From de27dbc0f8fd31cc428468e75b129bd751e03f43 Mon Sep 17 00:00:00 2001
From: John MacFarlane <fiddlosopher@gmail.com>
Date: Fri, 11 May 2012 20:08:26 -0700
Subject: DocBook reader.  Handle block elements inside paragraphs.

---
 src/Text/Pandoc/Readers/DocBook.hs | 34 +++++++++++++++++++++++++++++-----
 1 file changed, 29 insertions(+), 5 deletions(-)

(limited to 'src/Text/Pandoc')

diff --git a/src/Text/Pandoc/Readers/DocBook.hs b/src/Text/Pandoc/Readers/DocBook.hs
index 30e926f3d..60d27afc6 100644
--- a/src/Text/Pandoc/Readers/DocBook.hs
+++ b/src/Text/Pandoc/Readers/DocBook.hs
@@ -547,6 +547,20 @@ attrValue attr elt =
 named :: String -> Element -> Bool
 named s e = qName (elName e) == s
 
+isBlockElement :: Content -> Bool
+isBlockElement (Elem e) = qName (elName e) `elem` blocktags
+  where blocktags = ["toc","index","para","formalpara","simpara",
+           "ackno","epigraph","blockquote","bibliography","bibliodiv",
+           "biblioentry","glossee","glosseealso","glossary",
+           "glossdiv","glosslist","chapter","appendix","preface",
+           "sect1","sect2","sect3","sect4","sect5","section",
+           "refsect1","refsect2","refsect3","refsection",
+           "important","caution","note","tip","warning","qandadiv",
+           "question","answer","abstract","itemizedlist","orderedlist",
+           "variablelist","article","book","table","informaltable",
+           "computeroutput","screen","programlisting"]
+isBlockElement _ = False
+
 -- Trim leading and trailing newline characters
 trimNl :: String -> String
 trimNl = reverse . go . reverse . go
@@ -586,15 +600,15 @@ parseBlock (Elem e) =
   case qName (elName e) of
         "toc"   -> return mempty -- skip TOC, since in pandoc it's autogenerated
         "index" -> return mempty -- skip index, since page numbers meaningless
-        "para"  -> para <$> getInlines e
+        "para"  -> parsePara (elContent e)
         "formalpara" -> do
            tit <- case filterChild (named "title") e of
                         Just t  -> (<> str "." <> linebreak) <$> emph
                                       <$> getInlines t
                         Nothing -> return mempty
-           addToStart tit <$> para <$> getInlines e
-        "simpara"  -> para <$> getInlines e
-        "ackno"  -> para <$> getInlines e
+           addToStart tit <$> parsePara (elContent e)
+        "simpara"  -> parsePara (elContent e)
+        "ackno"  -> parsePara (elContent e)
         "epigraph" -> parseBlockquote
         "blockquote" -> parseBlockquote
         "attribution" -> return mempty
@@ -603,7 +617,7 @@ parseBlock (Elem e) =
         "title" -> return mempty -- handled by getTitle or sect
         "bibliography" -> sect 0
         "bibliodiv" -> sect 1
-        "biblioentry" -> para <$> getInlines e
+        "biblioentry" -> parsePara (elContent e)
         "glosssee" -> para . (\ils -> text "See " <> ils <> str ".")
                          <$> getInlines e
         "glossseealso" -> para . (\ils -> text "See also " <> ils <> str ".")
@@ -689,6 +703,16 @@ parseBlock (Elem e) =
         "?xml"  -> return mempty
         _       -> getBlocks e
    where getBlocks e' =  mconcat <$> (mapM parseBlock $ elContent e')
+         parsePara conts = do
+           let (ils,rest) = break isBlockElement conts
+           ils' <- (trimInlines . mconcat) <$> mapM parseInline ils
+           let p = if ils' == mempty then mempty else para ils'
+           case rest of
+                 []     -> return p
+                 (r:rs) -> do
+                    b <- parseBlock r
+                    x <- parsePara rs
+                    return $ p <> b <> x
          codeBlockWithLang classes = do
            let classes' = case attrValue "language" e of
                                 ""   -> classes
-- 
cgit v1.2.3