From 87c66d4623200a48eab6e550f73372aaf3608289 Mon Sep 17 00:00:00 2001
From: fiddlosopher <fiddlosopher@788f1e2b-df1e-0410-8736-df70ead52e1b>
Date: Fri, 1 Jan 2010 04:11:48 +0000
Subject: [PATCH] Finished converting HTML writer to use unicode instead of
entities.
git-svn-id: https://pandoc.googlecode.com/svn/trunk@1767 788f1e2b-df1e-0410-8736-df70ead52e1b
---
src/Text/Pandoc/Writers/HTML.hs | 27 +++++++-------
tests/writer.html | 66 ++++++++++++++++-----------------
2 files changed, 46 insertions(+), 47 deletions(-)
diff --git a/src/Text/Pandoc/Writers/HTML.hs b/src/Text/Pandoc/Writers/HTML.hs
index 9a093ad96..c80a3b32f 100644
--- a/src/Text/Pandoc/Writers/HTML.hs
+++ b/src/Text/Pandoc/Writers/HTML.hs
@@ -60,9 +60,8 @@ renderFragment opts = if writerWrapText opts
then renderHtmlFragment
else showHtmlFragment
--- | Slightly modified version of Text.XHtml's stringToHtml.
--- Only uses numerical entities for 0xff and greater.
--- Adds .
+-- | Modified version of Text.XHtml's stringToHtml.
+-- Use unicode characters wherever possible.
stringToHtml :: String -> Html
stringToHtml = primHtml . concatMap fixChar
where
@@ -248,7 +247,7 @@ obfuscateLink opts txt s =
linkText ++ "+'<\\/'+'a'+'>');\n// -->\n")) +++
noscript (primHtml $ obfuscateString altText)
_ -> error $ "Unknown obfuscation method: " ++ show meth
- _ -> anchor ! [href s] $ primHtml txt -- malformed email
+ _ -> anchor ! [href s] $ stringToHtml txt -- malformed email
-- | Obfuscate character as entity.
obfuscateChar :: Char -> String
@@ -410,11 +409,11 @@ inlineToHtml opts inline =
case inline of
(Str str) -> return $ stringToHtml str
(Space) -> return $ stringToHtml " "
- (LineBreak) -> return $ br
- (EmDash) -> return $ primHtmlChar "mdash"
- (EnDash) -> return $ primHtmlChar "ndash"
- (Ellipses) -> return $ primHtmlChar "hellip"
- (Apostrophe) -> return $ primHtmlChar "rsquo"
+ (LineBreak) -> return br
+ (EmDash) -> return $ stringToHtml "—"
+ (EnDash) -> return $ stringToHtml "–"
+ (Ellipses) -> return $ stringToHtml "…"
+ (Apostrophe) -> return $ stringToHtml "’"
(Emph lst) -> inlineListToHtml opts lst >>= return . emphasize
(Strong lst) -> inlineListToHtml opts lst >>= return . strong
(Code str) -> return $ thecode << str
@@ -426,10 +425,10 @@ inlineToHtml opts inline =
(Subscript lst) -> inlineListToHtml opts lst >>= return . sub
(Quoted quoteType lst) ->
let (leftQuote, rightQuote) = case quoteType of
- SingleQuote -> (primHtmlChar "lsquo",
- primHtmlChar "rsquo")
- DoubleQuote -> (primHtmlChar "ldquo",
- primHtmlChar "rdquo")
+ SingleQuote -> (stringToHtml "‘",
+ stringToHtml "’")
+ DoubleQuote -> (stringToHtml "“",
+ stringToHtml "”")
in do contents <- inlineListToHtml opts lst
return $ leftQuote +++ contents +++ rightQuote
(Math t str) ->
@@ -502,7 +501,7 @@ blockListToNote opts ref blocks =
-- that block. Otherwise, insert a new Plain block with the backlink.
