ODT support for endnotes, reader and writer

MANUAL.txt

@@ -5923,6 +5923,33 @@ they cannot contain multiple paragraphs).  The syntax is as follows:
 
 Inline and regular footnotes may be mixed freely.
 
+## Endnotes
+
+You can use the following convention to specify endnotes: surround a
+footnote with a `Span` of class "endnote", like this:
+
+    Here' and endnote[[^1]]{.endnote}.
+
+    [^1]: This is the endnote text.
+
+When a reader or a writer does not know about this convention, those
+notes are just regular footnotes, just with a transparent `Span` wrapper
+around them. 
+
+### Extension: `endnotes` ###
+
+Enabling this extension tells some readers and writers to use that
+convention to distinguish endnotes from footnotes.
+
+This extension is supported by the docx and odt readers:
+when you convert with `-f docx+endnotes` or `-f odt+endnotes`,
+the endnotes of the docx or odt file will become
+notes embedded in a span with class "endnote".
+
+It is also supported by the docx and odt writers: when you convert
+with `-t docx+endnotes` or `-t odt+endnotes`, all the notes embedded
+in a span of class "endnote" will be endnotes in the docx or odt file.
+
 ## Citation syntax
 
 ### Extension: `citations` ###

data/docx/word/endnotes.xml

@@ -0,0 +1,21 @@
+<?xml version="1.0" encoding="UTF-8"?>
+<w:endnotes xmlns:w="http://schemas.openxmlformats.org/wordprocessingml/2006/main" xmlns:m="http://schemas.openxmlformats.org/officeDocument/2006/math" xmlns:r="http://schemas.openxmlformats.org/officeDocument/2006/relationships" xmlns:o="urn:schemas-microsoft-com:office:office" xmlns:v="urn:schemas-microsoft-com:vml" xmlns:w10="urn:schemas-microsoft-com:office:word" xmlns:a="http://schemas.openxmlformats.org/drawingml/2006/main" xmlns:pic="http://schemas.openxmlformats.org/drawingml/2006/picture" xmlns:wp="http://schemas.openxmlformats.org/drawingml/2006/wordprocessingDrawing">
+  <w:endnote w:type="continuationSeparator" w:id="0"><w:p><w:r><w:continuationSeparator /></w:r></w:p></w:endnote>
+  <w:endnote w:type="separator" w:id="-1"><w:p><w:r><w:separator /></w:r></w:p></w:endnote>
+  <w:endnote w:id="32">
+    <w:p>
+      <w:pPr><w:pStyle w:val="EndnoteText" /></w:pPr>
+      <w:r>
+        <w:rPr>
+          <w:rStyle w:val="EndnoteReference" />
+        </w:rPr>
+      <w:endnoteRef />
+      </w:r>
+      <w:r><w:t>Endnote Text.</w:t></w:r>
+    </w:p>
+    <w:p>
+      <w:pPr><w:pStyle w:val="EndnoteBlockText"/></w:pPr>
+      <w:r><w:t>Endnote Block Text</w:t></w:r>
+    </w:p>
+  </w:endnote>
+</w:endnotes>

data/docx/word/settings.xml

@@ -16,6 +16,10 @@
     <w:footnote w:id="-1" />
     <w:footnote w:id="0" />
   </w:footnotePr>
+  <w:endnotePr>
+    <w:endnote w:id="-1" />
+    <w:endnote w:id="0" />
+  </w:endnotePr>
   <w:rsids>
   </w:rsids>
   <m:mathPr>

pandoc-lua-engine/test/lua/module/pandoc-format.lua

@@ -24,6 +24,7 @@ return {
         'citations',
         'east_asian_line_breaks',
         'empty_paragraphs',
+        'endnotes',
         'gfm_auto_identifiers',
         'native_numbering',
         'styles',

