Markdown reader: use CommonMark rules for list item nesting.

Closes #3511.

Previously pandoc used the four-space rule: continuation paragraphs,
sublists, and other block level content had to be indented 4
spaces.  Now the indentation required is determined by the
first line of the list item:  to be included in the list item,
blocks must be indented to the level of the first non-space
content after the list marker. Exception: if are 5 or more spaces
after the list marker, then the content is interpreted as an
indented code block, and continuation paragraphs must be indented
two spaces beyond the end of the list marker.  See the CommonMark
spec for more details and examples.

Documents that adhere to the four-space rule should, in most cases,
be parsed the same way by the new rules.  Here are some examples
of texts that will be parsed differently:

    - a
      - b

will be parsed as a list item with a sublist; under the four-space
rule, it would be a list with two items.

    - a

          code

Here we have an indented code block under the list item, even though it
is only indented six spaces from the margin, because it is four spaces
past the point where a continuation paragraph could begin.  With the
four-space rule, this would be a regular paragraph rather than a code
block.

    - a

            code

Here the code block will start with two spaces, whereas under
the four-space rule, it would start with `code`.  With the four-space
rule, indented code under a list item always must be indented eight
spaces from the margin, while the new rules require only that it
be indented four spaces from the beginning of the first non-space
text after the list marker (here, `a`).

This change was motivated by a slew of bug reports from people
who expected lists to work differently (#3125, #2367, #2575, #2210,
 #1990, #1137, #744, #172, #137, #128) and by the growing prevalance
of CommonMark (now used by GitHub, for example).

Users who want to use the old rules can select the `four_space_rule`
extension.

* Added `four_space_rule` extension.
* Added `Ext_four_space_rule` to `Extensions`.
* `Parsing` now exports `gobbleAtMostSpaces`, and the type
  of `gobbleSpaces` has been changed so that a `ReaderOptions`
  parameter is not needed.

8 files changed, 178 insertions, 117 deletions

diff --git a/MANUAL.txt b/MANUAL.txt
index 596ba73f2..7e0bdfcdd 100644
--- a/MANUAL.txt
+++ b/MANUAL.txt
@@ -2078,12 +2078,12 @@ But Markdown also allows a "lazy" format:
     list item.
     * and my second.
 
-### The four-space rule ###
+### Block content in list items ###
 
 A list item may contain multiple paragraphs and other block-level
 content. However, subsequent paragraphs must be preceded by a blank line
-and indented four spaces or a tab. The list will look better if the first
-paragraph is aligned with the rest:
+and indented to line up with the first non-space content after
+the list marker.
 
       * First paragraph.
 
@@ -2094,19 +2094,29 @@ paragraph is aligned with the rest:
 
             { code }
 
+Exception: if the list marker is followed by an indented code
+block, which must begin 5 spaces after the list marker, then
+subsequent paragraphs must begin two columns after the last
+character of the list marker:
+
+    *     code
+
+      continuation paragraph
+
 List items may include other lists.  In this case the preceding blank
-line is optional.  The nested list must be indented four spaces or
-one tab:
+line is optional.  The nested list must be indented to line up with
+the first non-space character after the list marker of the
+containing list item.
 
     * fruits
-        + apples
-            - macintosh
-            - red delicious
-        + pears
-        + peaches
+      + apples
+        - macintosh
+        - red delicious
+      + pears
+      + peaches
     * vegetables
-        + broccoli
-        + chard
+      + broccoli
+      + chard
 
 As noted above, Markdown allows you to write list items "lazily," instead of
 indenting continuation lines. However, if there are multiple paragraphs or
@@ -2121,21 +2131,6 @@ other blocks in a list item, the first line of each must be indented.
         Second paragraph of second
     list item.
 
-**Note:**  Although the four-space rule for continuation paragraphs
-comes from the official [Markdown syntax guide], the reference implementation,
-`Markdown.pl`, does not follow it. So pandoc will give different results than
-`Markdown.pl` when authors have indented continuation paragraphs fewer than
-four spaces.
-
-The [Markdown syntax guide] is not explicit whether the four-space
-rule applies to *all* block-level content in a list item; it only
-mentions paragraphs and code blocks.  But it implies that the rule
-applies to all block-level content (including nested lists), and
-pandoc interprets it that way.
-
-  [Markdown syntax guide]:
-    http://daringfireball.net/projects/markdown/syntax#list
-
 ### Ordered lists ###
 
 Ordered lists work just like bulleted lists, except that the items
@@ -3606,6 +3601,12 @@ implied by pandoc's default `all_symbols_escapable`.
 Allow a list to occur right after a paragraph, with no intervening
 blank space.
 
