Support complex figures. [API change]

Thanks and credit go to Aner Lucero, who laid the groundwork for this
feature in the 2021 GSoC project. He contributed many changes, including
modifications to the readers for HTML, JATS, and LaTeX, and to the HTML
and JATS writers.

Shared (Albert Krewinkel):

- The new function `figureDiv`, exported from `Text.Pandoc.Shared`,
  offers a standardized way to convert a figure into a Div element.

Readers (Aner Lucero):

- HTML reader: `<figure>` elements are parsed as figures, with the
  caption taken from the respective `<figcaption>` elements.

- JATS reader: The `<fig>` and `<caption>` elements are parsed into
  figure elements, even if the contents is more complex.

- LaTeX reader: support for figures with non-image contents and for
  subfigures.

- Markdown reader: paragraphs containing just an image are treated as
  figures if the `implicit_figures` extension is enabled. The identifier
  is used as the figure's identifier and the image description is also
  used as figure caption; all other attributes are treated as belonging
  to the image.

Writers (Aner Lucero, Albert Krewinkel):

- DokuWiki, Haddock, Jira, Man, MediaWiki, Ms, Muse, PPTX, RTF, TEI,
  ZimWiki writers: Figures are rendered like Div elements.

- Asciidoc writer: The figure contents is unwrapped; each image in the
  the figure becomes a separate figure.

- Classic custom writers: Figures are passed to the global function
  `Figure(caption, contents, attr)`, where `caption` and `contents` are
  strings and `attr` is a table of key-value pairs.

- ConTeXt writer: Figures are wrapped in a "placefigure" environment
  with `\startplacefigure`/`\endplacefigure`, adding the features
  caption and listing title as properties. Subfigures are place in a
  single row with the `\startfloatcombination` environment.

- DocBook writer: Uses `mediaobject` elements, unless the figure contains
  subfigures or tables, in which case the figure content is unwrapped.

- Docx writer: figures with multiple content blocks are rendered as
  tables with style `FigureTable`; like before, single-image figures are
  still output as paragraphs with style `Figure` or `Captioned Figure`,
  depending on whether a caption is attached.

- DokuWiki writer: Caption and "alt-text" are no longer combined. The
  alt text of a figure will now be lost in the conversion.

- FB2 writer: The figure caption is added as alt text to the images in
  the figure; pre-existing alt texts are kept.

- ICML writer: Only single-image figures are supported. The contents of
  figures with additional elements gets unwrapped.

- HTML writer: the alt text is no longer constructed from the caption,
  as was the case with implicit figures. This reduces duplication, but
  comes at the risk of images that are missing alt texts. Authors should
  take care to provide alt texts for all images.

  Some readers, most notably the Markdown reader with the
  `implicit_figures` extension, add a caption that's identical to the
  image description. The writer checks for this and adds an
  `aria-hidden` attribute to the `<figcaption>` element in that case.

- JATS writer: The `<fig>` and `<caption>` elements are used write
  figures.

- LaTeX writer: complex figures, e.g. with non-image contents and
  subfigures, are supported. The `subfigure` template variable is set if
  the document contains subfigures, triggering the conditional loading
  of the *subcaption* package. Contants of figures that contain tables
  are become unwrapped, as longtable environments are not allowed within
  figures.

- Markdown writer: figures are output as implicit figures if possible,
  via HTML if the `raw_html` extension is enabled, and as Div elements
  otherwise.

- OpenDocument writer: A separate paragraph is generated for each block
  element in a figure, each with style `FigureWithCaption`. Behavior for
  single-image figures therefore remains unchanged.

- Org writer: Only the first element in a figure is given a caption;
  additional block elements in the figure are appended without any
  caption being added.

- RST writer: Single-image figures are supported as before; the contents
  of more complex images become nested in a container of type `float`.

- Texinfo writer: Figures are rendered as float with type `figure`.

- Textile writer: Figures are rendered with the help of HTML elements.

- XWiki: Figures are placed in a group.

