5a71632 · typeable/pandoc

git.iko.soy/typeable/pandoc

Parsing: `many1Till`: Check for the end condition before parsing

By not checking for the end condition before the first parse, the parser was applied too often, consuming too much of the input. This fixes the behaviour of `testStringWith (many1Till (oneOf "ab") (string "aa")) "aaa"` which before incorrectly returned `Right "a"`. With this change, it instead correctly fails with `Left (PandocParsecError ...)` because it is not able to parse at least one occurence of `oneOf "ab"` that is not `"aa"`. Note that this only affects `many1Till p end` where `p` matches on a prefix of `end`.

commit5a71632d11f86cb6ed700fe3a4d717a56bad1b9e authorHerwig Stuetz <herwig@herwigstuetz.com> dateTue 23 May 2017, 23:21 (9 years ago) parentafb5514

4 files changed+8−7

compare with master

unifiedsplit

src/Text/Pandoc/Parsing.hs+3−2 src/Text/Pandoc/Readers/Org/Inlines.hs+2−2 src/Text/Pandoc/Readers/TWiki.hs+2−2 src/Text/Pandoc/Readers/Textile.hs+1−1

src/Text/Pandoc/Parsing.hs+3−2

@@ -274,11 +274,12 @@ indentWith num = do

274274 , try (char '\t' >> indentWith (num - tabStop)) ]

275275

276276-- | Like @manyTill@, but reads at least one item.

277−many1Till :: Stream s m t

277+many1Till :: (Show end, Stream s m t)

278278 => ParserT s st m a

279279 -> ParserT s st m end

280280 -> ParserT s st m [a]

281281many1Till p end = do

282+ notFollowedBy' end

282283 first <- p

283284 rest <- manyTill p end

284285 return (first:rest)

@@ -343,7 +344,7 @@ blanklines :: Stream s m Char => ParserT s st m [Char]

343344blanklines = many1 blankline

344345

345346-- | Parses material enclosed between start and end parsers.

346−enclosed :: Stream s m Char => ParserT s st m t -- ^ start parser

347+enclosed :: (Show end, Stream s m Char) => ParserT s st m t -- ^ start parser

347348 -> ParserT s st m end -- ^ end parser

348349 -> ParserT s st m a -- ^ content parser (to be used repeatedly)

349350 -> ParserT s st m [a]

src/Text/Pandoc/Readers/Org/Inlines.hs+2−2

@@ -687,13 +687,13 @@ mathEnd c = try $ do

687687 return res

688688

689689

690−enclosedInlines :: PandocMonad m => OrgParser m a

690+enclosedInlines :: (PandocMonad m, Show b) => OrgParser m a

691691 -> OrgParser m b

692692 -> OrgParser m (F Inlines)

693693enclosedInlines start end = try $

694694 trimInlinesF . mconcat <$> enclosed start end inline

695695

696−enclosedRaw :: PandocMonad m => OrgParser m a

696+enclosedRaw :: (PandocMonad m, Show b) => OrgParser m a

697697 -> OrgParser m b

698698 -> OrgParser m String

699699enclosedRaw start end = try $

src/Text/Pandoc/Readers/TWiki.hs+2−2

@@ -349,13 +349,13 @@ linebreak = newline >> notFollowedBy newline >> (lastNewline <|> innerNewline)

349349 where lastNewline = eof >> return mempty

350350 innerNewline = return B.space

351351

352−between :: (Monoid c, PandocMonad m)

352+between :: (Monoid c, PandocMonad m, Show b)

353353 => TWParser m a -> TWParser m b -> (TWParser m b -> TWParser m c)

354354 -> TWParser m c

355355between start end p =

356356 mconcat <$> try (start >> notFollowedBy whitespace >> many1Till (p end) end)

357357

358−enclosed :: (Monoid b, PandocMonad m)

358+enclosed :: (Monoid b, PandocMonad m, Show a)

359359 => TWParser m a -> (TWParser m a -> TWParser m b) -> TWParser m b

360360enclosed sep p = between sep (try $ sep <* endMarker) p

361361 where

src/Text/Pandoc/Readers/Textile.hs+1−1

@@ -692,7 +692,7 @@ langAttr = do

692692 return $ \(id',classes,keyvals) -> (id',classes,("lang",lang):keyvals)

693693

694694-- | Parses material surrounded by a parser.

695−surrounded :: PandocMonad m

695+surrounded :: (PandocMonad m, Show t)

696696 => ParserT [Char] st m t -- ^ surrounding parser

697697 -> ParserT [Char] st m a -- ^ content parser (to be used repeatedly)

698698 -> ParserT [Char] st m [a]