+#### Extension: `four_space_rule` ####
+
+Selects the pandoc <= 2.0 behavior for parsing lists, so that
+four spaces indent are needed for list item continuation
+paragraphs.
+
 #### Extension: `spaced_reference_links` ####
 
 Allow whitespace between the two components of a reference link,
diff --git a/src/Text/Pandoc/Extensions.hs b/src/Text/Pandoc/Extensions.hs
index e6a3ca044..95e59063b 100644
--- a/src/Text/Pandoc/Extensions.hs
+++ b/src/Text/Pandoc/Extensions.hs
@@ -111,6 +111,7 @@ data Extension =
     | Ext_autolink_bare_uris  -- ^ Make all absolute URIs into links
     | Ext_fancy_lists         -- ^ Enable fancy list numbers and delimiters
     | Ext_lists_without_preceding_blankline -- ^ Allow lists without preceding blank
+    | Ext_four_space_rule     -- ^ Require 4-space indent for list contents
     | Ext_startnum            -- ^ Make start number of ordered list significant
     | Ext_definition_lists    -- ^ Definition lists as in pandoc, mmd, php
     | Ext_compact_definition_lists  -- ^ Definition lists without
diff --git a/src/Text/Pandoc/Parsing.hs b/src/Text/Pandoc/Parsing.hs
index 37a0b53b4..9ed18d4e0 100644
--- a/src/Text/Pandoc/Parsing.hs
+++ b/src/Text/Pandoc/Parsing.hs
@@ -50,6 +50,7 @@ module Text.Pandoc.Parsing ( takeWhileP,
                              blankline,
                              blanklines,
                              gobbleSpaces,
+                             gobbleAtMostSpaces,
                              enclosed,
                              stringAnyCase,
                              parseFromString,
@@ -380,14 +381,33 @@ blanklines = many1 blankline
 
 -- | Gobble n spaces; if tabs are encountered, expand them
 -- and gobble some or all of their spaces, leaving the rest.
-gobbleSpaces :: Monad m => ReaderOptions -> Int -> ParserT [Char] st m ()
-gobbleSpaces _    0 = return ()
-gobbleSpaces opts n = try $ do
-  char ' ' <|> do char '\t'
-                  inp <- getInput
-                  setInput $ replicate (readerTabStop opts - 1) ' ' ++ inp
-                  return ' '
-  gobbleSpaces opts (n - 1)
+gobbleSpaces :: (HasReaderOptions st, Monad m)
+             => Int -> ParserT [Char] st m ()
+gobbleSpaces 0 = return ()
+gobbleSpaces n
+  | n < 0     = error "gobbleSpaces called with negative number"
+  | otherwise = try $ do
+      char ' ' <|> eatOneSpaceOfTab
+      gobbleSpaces (n - 1)
+
+eatOneSpaceOfTab :: (HasReaderOptions st, Monad m) => ParserT [Char] st m Char
+eatOneSpaceOfTab = do
+  char '\t'
+  tabstop <- getOption readerTabStop
+  inp <- getInput
+  setInput $ replicate (tabstop - 1) ' ' ++ inp
+  return ' '
+
+-- | Gobble up to n spaces; if tabs are encountered, expand them
+-- and gobble some or all of their spaces, leaving the rest.
+gobbleAtMostSpaces :: (HasReaderOptions st, Monad m)
+                   => Int -> ParserT [Char] st m Int
+gobbleAtMostSpaces 0 = return 0
+gobbleAtMostSpaces n
+  | n < 0     = error "gobbleAtMostSpaces called with negative number"
+  | otherwise = option 0 $ do
+      char ' ' <|> eatOneSpaceOfTab
+      (+ 1) <$> gobbleAtMostSpaces (n - 1)
 