Co-authored-by: Aner Lucero <[email protected]>

Author: tarleb

cabal.project

@@ -16,3 +16,13 @@ source-repository-package
   type: git
   location: https://github.com/jgm/texmath
   tag: 1a77db688bd3285228299e5aeefc93d6c0d8c0b9
+
+source-repository-package
+  type: git
+  location: https://github.com/tarleb/pandoc-types
+  tag: f84b7359765a2798f22efe4e9457538cda7a8d4a
+
+source-repository-package
+  type: git
+  location: https://github.com/pandoc/pandoc-lua-marshal
+  tag: a2a97e2af78326ea7841101d4ef56e74426b66c4

data/templates/default.latex

@@ -293,6 +293,9 @@ $if(numbersections)$
 $else$
 \setcounter{secnumdepth}{-\maxdimen} % remove section numbering
 $endif$
+$if(subfigure)$
+\usepackage{subcaption}
+$endif$
 $if(beamer)$
 $else$
 $if(block-headings)$

pandoc-lua-engine/src/Text/Pandoc/Lua/Writer/Classic.hs

@@ -157,6 +157,12 @@ blockToCustom (CodeBlock attr str) =
 blockToCustom (BlockQuote blocks) =
   invoke "BlockQuote" (Stringify blocks)
 
+blockToCustom (Figure attr (Caption _ cbody) content) =
+  invoke "Figure"
+    (Stringify cbody)
+    (Stringify content)
+    (attrToMap attr)
+
 blockToCustom (Table _ blkCapt specs thead tbody tfoot) =
   let (capt, aligns, widths, headers, rows) = toLegacyTable blkCapt specs thead tbody tfoot
       aligns' = map show aligns

pandoc-lua-engine/test/sample.lua

@@ -295,6 +295,12 @@ function CaptionedImage(src, tit, caption, attr)
   end
 end
 
+function Figure(caption, contents, attr)
+  return '<figure' .. attributes(attr) .. '>\n' .. contents ..
+    '\n<figcaption>' .. caption .. '</figcaption>\n' ..
+    '</figure>'
+end
+
 -- Caption is a string, aligns is an array of strings,
 -- widths is an array of floats, headers is an array of
 -- strings, rows is an array of arrays of strings.

pandoc-lua-engine/test/writer.custom

@@ -737,7 +737,8 @@ So is &lsquo;pine.&rsquo;</p>
 <p>From &ldquo;Voyage dans la Lune&rdquo; by Georges Melies (1902):</p>
 
 <figure>
-<img src="lalune.jpg" id="" alt="lalune"/><figcaption>lalune</figcaption>
+<img src="lalune.jpg" title="Voyage dans la Lune"/>
+<figcaption>lalune</figcaption>
 </figure>
 
 <p>Here is a movie <img src="movie.jpg" title=""/> icon.</p>

src/Text/Pandoc/Readers/HTML.hs

@@ -25,7 +25,7 @@ module Text.Pandoc.Readers.HTML ( readHtml
                                 ) where
 
 import Control.Applicative ((<|>))
-import Control.Monad (guard, msum, mzero, unless, void)
+import Control.Monad (guard, mzero, unless, void)
 import Control.Monad.Except (throwError, catchError)
 import Control.Monad.Reader (ask, asks, lift, local, runReaderT)
 import Data.Text.Encoding.Base64 (encodeBase64)
@@ -36,6 +36,7 @@ import Data.List.Split (splitWhen)
 import Data.List (foldl')
 import qualified Data.Map as M
 import Data.Maybe (fromMaybe, isJust, isNothing)
+import Data.Either (partitionEithers)
 import Data.Monoid (First (..))
 import qualified Data.Set as Set
 import Data.Text (Text)
@@ -63,8 +64,8 @@ import Text.Pandoc.Options (
     extensionEnabled)
 import Text.Pandoc.Parsing hiding ((<|>))
 import Text.Pandoc.Shared (
-    addMetaField, blocksToInlines', extractSpaces,
-    htmlSpanLikeElements, renderTags', safeRead, tshow, formatCode)
+    addMetaField, extractSpaces, htmlSpanLikeElements, renderTags',
+    safeRead, tshow, formatCode)
 import Text.Pandoc.URI (escapeURI)
 import Text.Pandoc.Walk
 import Text.TeXMath (readMathML, writeTeX)
@@ -581,24 +582,15 @@ pPara = do
     <|> return (B.para contents)
 
