New treatment of dashes in --smart mode.

* `---` is always em-dash, `--` is always en-dash.
* pandoc no longer tries to guess when `-` should be en-dash.
* A new option, `--old-dashes`, is provided for legacy documents.

Rationale: The rules for en-dash are too complex and
language-dependent for a guesser to work reliably.  This
change gives users greater control.  The alternative of
using unicode isn't very good, since unicode em- and en-
dashes are barely distinguishable in a monospace font.

6 files changed, 56 insertions, 15 deletions

diff --git a/README b/README
index 9380c895e..5b4de3942 100644
--- a/README
+++ b/README
@@ -206,11 +206,17 @@ Options
 
 `-S`, `--smart`
 :   Produce typographically correct output, converting straight quotes
-    to curly quotes, `---` and `--` to dashes, and `...` to ellipses.
-    Nonbreaking spaces are inserted after certain abbreviations, such
-    as "Mr." (Note: This option is significant only when the input format is
-    `markdown` or `textile`.  It is selected automatically when the input
-    format is `textile` or the output format is `latex` or `context`.)
+    to curly quotes, `---` to em-dashes, `--` to en-dashes, and
+    `...` to ellipses. Nonbreaking spaces are inserted after certain
+    abbreviations, such as "Mr." (Note: This option is significant only when
+    the input format is `markdown` or `textile`. It is selected automatically
+    when the input format is `textile` or the output format is `latex` or
+    `context`.)
+
+`--old-dashes`
+:   Selects the pandoc <= 1.8.2.1 behavior for parsing smart dashes: `-` before
+    a numeral is an en-dash, and `--` is an em-dash.  This option is selected
+    automatically for `textile` input.
 
 `-5`, `--html5`
 :   Produce HTML5 instead of HTML4.  This option has no effect for writers
diff --git a/src/Text/Pandoc/Parsing.hs b/src/Text/Pandoc/Parsing.hs
index c2c512033..71da3a730 100644
--- a/src/Text/Pandoc/Parsing.hs
+++ b/src/Text/Pandoc/Parsing.hs
@@ -614,6 +614,9 @@ data ParserState = ParserState
       stateDate            :: [Inline],      -- ^ Date of document
       stateStrict          :: Bool,          -- ^ Use strict markdown syntax?
       stateSmart           :: Bool,          -- ^ Use smart typography?
+      stateOldDashes       :: Bool,          -- ^ Use pandoc <= 1.8.2.1 behavior
+                                             --   in parsing dashes; -- is em-dash;
+                                             --   before numeral is en-dash
       stateLiterateHaskell :: Bool,          -- ^ Treat input as literate haskell
       stateColumns         :: Int,           -- ^ Number of columns in terminal
       stateHeaderTable     :: [HeaderType],  -- ^ Ordered list of header types used
@@ -642,6 +645,7 @@ defaultParserState =
                   stateDate            = [],
                   stateStrict          = False,
                   stateSmart           = False,
+                  stateOldDashes       = False,
                   stateLiterateHaskell = False,
                   stateColumns         = 80,
                   stateHeaderTable     = [],
@@ -788,17 +792,37 @@ ellipses = do
   try (charOrRef "\8230\133") <|> try (string "..." >> return '…')
   return (Str "\8230")
 
-dash :: GenParser Char st Inline
-dash = enDash <|> emDash
+dash :: GenParser Char ParserState Inline
+dash = do
+  oldDashes <- stateOldDashes `fmap` getState
+  if oldDashes
+     then emDashOld <|> enDashOld
+     else Str `fmap` (hyphenDash <|> emDash <|> enDash)
 
-enDash :: GenParser Char st Inline
+-- Two hyphens = en-dash, three = em-dash
+hyphenDash :: GenParser Char st String
+hyphenDash = do
+  try $ string "--"
+  option "\8211" (char '-' >> return "\8212")
+
+emDash :: GenParser Char st String
+emDash = do
+  try (charOrRef "\8212\151")
+  return "\8212"
+
+enDash :: GenParser Char st String
 enDash = do
+  try (charOrRef "\8212\151")
+  return "\8211"
+
+enDashOld :: GenParser Char st Inline
+enDashOld = do
   try (charOrRef "\8211\150") <|>
     try (char '-' >> lookAhead (satisfy isDigit) >> return '–')
   return (Str "\8211")
 
