From 7a602d222f6ba005a901d6b3322f9392602d6b9d Mon Sep 17 00:00:00 2001
From: John MacFarlane <fiddlosopher@gmail.com>
Date: Tue, 7 Feb 2012 21:50:55 -0800
Subject: Limit nesting of strong/emph.

This avoids exponential lookahead in parasitic cases, like
a**a*a**a*a**a*a**a*a**a*a**a*a**a*a**.

Added stateMaxNestingLevel to ParserState.

We set this to 6, so you can still have Emph inside Emph, just not
indefinitely.
---
 src/Text/Pandoc/Parsing.hs          |  2 ++
 src/Text/Pandoc/Readers/Markdown.hs | 16 ++++++++++++++--
 2 files changed, 16 insertions(+), 2 deletions(-)

(limited to 'src')

diff --git a/src/Text/Pandoc/Parsing.hs b/src/Text/Pandoc/Parsing.hs
index 0468ceec5..725621ce2 100644
--- a/src/Text/Pandoc/Parsing.hs
+++ b/src/Text/Pandoc/Parsing.hs
@@ -629,6 +629,7 @@ data ParserState = ParserState
     { stateParseRaw        :: Bool,          -- ^ Parse raw HTML and LaTeX?
       stateParserContext   :: ParserContext, -- ^ Inside list?
       stateQuoteContext    :: QuoteContext,  -- ^ Inside quoted environment?
+      stateMaxNestingLevel :: Int,           -- ^ Max # of nested Strong/Emph
       stateLastStrPos      :: Maybe SourcePos, -- ^ Position after last str parsed
       stateKeys            :: KeyTable,      -- ^ List of reference keys
       stateCitations       :: [String],      -- ^ List of available citations
@@ -660,6 +661,7 @@ defaultParserState =
     ParserState { stateParseRaw        = False,
                   stateParserContext   = NullState,
                   stateQuoteContext    = NoQuote,
+                  stateMaxNestingLevel = 6,
                   stateLastStrPos      = Nothing,
                   stateKeys            = M.empty,
                   stateCitations       = [],
diff --git a/src/Text/Pandoc/Readers/Markdown.hs b/src/Text/Pandoc/Readers/Markdown.hs
index 607d0971a..31db51028 100644
--- a/src/Text/Pandoc/Readers/Markdown.hs
+++ b/src/Text/Pandoc/Readers/Markdown.hs
@@ -1041,8 +1041,20 @@ inlinesBetween start end =
     where inner      = innerSpace <|> (notFollowedBy' whitespace >> inline)
           innerSpace = try $ whitespace >>~ notFollowedBy' end
 
+-- This is used to prevent exponential blowups for things like:
+-- a**a*a**a*a**a*a**a*a**a*a**a*a**
+nested :: GenParser Char ParserState a
+       -> GenParser Char ParserState a
+nested p = do
+  nestlevel <- stateMaxNestingLevel `fmap` getState
+  guard $ nestlevel > 0
+  updateState $ \st -> st{ stateMaxNestingLevel = stateMaxNestingLevel st - 1 }
+  res <- p
+  updateState $ \st -> st{ stateMaxNestingLevel = nestlevel }
+  return res
+
 emph :: GenParser Char ParserState Inline
-emph = Emph `liftM`
+emph = Emph `fmap` nested
   (inlinesBetween starStart starEnd <|> inlinesBetween ulStart ulEnd)
     where starStart = char '*' >> lookAhead nonspaceChar
           starEnd   = notFollowedBy' strong >> char '*'
@@ -1050,7 +1062,7 @@ emph = Emph `liftM`
           ulEnd     = notFollowedBy' strong >> char '_'
 
 strong :: GenParser Char ParserState Inline
-strong = Strong `liftM`
+strong = Strong `liftM` nested
   (inlinesBetween starStart starEnd <|> inlinesBetween ulStart ulEnd)
     where starStart = string "**" >> lookAhead nonspaceChar
           starEnd   = try $ string "**"
-- 
cgit v1.2.3