 -- | Parses material enclosed between start and end parsers.
 enclosed :: (Show end, Stream s  m Char) => ParserT s st m t   -- ^ start parser
diff --git a/src/Text/Pandoc/Readers/Markdown.hs b/src/Text/Pandoc/Readers/Markdown.hs
index 26263d674..664691c8c 100644
--- a/src/Text/Pandoc/Readers/Markdown.hs
+++ b/src/Text/Pandoc/Readers/Markdown.hs
@@ -138,12 +138,7 @@ nonindentSpaces = do
 skipNonindentSpaces :: PandocMonad m => MarkdownParser m Int
 skipNonindentSpaces = do
   tabStop <- getOption readerTabStop
-  atMostSpaces (tabStop - 1) <* notFollowedBy spaceChar
-
-atMostSpaces :: PandocMonad m => Int -> MarkdownParser m Int
-atMostSpaces n
-  | n > 0     = (char ' ' >> (+1) <$> atMostSpaces (n-1)) <|> return 0
-  | otherwise = return 0
+  gobbleAtMostSpaces (tabStop - 1) <* notFollowedBy spaceChar
 
 litChar :: PandocMonad m => MarkdownParser m Char
 litChar = escapedChar'
@@ -809,49 +804,51 @@ blockQuote = do
 bulletListStart :: PandocMonad m => MarkdownParser m ()
 bulletListStart = try $ do
   optional newline -- if preceded by a Plain block in a list context
-  startpos <- sourceColumn <$> getPosition
   skipNonindentSpaces
   notFollowedBy' (() <$ hrule)     -- because hrules start out just like lists
   satisfy isBulletListMarker
-  endpos <- sourceColumn <$> getPosition
-  tabStop <- getOption readerTabStop
-  lookAhead (newline <|> spaceChar)
-  () <$ atMostSpaces (tabStop - (endpos - startpos))
+  gobbleSpaces 1 <|> () <$ lookAhead newline
+  try (gobbleAtMostSpaces 3 >> notFollowedBy spaceChar) <|> return ()
 
-anyOrderedListStart :: PandocMonad m => MarkdownParser m (Int, ListNumberStyle, ListNumberDelim)
-anyOrderedListStart = try $ do
+orderedListStart :: PandocMonad m
+                 => Maybe (ListNumberStyle, ListNumberDelim)
+                 -> MarkdownParser m (Int, ListNumberStyle, ListNumberDelim)
+orderedListStart mbstydelim = try $ do
   optional newline -- if preceded by a Plain block in a list context
-  startpos <- sourceColumn <$> getPosition
   skipNonindentSpaces
   notFollowedBy $ string "p." >> spaceChar >> digit  -- page number
-  res <- do guardDisabled Ext_fancy_lists
-            start <- many1 digit >>= safeRead
-            char '.'
-            return (start, DefaultStyle, DefaultDelim)
-     <|> do (num, style, delim) <- anyOrderedListMarker
-            -- if it could be an abbreviated first name,
-            -- insist on more than one space
-            when (delim == Period && (style == UpperAlpha ||
-                 (style == UpperRoman &&
-                  num `elem` [1, 5, 10, 50, 100, 500, 1000]))) $
-               () <$ spaceChar
-            return (num, style, delim)
-  endpos <- sourceColumn <$> getPosition
-  tabStop <- getOption readerTabStop
-  lookAhead (newline <|> spaceChar)
-  atMostSpaces (tabStop - (endpos - startpos))
-  return res
+  (do guardDisabled Ext_fancy_lists
+      start <- many1 digit >>= safeRead
+      char '.'
+      gobbleSpaces 1 <|> () <$ lookAhead newline
+      optional $ try (gobbleAtMostSpaces 3 >> notFollowedBy spaceChar)
+      return (start, DefaultStyle, DefaultDelim))
+   <|>
+   (do (num, style, delim) <- maybe
+          anyOrderedListMarker
+          (\(sty,delim) -> (\start -> (start,sty,delim)) <$>
+               orderedListMarker sty delim)
+          mbstydelim
+       gobbleSpaces 1 <|> () <$ lookAhead newline
+       -- if it could be an abbreviated first name,
+       -- insist on more than one space
+       when (delim == Period && (style == UpperAlpha ||
+            (style == UpperRoman &&
+             num `elem` [1, 5, 10, 50, 100, 500, 1000]))) $
+              () <$ lookAhead (newline <|> spaceChar)
+       optional $ try (gobbleAtMostSpaces 3 >> notFollowedBy spaceChar)
+       return (num, style, delim))
 
 listStart :: PandocMonad m => MarkdownParser m ()
-listStart = bulletListStart <|> (anyOrderedListStart >> return ())
+listStart = bulletListStart <|> (orderedListStart Nothing >> return ())
 