src/Text/Pandoc/Extensions.hs

@@ -68,6 +68,7 @@ data Extension =
     | Ext_element_citations   -- ^ Use element-citation elements for JATS citations
     | Ext_emoji               -- ^ Support emoji like :smile:
     | Ext_empty_paragraphs -- ^ Allow empty paragraphs
+    | Ext_endnotes            -- ^ Endnotes support when footnotes are embedded in a Span.endnote
     | Ext_epub_html_exts      -- ^ Recognise the EPUB extended version of HTML
     | Ext_escaped_line_breaks     -- ^ Treat a backslash at EOL as linebreak
     | Ext_example_lists       -- ^ Markdown-style numbered examples
@@ -532,12 +533,14 @@ getAllExtensions f = universalExtensions <> getAll f
     [ Ext_raw_markdown ]
   getAll "docx"            = autoIdExtensions <> extensionsFromList
     [ Ext_empty_paragraphs
+    , Ext_endnotes
     , Ext_native_numbering
     , Ext_styles
     , Ext_citations
     ]
   getAll "opendocument"    = extensionsFromList
     [ Ext_empty_paragraphs
+    , Ext_endnotes
     , Ext_native_numbering
     , Ext_xrefs_name
     , Ext_xrefs_number

src/Text/Pandoc/Readers/Docx.hs

@@ -52,7 +52,9 @@ implemented, [-] means partially implemented):
   - [X] Link (links to an arbitrary bookmark create a span with the target as
         id and "anchor" class)
   - [X] Image
-  - [X] Note (Footnotes and Endnotes are silently combined.)
+  - [X] Note (Footnotes and Endnotes are silently combined,
+              unless Ext_endnotes is enabled: in that case a Note is embedded
+              in a Span with class "endnote")
 -}
 
 module Text.Pandoc.Readers.Docx
@@ -342,7 +344,12 @@ runToInlines (Run rs runElems)
       transform <- runStyleToTransform rPr
       return $ transform ils
 runToInlines (Footnote bps) = note . smushBlocks <$> mapM bodyPartToBlocks bps
-runToInlines (Endnote bps) = note . smushBlocks <$> mapM bodyPartToBlocks bps
+runToInlines (Endnote bps) = do
+  isEndnotesExtEnabled <- asks (isEnabled Ext_endnotes . docxOptions)
+  noteInlines <- note . smushBlocks <$> mapM bodyPartToBlocks bps
+  return $ if isEndnotesExtEnabled
+    then spanWith ("", ["endnote"], []) noteInlines
+    else noteInlines
 runToInlines (InlineDrawing fp title alt bs ext) = do
   (lift . lift) $ P.insertMedia fp Nothing bs
   return $ imageWith (extentToAttr ext) (T.pack fp) title $ text alt

src/Text/Pandoc/Readers/ODT.hs

@@ -70,20 +70,20 @@ makeFigure x = x
 readODT' :: ReaderOptions
          -> B.ByteString
          -> Either PandocError (Pandoc, MediaBag)
-readODT' _ bytes = bytesToODT bytes-- of
+readODT' opts bytes = bytesToODT opts bytes-- of
 --                    Right (pandoc, mediaBag) -> Right (pandoc , mediaBag)
 --                    Left  err                -> Left err
 
 --
-bytesToODT :: B.ByteString -> Either PandocError (Pandoc, MediaBag)
-bytesToODT bytes = case toArchiveOrFail bytes of
-  Right archive -> archiveToODT archive
+bytesToODT :: ReaderOptions -> B.ByteString -> Either PandocError (Pandoc, MediaBag)
+bytesToODT opts bytes = case toArchiveOrFail bytes of
+  Right archive -> archiveToODT opts archive
   Left err      -> Left $ PandocParseError
                         $ "Could not unzip ODT: " <> T.pack err
 
