aboutsummaryrefslogtreecommitdiff
path: root/src
diff options
context:
space:
mode:
authorAlbert Krewinkel <albert@zeitkraut.de>2017-05-27 15:24:01 +0200
committerAlbert Krewinkel <albert@zeitkraut.de>2017-05-27 15:38:08 +0200
commitbf93c07267bf138f4f4cab7625ff273fa2ac67cd (patch)
tree23214795499a342e9d52543ed98f9f2813f568b3 /src
parent8ec03cfc872d7ff58a2585740cecb4215313251a (diff)
downloadpandoc-bf93c07267bf138f4f4cab7625ff273fa2ac67cd.tar.gz
Org reader: subject full doc tree to headline transformations
Emacs parses org documents into a tree structure, which is then post-processed during exporting. The reader is changed to do the same, turning the document into a single tree of headlines starting at levelĀ 0. Fixes: #3695
Diffstat (limited to 'src')
-rw-r--r--src/Text/Pandoc/Readers/Org/Blocks.hs10
-rw-r--r--src/Text/Pandoc/Readers/Org/DocumentTree.hs33
2 files changed, 37 insertions, 6 deletions
diff --git a/src/Text/Pandoc/Readers/Org/Blocks.hs b/src/Text/Pandoc/Readers/Org/Blocks.hs
index fa2f7fac5..52e990584 100644
--- a/src/Text/Pandoc/Readers/Org/Blocks.hs
+++ b/src/Text/Pandoc/Readers/Org/Blocks.hs
@@ -33,7 +33,7 @@ module Text.Pandoc.Readers.Org.Blocks
) where
import Text.Pandoc.Readers.Org.BlockStarts
-import Text.Pandoc.Readers.Org.DocumentTree (headline, headlineToBlocks)
+import Text.Pandoc.Readers.Org.DocumentTree (documentTree, headlineToBlocks)
import Text.Pandoc.Readers.Org.Inlines
import Text.Pandoc.Readers.Org.Meta (metaExport, metaKey, metaLine)
import Text.Pandoc.Readers.Org.ParserState
@@ -62,11 +62,11 @@ import Data.Monoid ((<>))
-- | Get a list of blocks.
blockList :: PandocMonad m => OrgParser m [Block]
blockList = do
- initialBlocks <- blocks
- headlines <- sequence <$> manyTill (headline blocks inline 1) eof
+ headlines <- documentTree blocks inline
st <- getState
- headlineBlocks <- fmap mconcat . mapM headlineToBlocks $ runF headlines st
- return . B.toList $ (runF initialBlocks st) <> headlineBlocks
+ headlineBlocks <- headlineToBlocks $ runF headlines st
+ -- ignore first headline, it's the document's title
+ return . drop 1 . B.toList $ headlineBlocks
-- | Get the meta information saved in the state.
meta :: Monad m => OrgParser m Meta
diff --git a/src/Text/Pandoc/Readers/Org/DocumentTree.hs b/src/Text/Pandoc/Readers/Org/DocumentTree.hs
index 53ec2ef57..8c2a8482a 100644
--- a/src/Text/Pandoc/Readers/Org/DocumentTree.hs
+++ b/src/Text/Pandoc/Readers/Org/DocumentTree.hs
@@ -28,7 +28,7 @@ Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
Parsers for org-mode headlines and document subtrees
-}
module Text.Pandoc.Readers.Org.DocumentTree
- ( headline
+ ( documentTree
, headlineToBlocks
) where
@@ -43,11 +43,42 @@ import Text.Pandoc.Readers.Org.BlockStarts
import Text.Pandoc.Readers.Org.Parsing
import Text.Pandoc.Readers.Org.ParserState
+import qualified Data.Map as Map
import qualified Text.Pandoc.Builder as B
--
-- Org headers
--
+
+-- | Parse input as org document tree.
+documentTree :: PandocMonad m
+ => OrgParser m (F Blocks)
+ -> OrgParser m (F Inlines)
+ -> OrgParser m (F Headline)
+documentTree blocks inline = do
+ initialBlocks <- blocks
+ headlines <- sequence <$> manyTill (headline blocks inline 1) eof
+ title <- fmap (getTitle . unMeta) . orgStateMeta <$> getState
+ return $ do
+ headlines' <- headlines
+ initialBlocks' <- initialBlocks
+ title' <- title
+ return Headline
+ { headlineLevel = 0
+ , headlineTodoMarker = Nothing
+ , headlineText = B.fromList title'
+ , headlineTags = mempty
+ , headlineProperties = mempty
+ , headlineContents = initialBlocks'
+ , headlineChildren = headlines'
+ }
+ where
+ getTitle :: Map.Map String MetaValue -> [Inline]
+ getTitle metamap =
+ case Map.lookup "title" metamap of
+ Just (MetaInlines inlns) -> inlns
+ _ -> []
+
newtype Tag = Tag { fromTag :: String }
deriving (Show, Eq)