-listLine :: PandocMonad m => MarkdownParser m String
-listLine = try $ do
-  notFollowedBy' (do indentSpaces
-                     many spaceChar
+listLine :: PandocMonad m => Int -> MarkdownParser m String
+listLine continuationIndent = try $ do
+  notFollowedBy' (do gobbleSpaces continuationIndent
+                     skipMany spaceChar
                      listStart)
   notFollowedByHtmlCloser
-  optional (() <$ indentSpaces)
+  optional (() <$ gobbleSpaces continuationIndent)
   listLineCommon
 
 listLineCommon :: PandocMonad m => MarkdownParser m String
@@ -864,26 +861,39 @@ listLineCommon = concat <$> manyTill
 -- parse raw text for one list item, excluding start marker and continuations
 rawListItem :: PandocMonad m
             => MarkdownParser m a
-            -> MarkdownParser m String
+            -> MarkdownParser m (String, Int)
 rawListItem start = try $ do
+  pos1 <- getPosition
   start
+  pos2 <- getPosition
+  continuationIndent <- (4 <$ guardEnabled Ext_four_space_rule)
+                    <|> return (sourceColumn pos2 - sourceColumn pos1)
   first <- listLineCommon
   rest <- many (do notFollowedBy listStart
                    notFollowedBy (() <$ codeBlockFenced)
                    notFollowedBy blankline
-                   listLine)
+                   listLine continuationIndent)
   blanks <- many blankline
-  return $ unlines (first:rest) ++ blanks
+  let result = unlines (first:rest) ++ blanks
+  return (result, continuationIndent)
 
 -- continuation of a list item - indented and separated by blankline
 -- or (in compact lists) endline.
 -- note: nested lists are parsed as continuations
-listContinuation :: PandocMonad m => MarkdownParser m String
-listContinuation = try $ do
-  lookAhead indentSpaces
-  result <- many1 listContinuationLine
+listContinuation :: PandocMonad m => Int -> MarkdownParser m String
+listContinuation continuationIndent = try $ do
+  x <- try $ do
+         notFollowedBy blankline
+         notFollowedByHtmlCloser
+         gobbleSpaces continuationIndent
+         anyLineNewline
+  xs <- many $ try $ do
+         notFollowedBy blankline
+         notFollowedByHtmlCloser
+         gobbleSpaces continuationIndent <|> notFollowedBy' listStart
+         anyLineNewline
   blanks <- many blankline
-  return $ concat result ++ blanks
+  return $ concat (x:xs) ++ blanks
 
 notFollowedByHtmlCloser :: PandocMonad m => MarkdownParser m ()
 notFollowedByHtmlCloser = do
@@ -892,20 +902,12 @@ notFollowedByHtmlCloser = do
         Just t  -> notFollowedBy' $ htmlTag (~== TagClose t)
         Nothing -> return ()
 
-listContinuationLine :: PandocMonad m => MarkdownParser m String
-listContinuationLine = try $ do
-  notFollowedBy blankline
-  notFollowedBy' listStart
-  notFollowedByHtmlCloser
-  optional indentSpaces
-  anyLineNewline
-
 listItem :: PandocMonad m
          => MarkdownParser m a
          -> MarkdownParser m (F Blocks)
 listItem start = try $ do
-  first <- rawListItem start
-  continuations <- many listContinuation
+  (first, continuationIndent) <- rawListItem start
+  continuations <- many (listContinuation continuationIndent)
   -- parsing with ListItemState forces markers at beginning of lines to
   -- count as list item markers, even if not separated by blank space.
   -- see definition of "endline"
@@ -920,23 +922,14 @@ listItem start = try $ do
 
 orderedList :: PandocMonad m => MarkdownParser m (F Blocks)
 orderedList = try $ do
-  (start, style, delim) <- lookAhead anyOrderedListStart
+  (start, style, delim) <- lookAhead (orderedListStart Nothing)
   unless (style `elem` [DefaultStyle, Decimal, Example] &&
           delim `elem` [DefaultDelim, Period]) $
     guardEnabled Ext_fancy_lists
   when (style == Example) $ guardEnabled Ext_example_lists
   items <- fmap sequence $ many1 $ listItem
-                 ( try $ do
-                     optional newline -- if preceded by Plain block in a list
-                     startpos <- sourceColumn <$> getPosition
-                     skipNonindentSpaces
-                     res <- orderedListMarker style delim
-                     endpos <- sourceColumn <$> getPosition
-                     tabStop <- getOption readerTabStop
-                     lookAhead (newline <|> spaceChar)
-                     atMostSpaces (tabStop - (endpos - startpos))
-                     return res )
-  start' <- option 1 $ guardEnabled Ext_startnum >> return start
+                 (orderedListStart (Just (style, delim)))
+  start' <- (start <$ guardEnabled Ext_startnum) <|> return 1
   return $ B.orderedListWith (start', style, delim) <$> fmap compactify items
 