 --
-archiveToODT :: Archive -> Either PandocError (Pandoc, MediaBag)
-archiveToODT archive = do
+archiveToODT :: ReaderOptions -> Archive -> Either PandocError (Pandoc, MediaBag)
+archiveToODT opts archive = do
   let onFailure msg Nothing = Left $ PandocParseError msg
       onFailure _   (Just x) = Right x
   contentEntry <- onFailure "Could not find content.xml"
@@ -101,7 +101,7 @@ archiveToODT archive = do
         let (dir, name) = splitFileName fp
         in  (dir == "Pictures/") || (dir /= "./" && name == "content.xml")
   let media = filteredFilesFromArchive archive filePathIsODTMedia
-  let startState = readerState styles media
+  let startState = readerState opts styles media
   either (\_ -> Left $ PandocParseError "Could not convert opendocument") Right
     (runConverter' read_body startState contentElem)
 

src/Text/Pandoc/Readers/ODT/ContentReader.hs

@@ -44,6 +44,7 @@ import Text.Pandoc.Builder hiding (underline)
 import Text.Pandoc.MediaBag (MediaBag, insertMedia)
 import Text.Pandoc.Shared
 import Text.Pandoc.Extensions (extensionsFromList, Extension(..))
+import Text.Pandoc.Options (isEnabled, ReaderOptions)
 import qualified Text.Pandoc.UTF8 as UTF8
 
 import Text.Pandoc.Readers.Docx.Combine (combineBlocks)
@@ -96,11 +97,13 @@ data ReaderState
                  , envMedia         :: Media
                    -- | Hold binary resources used in the document
                  , odtMediaBag      :: MediaBag
+                   -- | Read endnotes as Note inside a Span of class "endnote"
+                 , readEndnotes     :: Bool
                  }
   deriving ( Show )
 
-readerState :: Styles -> Media -> ReaderState
-readerState styles media = ReaderState styles [] 0 M.empty Nothing M.empty media mempty
+readerState :: ReaderOptions -> Styles -> Media -> ReaderState
+readerState opts styles media = ReaderState styles [] 0 M.empty Nothing M.empty media mempty (isEnabled Ext_endnotes opts)
 
 --
 pushStyle'  :: Style -> ReaderState -> ReaderState
@@ -788,13 +791,23 @@ fixRelativeLink uri =
           _ -> uri
 
 -------------------------
--- Footnotes
+-- Footnotes and Endnotes
 -------------------------
 
+endnote :: Blocks -> Inlines
+endnote blocks = spanWith ("", ["endnote"], []) $ note blocks
+
+whichnote :: Bool -> T.Text -> (Blocks -> Inlines)
+whichnote True "endnote" = endnote
+whichnote _ _ = note
+
 read_note        :: InlineMatcher
 read_note         = matchingElement NsText "note"
-                    $ liftA note
+                    $ liftA3 whichnote readingEndnotes noteClass
                     $ matchChildContent' [ read_note_body ]
+                  where
+                    noteClass = findAttrTextWithDefault NsText "note-class" ""
+                    readingEndnotes = getExtraState >>^ readEndnotes
 
 read_note_body   :: BlockMatcher
 read_note_body    = matchingElement NsText "note-body"

src/Text/Pandoc/Writers/Docx.hs

@@ -105,7 +105,7 @@ writeDocx opts doc = do
         }
 
   -- Phase 5: Relationship extraction
-  (baserels, headers, footers, newMaxRelId) <- extractRelationships refArchive distArchive
+  (baserels, headers, footers, newMaxRelId) <- extractRelationships opts refArchive distArchive
 
   let initialSt = defaultWriterState {
           stStyleMaps  = styleMaps
@@ -123,7 +123,7 @@ writeDocx opts doc = do
                             [ mknode "w:numRestart" [("w:val","eachSect")] () ]
                           ]
 
