From 7d9b782f73edfc49fbe6f0c3d6ce61328811cbc7 Mon Sep 17 00:00:00 2001
From: Mauro Bieg <mb21@users.noreply.github.com>
Date: Sat, 22 Jul 2017 19:22:56 +0200
Subject: [PATCH] HTML Reader: parse figure and figcaption (#3813)
---
src/Text/Pandoc/Readers/HTML.hs | 20 ++++++++++++++
test/command/html-read-figure.md | 45 ++++++++++++++++++++++++++++++++
2 files changed, 65 insertions(+)
create mode 100644 test/command/html-read-figure.md
diff --git a/src/Text/Pandoc/Readers/HTML.hs b/src/Text/Pandoc/Readers/HTML.hs
index 734973e33..3a0d6eb14 100644
--- a/src/Text/Pandoc/Readers/HTML.hs
+++ b/src/Text/Pandoc/Readers/HTML.hs
@@ -188,6 +188,7 @@ block = do
, pBody
, pDiv
, pPlain
+ , pFigure
, pRawHtmlBlock
]
trace (take 60 $ show $ B.toList res)
@@ -553,6 +554,25 @@ pPara = do
contents <- trimInlines <$> pInTags "p" inline
return $ B.para contents
+pFigure :: PandocMonad m => TagParser m Blocks
+pFigure = do
+ TagOpen _ _ <- pSatisfy (matchTagOpen "figure" [])
+ skipMany pBlank
+ let pImg = pOptInTag "p" pImage <* skipMany pBlank
+ pCapt = option mempty $ pInTags "figcaption" inline <* skipMany pBlank
+ pImgCapt = do
+ img <- pImg
+ cap <- pCapt
+ return (img, cap)
+ pCaptImg = do
+ cap <- pCapt
+ img <- pImg
+ return (img, cap)
+ (imgMany, caption) <- pImgCapt <|> pCaptImg
+ TagClose _ <- pSatisfy (matchTagClose "figure")
+ let (Image attr _ (url, tit)):_ = B.toList imgMany
+ return $ B.para $ B.imageWith attr url ("fig:" ++ tit) caption
+
pCodeBlock :: PandocMonad m => TagParser m Blocks
pCodeBlock = try $ do
TagOpen _ attr' <- pSatisfy (matchTagOpen "pre" [])
diff --git a/test/command/html-read-figure.md b/test/command/html-read-figure.md
new file mode 100644
index 000000000..9c604c706
--- /dev/null
+++ b/test/command/html-read-figure.md
@@ -0,0 +1,45 @@
+```
+% pandoc -f html -t native
+<figure>
+ <img src="foo.png" title="voyage">
+ <figcaption>bar</figcaption>
+</figure>
+^D
+[Para [Image ("",[],[]) [Str "bar"] ("foo.png","fig:voyage")]]
+```
+
+```
+% pandoc -f html -t native
+<figure>
+ <figcaption>bar</figcaption>
+ <img src="foo.png" title="voyage">
+</figure>
+^D
+[Para [Image ("",[],[]) [Str "bar"] ("foo.png","fig:voyage")]]
+```
+
+```
+% pandoc -f html -t native
+<figure>
+ <img src="foo.png" title="voyage">
+</figure>
+^D
+[Para [Image ("",[],[]) [] ("foo.png","fig:voyage")]]
+```
+
+```
+% pandoc -f html -t native
+<figure>
+ <p><img src="foo.png" title="voyage"></p>
+ <figcaption>bar</figcaption>
+</figure>
+^D
+[Para [Image ("",[],[]) [Str "bar"] ("foo.png","fig:voyage")]]
+```
+
+```
+% pandoc -f html -t native
+<figure><img src="foo.png" title="voyage" alt="this is ignored"><figcaption>bar <strong>baz</strong></figcaption></figure>
+^D
+[Para [Image ("",[],[]) [Str "bar",Space,Strong [Str "baz"]] ("foo.png","fig:voyage")]]
+```