Tissevert/pandoc
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions 1

HTML reader: Parse appropriately styled span as SmallCaps.

Browse source
This commit is contained in:
John MacFarlane 2014-08-16 22:57:00 -07:00
parent cb4ae6112e
commit 9d52ecdd42
3 changed files with 8 additions and 1 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

7
src/Text/Pandoc/Readers/HTML.hs
View file

@ -570,7 +570,12 @@ pSpan = try $ do
guardEnabled Ext_native_spans
TagOpen _ attr <- lookAhead $ pSatisfy $ tagOpen (=="span") (const True)
contents <- pInTags "span" inline
return $ B.spanWith (mkAttr attr) contents
let attr' = mkAttr attr
return $ case attr' of
("",[],[("style",s)])
| filter (`notElem` " \t;") s == "font-variant:small-caps" ->
B.smallcaps contents
_ -> B.spanWith (mkAttr attr) contents
pRawHtmlInline :: TagParser Inlines
pRawHtmlInline = do

1
tests/html-reader.html
View file

@ -309,6 +309,7 @@ These should not be escaped: \$ \\ \> \[ \{
<p><strong><em>This is strong and em.</em></strong></p>
<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>
<hr />
<h1>Smart quotes, ellipses, dashes</h1>
<p>"Hello," said the spider. "'Shelob' is my name."</p>

1
tests/html-reader.native
View file

@ -193,6 +193,7 @@ Pandoc (Meta {unMeta = fromList [("generator",MetaInlines [Str "pandoc"]),("titl
,Para [Strong [Emph [Str "This",Space,Str "is",Space,Str "strong",Space,Str "and",Space,Str "em."]]]
,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 "."]
,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.\""]