-  ((contents, footnotes, comments), st) <- runStateT
+  ((contents, footnotes, endnotes, comments), st) <- runStateT
                          (runReaderT
                           (writeOpenXML opts{ writerWrapText = WrapNone }
                                         doc')
@@ -137,12 +137,14 @@ writeDocx opts doc = do
   -- because Word sometimes changes these files when a reference.docx is modified,
   -- e.g. deleting the reference to footnotes.xml or removing default entries
   -- for image content types.
-  let contentTypesEntry = mkContentTypesEntry epochtime imgs headers footers refArchive
+  let contentTypesEntry = mkContentTypesEntry epochtime imgs headers footers endnotes refArchive
   let relEntry = mkDocumentRelsEntry epochtime baserels imgs (stExternalLinks st)
   let contentEntry = toEntry "word/document.xml" epochtime
                        (BL.fromStrict $ UTF8.fromText contents)
   let footnotesEntry = mkFootnotesEntry epochtime footnotes
   let footnoteRelEntry = mkFootnoteRelsEntry epochtime (stExternalLinks st)
+  let endnotesEntry = mkEndnotesEntry epochtime endnotes
+  let endnoteRelEntry = mkEndnoteRelsEntry epochtime (stExternalLinks st)
   let commentsEntry = mkCommentsEntry epochtime comments
   let styleEntry = mkStylesEntry epochtime styledoc styleMaps st opts
   numEntry <- mkNumberingEntry refArchive distArchive epochtime (stLists st)
@@ -165,10 +167,12 @@ writeDocx opts doc = do
   let archive = foldr addEntryToArchive emptyArchive $
                   contentTypesEntry : relsEntry : contentEntry : relEntry :
                   footnoteRelEntry : numEntry : styleEntry : footnotesEntry :
-                  commentsEntry :
-                  docPropsEntry : customPropsEntry :
-                  settingsEntry :
-                  imageEntries ++ refEntries
+                  commentsEntry : docPropsEntry : customPropsEntry : settingsEntry :
+                  imageEntries
+                  ++ refEntries
+                  ++ if (isEnabled Ext_endnotes opts)
+                      then [endnoteRelEntry, endnotesEntry]
+                      else []
   return $ fromArchive archive
 
 newParaPropToOpenXml :: ParaStyleName -> Element
@@ -525,9 +529,9 @@ extractPageLayout refArchive distArchive = do
 
 -- | Parse and augment relationships from reference.docx
 extractRelationships :: PandocMonad m
-                     => Archive -> Archive
+                     => WriterOptions -> Archive -> Archive
                      -> m ([Element], [Element], [Element], Int)
-extractRelationships refArchive distArchive = do
+extractRelationships opts refArchive distArchive = do
   let isImageNode e = findAttr (QName "Type" Nothing Nothing) e == Just "http://schemas.openxmlformats.org/officeDocument/2006/relationships/image"
   let isHeaderNode e = findAttr (QName "Type" Nothing Nothing) e == Just "http://schemas.openxmlformats.org/officeDocument/2006/relationships/header"
   let isFooterNode e = findAttr (QName "Type" Nothing Nothing) e == Just "http://schemas.openxmlformats.org/officeDocument/2006/relationships/footer"
@@ -554,7 +558,7 @@ extractRelationships refArchive distArchive = do
                             ,("Target",target')] () : rels)
           _ -> (maxId, rels)
 
-  let (newMaxRelId, baserels) = foldr addBaseRel (maxRelId, parsedRels)
+  let (newMaxRelId, baserels) = foldr addBaseRel (maxRelId, parsedRels) $
                     [("http://schemas.openxmlformats.org/officeDocument/2006/relationships/numbering",
                       "numbering.xml")
                     ,("http://schemas.openxmlformats.org/officeDocument/2006/relationships/styles",
@@ -571,7 +575,8 @@ extractRelationships refArchive distArchive = do
                       "footnotes.xml")
                     ,("http://schemas.openxmlformats.org/officeDocument/2006/relationships/comments",
                       "comments.xml")
-                    ]
+                    ] ++ [("http://schemas.openxmlformats.org/officeDocument/2006/relationships/endnotes",
+                                                    "endnotes.xml") | isEnabled Ext_endnotes opts]
 
   return (baserels, headers, footers, newMaxRelId)
 
@@ -595,6 +600,26 @@ mkFootnoteRelsEntry epochtime externalLinks =
            [("xmlns","http://schemas.openxmlformats.org/package/2006/relationships")]
            linkrels
 