 bulletList :: PandocMonad m => MarkdownParser m (F Blocks)
@@ -1122,7 +1115,7 @@ rawHtmlBlocks = do
   updateState $ \st -> st{ stateInHtmlBlock = Just tagtype }
   let closer = htmlTag (\x -> x ~== TagClose tagtype)
   let block' = do notFollowedBy' closer
-                  atMostSpaces indentlevel
+                  gobbleAtMostSpaces indentlevel
                   block
   contents <- mconcat <$> many block'
   result <-
diff --git a/test/command/2434.md b/test/command/2434.md
index aa03e5fc3..4f12b6f56 100644
--- a/test/command/2434.md
+++ b/test/command/2434.md
@@ -31,7 +31,7 @@
 
 ```
 % pandoc -t opendocument
-(@)  text
+(@) text
 
     some text
 
diff --git a/test/command/3511.md b/test/command/3511.md
new file mode 100644
index 000000000..b8bcedbb0
--- /dev/null
+++ b/test/command/3511.md
@@ -0,0 +1,46 @@
+```
+% pandoc -t native
+- a
+  - b
+    - c
+
+-     code
+
+1000. one
+
+    not continuation
+^D
+[BulletList
+ [[Plain [Str "a"]
+  ,BulletList
+   [[Plain [Str "b"]
+    ,BulletList
+     [[Plain [Str "c"]]]]]]
+ ,[CodeBlock ("",[],[]) "code"]]
+,OrderedList (1000,Decimal,Period)
+ [[Plain [Str "one"]]]
+,CodeBlock ("",[],[]) "not continuation"]
+```
+
+```
+% pandoc -t native -f markdown+four_space_rule
+- a
+  - b
+    - c
+
+-     not code
+
+1000. one
+
+    continuation
+^D
+[BulletList
+ [[Plain [Str "a"]]
+ ,[Plain [Str "b"]
+  ,BulletList
+   [[Plain [Str "c"]]]]
+ ,[CodeBlock ("",[],[]) "not code"]]
+,OrderedList (1000,Decimal,Period)
+ [[Para [Str "one"]
+  ,Para [Str "continuation"]]]]
+```
diff --git a/test/markdown-reader-more.native b/test/markdown-reader-more.native
index a24417ffe..2e55dbb18 100644
--- a/test/markdown-reader-more.native
+++ b/test/markdown-reader-more.native
@@ -29,13 +29,13 @@ Pandoc (Meta {unMeta = fromList [("author",MetaList [MetaInlines [Str "Author",S
  ,[Plain [Str "three"]]]
 ,Header 2 ("indented-code-at-beginning-of-list",[],[]) [Str "Indented",Space,Str "code",Space,Str "at",Space,Str "beginning",Space,Str "of",Space,Str "list"]
 ,BulletList
- [[CodeBlock ("",[],[]) "code\ncode"]]
-,OrderedList (1,Decimal,Period)
- [[CodeBlock ("",[],[]) "code\ncode"]
- ,[CodeBlock ("",[],[]) "code\ncode"]]
-,BulletList
- [[CodeBlock ("",[],[]) "code\ncode"]
- ,[Plain [Str "no",Space,Str "code"]]]
+ [[CodeBlock ("",[],[]) "code\ncode"
+  ,OrderedList (1,Decimal,Period)
+   [[CodeBlock ("",[],[]) "code\ncode"]
+   ,[CodeBlock ("",[],[]) "code\ncode"]]
+  ,BulletList
+   [[CodeBlock ("",[],[]) "code\ncode"]
+   ,[Plain [Str "no",Space,Str "code"]]]]]
 ,Header 2 ("backslash-newline",[],[]) [Str "Backslash",Space,Str "newline"]
 ,Para [Str "hi",LineBreak,Str "there"]
 ,Header 2 ("code-spans",[],[]) [Str "Code",Space,Str "spans"]
diff --git a/test/markdown-reader-more.txt b/test/markdown-reader-more.txt
index 73c9500a0..0480e41cc 100644
--- a/test/markdown-reader-more.txt
+++ b/test/markdown-reader-more.txt
@@ -84,14 +84,14 @@ $PATH 90 $PATH
 
 ## Indented code at beginning of list
 
--       code
-        code
+-     code
+      code
 
-  1.    code
-        code
+  1.     code
+         code
 
-  12345678.    code
-        code
+  12345678.     code
+                code
 
   -     code
         code