeson/Parser/Internal.hs+7−1
@@ -39,7 +39,7 @@ import Prelude.Compat
39import Control.Applicative ((<|>))
39import Control.Applicative ((<|>))
40import Control.Monad (void, when)
40import Control.Monad (void, when)
41import Data.Aeson.Types.Internal (IResult(..), JSONPath, Result(..), Value(..))
41import Data.Aeson.Types.Internal (IResult(..), JSONPath, Result(..), Value(..))
42−import Data.Attoparsec.ByteString.Char8 (Parser, char, decimal, endOfInput, isDigit_w8, signed, skipSpace, string)
42+import Data.Attoparsec.ByteString.Char8 (Parser, char, decimal, endOfInput, isDigit_w8, signed, string)
43import Data.Scientific (Scientific)
43import Data.Scientific (Scientific)
44import Data.Text (Text)
44import Data.Text (Text)
45import Data.Vector as Vector (Vector, empty, fromListN, reverse)
45import Data.Vector as Vector (Vector, empty, fromListN, reverse)
@@ -295,6 +295,12 @@ jsonEOF = json <* skipSpace <* endOfInput
295jsonEOF' :: Parser Value
295jsonEOF' :: Parser Value
296jsonEOF' = json' <* skipSpace <* endOfInput
296jsonEOF' = json' <* skipSpace <* endOfInput
297
297
298+-- | The only valid whitespace in a JSON document is space, newline,
299+-- carriage return, and tab.
300+skipSpace :: Parser ()
301+skipSpace = A.skipWhile $ \w -> w == 0x20 || w == 0x0a || w == 0x0d || w == 0x09
302+{-# INLINE skipSpace #-}
303+
298------------------ Copy-pasted and adapted from attoparsec ------------------
304------------------ Copy-pasted and adapted from attoparsec ------------------
299
305
300-- A strict pair
306-- A strict pair
tests/UnitTests.hs−1
@@ -670,7 +670,6 @@ _blacklist = HashSet.fromList [
670 , "n_string_unescaped_crtl_char.json"
670 , "n_string_unescaped_crtl_char.json"
671 , "n_string_unescaped_newline.json"
671 , "n_string_unescaped_newline.json"
672 , "n_string_unescaped_tab.json"
672 , "n_string_unescaped_tab.json"
673− , "n_structure_whitespace_formfeed.json"
674 , "string_1_escaped_invalid_codepoint.json"
673 , "string_1_escaped_invalid_codepoint.json"
675 , "string_1_invalid_codepoint.json"
674 , "string_1_invalid_codepoint.json"
676 , "string_1_invalid_codepoints.json"
675 , "string_1_invalid_codepoints.json"