HTML Reader: Use underlineSpan and add test case

hftf · hftf · commit f78c09f1d32b · 2015-07-04T05:57:02.000-04:00
diff --git a/src/Text/Pandoc/Readers/HTML.hs b/src/Text/Pandoc/Readers/HTML.hs
@@ -44,7 +44,7 @@ import Text.Pandoc.Definition
 import qualified Text.Pandoc.Builder as B
 import Text.Pandoc.Builder (Blocks, Inlines, trimInlines, HasMeta(..))
 import Text.Pandoc.Shared ( extractSpaces, renderTags'
-                          , escapeURI, safeRead, mapLeft )
+                          , escapeURI, safeRead, mapLeft, underlineSpan )
 import Text.Pandoc.Options (ReaderOptions(readerParseRaw, readerTrace)
                            , Extension (Ext_epub_html_exts,
                                Ext_native_divs, Ext_native_spans))
@@ -495,6 +495,7 @@ inline = choice
            , pSuperscript
            , pSubscript
            , pStrikeout
+           , pUnderline
            , pLineBreak
            , pLink
            , pImage
@@ -563,6 +564,9 @@ pStrikeout = do
             contents <- mconcat <$> manyTill inline (pCloses "span")
             return $ B.strikeout contents)
 
+pUnderline :: TagParser Inlines
+pUnderline = pInlinesInTags "u" underlineSpan <|> pInlinesInTags "ins" underlineSpan
+
 pLineBreak :: TagParser Inlines
 pLineBreak = do
   pSelfClosing (=="br") (const True)
diff --git a/tests/html-reader.html b/tests/html-reader.html
@@ -310,6 +310,8 @@ <h1>Inline Markup</h1>
 <p>So is <strong><em>this</em></strong> word.</p>
 <p>This is code: <code>&gt;</code>, <code>$</code>, <code>\</code>, <code>\$</code>, <code>&lt;html&gt;</code>.</p>
 <p>This is <span style="font-variant: small-caps;">small caps</span>.</p>
+<p>These are all underlined: <u>foo</u> and <ins>bar</ins>.</p>
+<p>These are all strikethrough: <s>foo</s>, <strike>bar</strike>, and <del>baz</del>.</p>
 <hr />
 <h1>Smart quotes, ellipses, dashes</h1>
 <p>"Hello," said the spider. "'Shelob' is my name."</p>
diff --git a/tests/html-reader.native b/tests/html-reader.native
@@ -194,6 +194,8 @@ Pandoc (Meta {unMeta = fromList [("generator",MetaInlines [Str "pandoc"]),("titl
 ,Para [Str "So",Space,Str "is",Space,Strong [Emph [Str "this"]],Space,Str "word."]
 ,Para [Str "This",Space,Str "is",Space,Str "code:",Space,Code ("",[],[]) ">",Str ",",Space,Code ("",[],[]) "$",Str ",",Space,Code ("",[],[]) "\\",Str ",",Space,Code ("",[],[]) "\\$",Str ",",Space,Code ("",[],[]) "<html>",Str "."]
 ,Para [Str "This",Space,Str "is",Space,SmallCaps [Str "small",Space,Str "caps"],Str "."]
+,Para [Str "These",Space,Str "are",Space,Str "all",Space,Str "underlined:",Space,Span ("",["underline"],[]) [Str "foo"],Space,Str "and",Space,Span ("",["underline"],[]) [Str "bar"],Str "."]
+,Para [Str "These",Space,Str "are",Space,Str "all",Space,Str "strikethrough:",Space,Strikeout [Str "foo"],Str ",",Space,Strikeout [Str "bar"],Str ",",Space,Str "and",Space,Strikeout [Str "baz"],Str "."]
 ,HorizontalRule
 ,Header 1 ("",[],[]) [Str "Smart",Space,Str "quotes,",Space,Str "ellipses,",Space,Str "dashes"]
 ,Para [Str "\"Hello,\"",Space,Str "said",Space,Str "the",Space,Str "spider.",Space,Str "\"'Shelob'",Space,Str "is",Space,Str "my",Space,Str "name.\""]