+-- | Create endnotes XML entry
+mkEndnotesEntry :: Integer -> [Element] -> Entry
+mkEndnotesEntry epochtime endnotes =
+  let notes = mknode "w:endnotes" stdAttributes endnotes
+  in toEntry "word/endnotes.xml" epochtime $ renderXml notes
+
+-- | Create endnote relationships entry
+mkEndnoteRelsEntry :: Integer -> M.Map Text Text -> Entry
+mkEndnoteRelsEntry epochtime externalLinks =
+  let linkrels = map toLinkRel $ M.toList externalLinks
+      toLinkRel (src, ident) = mknode "Relationship"
+        [("Type","http://schemas.openxmlformats.org/officeDocument/2006/relationships/hyperlink")
+        ,("Id",ident)
+        ,("Target",src)
+        ,("TargetMode","External")] ()
+  in toEntry "word/_rels/endnotes.xml.rels" epochtime
+       $ renderXml $ mknode "Relationships"
+           [("xmlns","http://schemas.openxmlformats.org/package/2006/relationships")]
+           linkrels
+
 -- | Create comments XML entry
 mkCommentsEntry :: Integer -> [Element] -> Entry
 mkCommentsEntry epochtime comments =
@@ -627,9 +652,10 @@ mkContentTypesEntry :: Integer
                     -> [(String, String, Maybe MimeType, B.ByteString)]  -- imgs
                     -> [Element]  -- headers
                     -> [Element]  -- footers
+                    -> [Element]  -- endnotes
                     -> Archive    -- refArchive
                     -> Entry
-mkContentTypesEntry epochtime imgs headers footers refArchive =
+mkContentTypesEntry epochtime imgs headers footers endnotes refArchive =
   let mkOverrideNode (part', contentType') = mknode "Override"
            [("PartName", T.pack part')
            ,("ContentType", contentType')] ()
@@ -638,6 +664,11 @@ mkContentTypesEntry epochtime imgs headers footers refArchive =
                           fromMaybe "application/octet-stream" mbMimeType)
       mkMediaOverride imgpath =
           mkOverrideNode ("/" <> imgpath, getMimeTypeDef imgpath)
+      endnotesOverride = if null endnotes
+          then [("/word/endnotes.xml",
+                    "application/vnd.openxmlformats-officedocument.wordprocessingml.endnotes+xml")]
+          else []
+
       overrides = map mkOverrideNode (
                   [("/word/webSettings.xml",
                     "application/vnd.openxmlformats-officedocument.wordprocessingml.webSettings+xml")
@@ -664,6 +695,7 @@ mkContentTypesEntry epochtime imgs headers footers refArchive =
                   ,("/word/footnotes.xml",
                     "application/vnd.openxmlformats-officedocument.wordprocessingml.footnotes+xml")
                   ] ++
+                  endnotesOverride ++
                   map (\x -> (maybe "" (T.unpack . ("/word/" <>)) (extractTarget x),
                        "application/vnd.openxmlformats-officedocument.wordprocessingml.header+xml")) headers ++
                   map (\x -> (maybe "" (T.unpack . ("/word/" <>)) (extractTarget x),
@@ -723,6 +755,7 @@ mkStylesEntry epochtime styledoc styleMaps st opts =
         (\sty -> not $ hasStyleName sty $ smCharStyle styleMaps)
         (Set.toList $ stDynamicTextProps st)
 
+      -- TODO: add styles for endnotes, when Ext_endnotes is enabled
       newstyles = map newParaPropToOpenXml newDynamicParaProps ++
                   map newTextPropToOpenXml newDynamicTextProps ++
                   (case writerHighlightMethod opts of
@@ -836,7 +869,8 @@ collectReferenceEntries refArchive distArchive headers footers = do
                        , "word/_rels/" `isPrefixOf` eRelativePath e
                        , ".xml.rels" `isSuffixOf` eRelativePath e
                        , eRelativePath e /= "word/_rels/document.xml.rels"
-                       , eRelativePath e /= "word/_rels/footnotes.xml.rels" ]
+                       , eRelativePath e /= "word/_rels/footnotes.xml.rels"
+                       , eRelativePath e /= "word/_rels/endnotes.xml.rels" ]
   let otherMediaEntries = [ e | e <- zEntries refArchive
                           , "word/media/" `isPrefixOf` eRelativePath e ]
   return $ docPropsAppEntry : themeEntry : fontTableEntry : webSettingsEntry