-emDash :: GenParser Char st Inline
-emDash = do
+emDashOld :: GenParser Char st Inline
+emDashOld = do
   try (charOrRef "\8212\151") <|> (try $ string "--" >> optional (char '-') >> return '-')
   return (Str "\8212")
 
diff --git a/src/Text/Pandoc/Readers/Textile.hs b/src/Text/Pandoc/Readers/Textile.hs
index 4693bd06d..3b5954368 100644
--- a/src/Text/Pandoc/Readers/Textile.hs
+++ b/src/Text/Pandoc/Readers/Textile.hs
@@ -68,7 +68,8 @@ import Control.Monad ( guard, liftM )
 readTextile :: ParserState -- ^ Parser state, including options for parser
              -> String      -- ^ String to parse (assuming @'\n'@ line endings)
              -> Pandoc
-readTextile state s = (readWith parseTextile) state (s ++ "\n\n")
+readTextile state s =
+  (readWith parseTextile) state{ stateOldDashes = True } (s ++ "\n\n")
 
 
 --
diff --git a/src/pandoc.hs b/src/pandoc.hs
index fc28c4c3f..3660fc167 100644
--- a/src/pandoc.hs
+++ b/src/pandoc.hs
@@ -103,6 +103,7 @@ data Opt = Opt
     , optSelfContained     :: Bool    -- ^ Make HTML accessible offline
     , optXeTeX             :: Bool    -- ^ Format latex for xetex
     , optSmart             :: Bool    -- ^ Use smart typography
+    , optOldDashes         :: Bool    -- ^ Parse dashes like pandoc <=1.8.2.1
     , optHtml5             :: Bool    -- ^ Produce HTML5 in HTML
     , optHighlight         :: Bool    -- ^ Highlight source code
     , optHighlightStyle    :: Style   -- ^ Style to use for highlighted code
@@ -149,6 +150,7 @@ defaultOpts = Opt
     , optSelfContained     = False
     , optXeTeX             = False
     , optSmart             = False
+    , optOldDashes         = False
     , optHtml5             = False
     , optHighlight         = True
     , optHighlightStyle    = pygments
@@ -245,6 +247,12 @@ options =
                   (\opt -> return opt { optSmart = True }))
                  "" -- "Use smart quotes, dashes, and ellipses"
 
+    , Option "" ["old-dashes"]
+                 (NoArg
+                  (\opt -> return opt { optSmart = True
+                                      , optOldDashes = True }))
+                 "" -- "Use smart quotes, dashes, and ellipses"
+
     , Option "5" ["html5"]
                  (NoArg
                   (\opt -> do
@@ -735,6 +743,7 @@ main = do
               , optIncremental       = incremental
               , optSelfContained     = selfContained
               , optSmart             = smart
+              , optOldDashes         = oldDashes
               , optHtml5             = html5
               , optHighlight         = highlight
               , optHighlightStyle    = highlightStyle
@@ -858,6 +867,7 @@ main = do
                               stateCitations       = map CSL.refId refs,
                               stateSmart           = smart || writerName' `elem`
                                                      ["latex", "context", "latex+lhs", "beamer"],
+                              stateOldDashes       = oldDashes,
                               stateColumns         = columns,
                               stateStrict          = strict,
                               stateIndentedCodeClasses = codeBlockClasses,
diff --git a/tests/rst-reader.native b/tests/rst-reader.native
index 8d273a1d7..e0eb4d438 100644
--- a/tests/rst-reader.native
+++ b/tests/rst-reader.native
@@ -165,14 +165,14 @@ Pandoc (Meta {docTitle = [Str "Pandoc",Space,Str "Test",Space,Str "Suite",Str ":
   ,([Str "city"],
     [[Para [Emph [Str "Nowhere"],Str ",",Space,Str "MA,",Space,Str "USA"]]])
   ,([Str "phone"],
-    [[Para [Str "123",Str "\8211",Str "4567"]]])]]
+    [[Para [Str "123",Str "-",Str "4567"]]])]]
 ,DefinitionList
  [([Str "address"],
    [[Para [Str "61",Space,Str "Main",Space,Str "St",Str "."]]])
  ,([Str "city"],
    [[Para [Emph [Str "Nowhere"],Str ",",Space,Str "MA,",Space,Str "USA"]]])
  ,([Str "phone"],
-   [[Para [Str "123",Str "\8211",Str "4567"]]])]
+   [[Para [Str "123",Str "-",Str "4567"]]])]
 ,Header 1 [Str "HTML",Space,Str "Blocks"]
 ,Para [Str "Simple",Space,Str "block",Space,Str "on",Space,Str "one",Space,Str "line",Str ":"]
 ,RawBlock "html" "<div>foo</div>\n"
diff --git a/tests/testsuite.txt b/tests/testsuite.txt
index ccee0764a..3bb5d8cb5 100644
--- a/tests/testsuite.txt
+++ b/tests/testsuite.txt
@@ -492,9 +492,9 @@ So is 'pine.'
 
 Here is some quoted '`code`' and a "[quoted link][1]".
 
-Some dashes:  one---two --- three--four -- five.
+Some dashes:  one---two --- three---four --- five.
 
-Dashes between numbers: 5-7, 255-66, 1987-1999. 
+Dashes between numbers: 5--7, 255--66, 1987--1999. 
 
 Ellipses...and...and....