From 5527465c778736e91e88fae414ea26558b966724 Mon Sep 17 00:00:00 2001
From: Jesse Rosenthal <jrosenthal@jhu.edu>
Date: Wed, 18 Nov 2015 13:02:57 -0500
Subject: [PATCH] Docx reader: Handle dummy list items.
These come up when people create a list item and then delete the
bullet. It doesn't refer to any real list item, and we used to ignore
it.
We handle it with a DummyListItem type, which, in Docx.hs, is turned
into a normal paragraph with a "ListParagraph" class. If it follow
another list item, it is folded as another paragraph into that item. If
it doesn't, it's just its own (usually indented, and therefore
block-quoted) paragraph.
---
src/Text/Pandoc/Readers/Docx.hs | 4 ++++
src/Text/Pandoc/Readers/Docx/Parse.hs | 17 +++++++++++------
2 files changed, 15 insertions(+), 6 deletions(-)
diff --git a/src/Text/Pandoc/Readers/Docx.hs b/src/Text/Pandoc/Readers/Docx.hs
index ab49bf002..9f1c7af0a 100644
--- a/src/Text/Pandoc/Readers/Docx.hs
+++ b/src/Text/Pandoc/Readers/Docx.hs
@@ -503,6 +503,10 @@ bodyPartToBlocks (ListItem pPr numId lvl levelInfo parparts) = do
]
blks <- bodyPartToBlocks (Paragraph pPr parparts)
return $ divWith ("", ["list-item"], kvs) blks
+bodyPartToBlocks (DummyListItem pPr _ parparts) =
+ let pPr' = pPr {pStyle = "ListParagraph": (pStyle pPr)}
+ in
+ bodyPartToBlocks $ Paragraph pPr' parparts
bodyPartToBlocks (Tbl _ _ _ []) =
return $ para mempty
bodyPartToBlocks (Tbl cap _ look (r:rs)) = do
diff --git a/src/Text/Pandoc/Readers/Docx/Parse.hs b/src/Text/Pandoc/Readers/Docx/Parse.hs
index 02397d658..3723c71a5 100644
--- a/src/Text/Pandoc/Readers/Docx/Parse.hs
+++ b/src/Text/Pandoc/Readers/Docx/Parse.hs
@@ -144,9 +144,6 @@ type Level = (String, String, String, Maybe Integer)
data DocumentLocation = InDocument | InFootnote | InEndnote
deriving (Eq,Show)
--- data RelationshipType = DocumentRel | FootnoteRel | EndnoteRel
--- deriving Show
-
data Relationship = Relationship DocumentLocation RelId Target
deriving Show
@@ -181,6 +178,7 @@ defaultParagraphStyle = ParagraphStyle { pStyle = []
data BodyPart = Paragraph ParagraphStyle [ParPart]
| ListItem ParagraphStyle String String Level [ParPart]
+ | DummyListItem ParagraphStyle String [ParPart]
| Tbl String TblGrid TblLook [Row]
| OMathPara [Exp]
deriving Show
@@ -245,7 +243,6 @@ defaultRunStyle = RunStyle { isBold = Nothing
, rUnderline = Nothing
, rStyle = Nothing}
-
type Target = String
type Anchor = String
type URL = String
@@ -411,6 +408,13 @@ archiveToMedia :: Archive -> Media
archiveToMedia zf =
mapMaybe (getMediaPair zf) (filter filePathIsMedia (filesInArchive zf))
+-- lookupLevel :: String -> String -> Numbering -> Maybe Level
+-- lookupLevel numId ilvl (Numbering _ numbs absNumbs) = do
+-- absNumId <- lookup numId $ map (\(Numb nid absnumid) -> (nid, absnumid)) numbs
+-- lvls <- lookup absNumId $ map (\(AbstractNumb aid ls) -> (aid, ls)) absNumbs
+-- lvl <- lookup ilvl $ map (\l@(i, _, _, _) -> (i, l)) lvls
+-- return lvl
+
lookupLevel :: String -> String -> Numbering -> Maybe Level
lookupLevel numId ilvl (Numbering _ numbs absNumbs) = do
absNumId <- lookup numId $ map (\(Numb nid absnumid) -> (nid, absnumid)) numbs
@@ -418,6 +422,7 @@ lookupLevel numId ilvl (Numbering _ numbs absNumbs) = do
lvl <- lookup ilvl $ map (\l@(i, _, _, _) -> (i, l)) lvls
return lvl
+
numElemToNum :: NameSpaces -> Element -> Maybe Numb
numElemToNum ns element |
qName (elName element) == "num" &&
@@ -569,7 +574,7 @@ elemToBodyPart ns element
num <- asks envNumbering
case lookupLevel numId lvl num of
Just levelInfo -> return $ ListItem parstyle numId lvl levelInfo parparts
- Nothing -> throwError WrongElem
+ Nothing -> return $ DummyListItem parstyle lvl parparts
elemToBodyPart ns element
| isElem ns "w" "p" element = do
sty <- asks envParStyles
@@ -582,7 +587,7 @@ elemToBodyPart ns element
Just levelInfo ->
return $ ListItem parstyle numId lvl levelInfo parparts
Nothing ->
- throwError WrongElem
+ return $ DummyListItem parstyle lvl parparts
Nothing -> return $ Paragraph parstyle parparts
elemToBodyPart ns element
| isElem ns "w" "tbl" element = do