 pFigure :: PandocMonad m => TagParser m Blocks
-pFigure = try $ do
-  TagOpen _ _ <- pSatisfy (matchTagOpen "figure" [])
-  skipMany pBlank
-  let pImg  = (\x -> (Just x, Nothing)) <$>
-               (pInTag TagsOmittable "p" pImage <* skipMany pBlank)
-      pCapt = (\x -> (Nothing, Just x)) <$> do
-                bs <- pInTags "figcaption" block
-                return $ blocksToInlines' $ B.toList bs
-      pSkip = (Nothing, Nothing) <$ pSatisfy (not . matchTagClose "figure")
-  res <- many (pImg <|> pCapt <|> pSkip)
-  let mbimg = msum $ map fst res
-  let mbcap = msum $ map snd res
-  TagClose _ <- pSatisfy (matchTagClose "figure")
-  let caption = fromMaybe mempty mbcap
-  case B.toList <$> mbimg of
-       Just [Image attr _ (url, tit)] ->
-         return $ B.simpleFigureWith attr caption url tit
-       _ -> mzero
+pFigure = do
+  TagOpen tag attrList <- pSatisfy $ matchTagOpen "figure" []
+  let parser = Left <$> pInTags "figcaption" block <|>
+             (Right <$> block)
+  (captions, rest) <- partitionEithers <$> manyTill parser (pCloses tag <|> eof)
+  -- Concatenate all captions together
+  return $ B.figureWith (toAttr attrList)
+                        (B.simpleCaption (mconcat captions))
+                        (mconcat rest)
 
 pCodeBlock :: PandocMonad m => TagParser m Blocks
 pCodeBlock = try $ do

src/Text/Pandoc/Readers/JATS.hs

@@ -38,7 +38,6 @@ import Text.TeXMath (readMathML, writeTeX)
 import qualified Data.Set as S (fromList, member)
 import Data.Set ((\\))
 import Text.Pandoc.Sources (ToSources(..), sourcesToText)
-import qualified Data.Foldable as DF
 
 type JATS m = StateT JATSState m
 
@@ -232,29 +231,17 @@ parseBlock (Elem e) =
                      terms' <- mapM getInlines terms
                      items' <- mapM getBlocks items
                      return (mconcat $ intersperse (str "; ") terms', items')
-         parseFigure =
-           -- if a simple caption and single graphic, we emit a standard
-           -- implicit figure.  otherwise, we emit a div with the contents
-           case filterChildren (named "graphic") e of
-                  [g] -> do
-                         capt <- case filterChild (named "caption") e of
-                                        Just t  -> mconcat .
-                                          intersperse linebreak <$>
-                                          mapM getInlines
-                                          (filterChildren (const True) t)
-                                        Nothing -> return mempty
-
-                         let figAttributes = DF.toList $
-                              ("alt", ) . strContent <$>
-                              filterChild (named "alt-text") e
-
-                         return $ simpleFigureWith
-                          (attrValue "id" e, [], figAttributes)
-                          capt
-                          (attrValue "href" g)
-                          (attrValue "title" g)
-
-                  _   -> divWith (attrValue "id" e, ["fig"], []) <$> getBlocks e
+         parseFigure = do
+           capt <- case filterChild (named "caption") e of
+                     Just t  -> mconcat . intersperse linebreak <$>
+                                mapM getInlines (filterChildren (const True) t)
+                     Nothing -> return mempty
+           contents <- getBlocks e
+
+           return $ figureWith
+             (attrValue "id" e, [], [])
+             (simpleCaption $ plain capt)
+             contents
          parseFootnoteGroup = do
            forM_ (filterChildren (named "fn") e) $ \fn -> do
              let id' = attrValue "id" fn

src/Text/Pandoc/Readers/LaTeX.hs

@@ -33,6 +33,7 @@ import Data.Maybe (fromMaybe, maybeToList)
 import qualified Data.Set as Set
 import Data.Text (Text)
 import qualified Data.Text as T
+import Data.Either (partitionEithers)
 import Skylighting (defaultSyntaxMap)
 import System.FilePath (addExtension, replaceExtension, takeExtension)
 import Text.Collate.Lang (renderLang)
@@ -1011,8 +1012,8 @@ environments = M.union (tableEnvironments blocks inline) $
    , ("letter", env "letter" letterContents)
    , ("minipage", env "minipage" $
           skipopts *> spaces *> optional braced *> spaces *> blocks)
-   , ("figure", env "figure" $ skipopts *> figure)
-   , ("subfigure", env "subfigure" $ skipopts *> tok *> figure)
+   , ("figure", env "figure" $ skipopts *> figure')
+   , ("subfigure", env "subfigure" $ skipopts *> tok *> figure')
    , ("center", divWith ("", ["center"], []) <$> env "center" blocks)
    , ("quote", blockQuote <$> env "quote" blocks)
    , ("quotation", blockQuote <$> env "quotation" blocks)
@@ -1164,37 +1165,33 @@ letterContents = do
                   _ -> mempty
   return $ addr <> bs -- sig added by \closing
 
-figure :: PandocMonad m => LP m Blocks
-figure = try $ do
+figure' :: PandocMonad m => LP m Blocks
+figure' = try $ do
   resetCaption
-  blocks >>= addImageCaption
-
-addImageCaption :: PandocMonad m => Blocks -> LP m Blocks
-addImageCaption = walkM go
-  where go p@(Para [Image attr@(_, cls, kvs) _ (src, tit)])
-            | not ("fig:" `T.isPrefixOf` tit) = do
-          st <- getState
-          case sCaption st of
-            Nothing -> return p
-            Just (Caption _mbshort bs) -> do
-              let mblabel = sLastLabel st
-              let attr' = case mblabel of
-                            Just lab -> (lab, cls, kvs)
-                            Nothing  -> attr
-              case attr' of
-                   ("", _, _)    -> return ()
-                   (ident, _, _) -> do
-                      num <- getNextNumber sLastFigureNum
-                      setState
-                        st{ sLastFigureNum = num
-                          , sLabels = M.insert ident
-                                     [Str (renderDottedNum num)] (sLabels st) }
-
-              return $ SimpleFigure attr'
-                       (maybe id removeLabel mblabel
-                         (blocksToInlines bs))
-                       (src, tit)
-        go x = return x
+  innerContent <- many $ try (Left <$> label) <|> (Right <$> block)
+  let content = walk go $ mconcat $ snd $ partitionEithers innerContent
+  st <- getState
+  let caption' = case sCaption st of
+                   Nothing   -> B.emptyCaption
+                   Just capt -> capt
+  let mblabel  = sLastLabel st
+  let attr     = case mblabel of
+                   Just lab -> (lab, [], [])
+                   Nothing  -> nullAttr
+  case mblabel of
+    Nothing   -> pure ()
+    Just lab  -> do
+      num <- getNextNumber sLastFigureNum
+      setState
+        st { sLastFigureNum = num
+           , sLabels = M.insert lab [Str (renderDottedNum num)] (sLabels st)
+           }
+  return $ B.figureWith attr caption' content
+
+  where
+  -- Remove the `Image` caption b.c. it's on the `Figure`
+  go (Para [Image attr _ target]) = Plain [Image attr [] target]
+  go x = x
 
 coloredBlock :: PandocMonad m => Text -> LP m Blocks
 coloredBlock stylename = try $ do

src/Text/Pandoc/Readers/LaTeX/Math.hs

@@ -214,7 +214,8 @@ addQed bs =
   qedSign = B.str "\xa0\x25FB"
 
 italicize :: Block -> Block
-italicize x@(Para [Image{}]) = x -- see #6925
+italicize x@(Para [Image{}])  = x -- see #6925
+italicize x@(Plain [Image{}]) = x -- ditto
 italicize (Para ils) = Para [Emph ils]
 italicize (Plain ils) = Plain [Emph ils]
 italicize x = x

src/Text/Pandoc/Readers/Markdown.hs

@@ -1046,7 +1046,7 @@ para = try $ do
           [Image attr figCaption (src, tit)]
             | extensionEnabled Ext_implicit_figures exts
             , not (null figCaption) -> do
-                B.simpleFigureWith attr (B.fromList figCaption) src tit
+                implicitFigure attr (B.fromList figCaption) src tit
 
           _ -> constr inlns
 
@@ -1077,6 +1077,17 @@ para = try $ do
 plain :: PandocMonad m => MarkdownParser m (F Blocks)
 plain = fmap B.plain . trimInlinesF <$> inlines1
 
+implicitFigure :: Attr -> Inlines -> Text -> Text -> Blocks
+implicitFigure (ident, classes, attribs) capt url title =
+  let alt = case "alt" `lookup` attribs of
+              Just alt'       -> B.text alt'
+              _               -> capt
+      attribs' = filter ((/= "alt") . fst) attribs
+      figattr = (ident, mempty, mempty)
+      caption = B.simpleCaption $ B.plain capt
+      figbody = B.plain $ B.imageWith ("", classes, attribs') url title alt
+  in B.figureWith figattr caption figbody
+
 --
 -- raw html
 --