let backlink = [HtmlInline $ " <a href=\"#" ++ writerIdentifierPrefix opts ++ "fnref" ++ ref ++
"\" class=\"footnoteBackLink\"" ++
- " title=\"Jump back to footnote " ++ ref ++ "\">↩</a>"]
+ " title=\"Jump back to footnote " ++ ref ++ "\">↩</a>"]
blocks' = if null blocks
then []
else let lastBlock = last blocks
diff --git a/tests/writer.html b/tests/writer.html
index c67cfbd42..cd5a687e0 100644
--- a/tests/writer.html
+++ b/tests/writer.html
@@ -11,7 +11,7 @@
<body>
<h1 class="title">Pandoc Test Suite</h1>
<p
->This is a set of tests for pandoc. Most of them are adapted from John Gruber’s markdown test suite.</p
+>This is a set of tests for pandoc. Most of them are adapted from John Gruber’s markdown test suite.</p
><hr
/><div id="headers"
><h1
@@ -64,11 +64,11 @@
><h1
>Paragraphs</h1
><p
- >Here’s a regular paragraph.</p
+ >Here’s a regular paragraph.</p
><p
>In Markdown 1.0.0 and earlier. Version 8. This line turns into a list item. Because a hard-wrapped line in the middle of a paragraph looked like a list item.</p
><p
- >Here’s one with a bullet. * criminey.</p
+ >Here’s one with a bullet. * criminey.</p
><p
>There should be a hard line break<br
/>here.</p
@@ -292,7 +292,7 @@ These should not be escaped: \$ \\ \> \[ \{
><p
>Item 1, graf one.</p
><p
- >Item 1. graf two. The quick brown fox jumped over the lazy dog’s back.</p
+ >Item 1. graf two. The quick brown fox jumped over the lazy dog’s back.</p
></li
><li
><p
@@ -320,7 +320,7 @@ These should not be escaped: \$ \\ \> \[ \{
></li
></ul
><p
- >Here’s another:</p
+ >Here’s another:</p
><ol style="list-style-type: decimal;"
><li
>First</li
@@ -634,7 +634,7 @@ These should not be escaped: \$ \\ \> \[ \{
<script type="text/javascript">document.write('This *should not* be interpreted as markdown');</script>
<p
- >Here’s a simple block:</p
+ >Here’s a simple block:</p
><div>
foo</div>
<p
@@ -692,7 +692,7 @@ Blah
</code
></pre
><p
- >Hr’s:</p
+ >Hr’s:</p
><hr>
<hr />
@@ -801,25 +801,25 @@ Blah
><h1
>Smart quotes, ellipses, dashes</h1
><p
- >“Hello,” said the spider. “‘Shelob’ is my name.”</p
+ >“Hello,” said the spider. “‘Shelob’ is my name.”</p
><p
- >‘A’, ‘B’, and ‘C’ are letters.</p
+ >‘A’, ‘B’, and ‘C’ are letters.</p
><p
- >‘Oak,’ ‘elm,’ and ‘beech’ are names of trees. So is ‘pine.’</p
+ >‘Oak,’ ‘elm,’ and ‘beech’ are names of trees. So is ‘pine.’</p
><p
- >‘He said, “I want to go.”’ Were you alive in the 70’s?</p
+ >‘He said, “I want to go.”’ Were you alive in the 70’s?</p
><p
- >Here is some quoted ‘<code
+ >Here is some quoted ‘<code
>code</code
- >’ and a “<a href="http://example.com/?foo=1&bar=2"
+ >’ and a “<a href="http://example.com/?foo=1&bar=2"
>quoted link</a
- >”.</p
+ >”.</p
><p
- >Some dashes: one—two — three—four — five.</p
+ >Some dashes: one—two — three—four — five.</p
><p
- >Dashes between numbers: 5–7, 255–66, 1987–1999.</p
+ >Dashes between numbers: 5–7, 255–66, 1987–1999.</p
><p
- >Ellipses…and…and….</p
+ >Ellipses…and…and….</p
><hr
/></div
><div id="latex"
@@ -855,7 +855,7 @@ Blah
></span
>-Tree</li
><li
- >Here’s some display math: <span class="math"
+ >Here’s some display math: <span class="math"
>\frac{<em
>d</em
>}{<em
@@ -883,7 +883,7 @@ Blah
>}</span
></li
><li
- >Here’s one that has a line break in it: <span class="math"
+ >Here’s one that has a line break in it: <span class="math"
>α+ω × <em
>x</em
><sup
@@ -892,7 +892,7 @@ Blah
>.</li
></ul
><p
- >These shouldn’t be math:</p
+ >These shouldn’t be math:</p
><ul
><li
>To get the famous equation, write <code
@@ -901,7 +901,7 @@ Blah
><li
>$22,000 is a <em
>lot</em
- > of money. So is $34,000. (It worked if “lot” is emphasized.)</li
+ > of money. So is $34,000. (It worked if “lot” is emphasized.)</li
><li
>Shoes ($20) and socks ($5).</li
><li
@@ -912,7 +912,7 @@ Blah
> 23$.</li
></ul
><p
- >Here’s a LaTeX table:</p
+ >Here’s a LaTeX table:</p
><p
></p
><hr
@@ -1083,19 +1083,19 @@ document.write('<a h'+'ref'+'="ma'+'ilto'+':'+e+'">'+'Email link'+'<\/'+'a'+'>')
><h2
>With ampersands</h2
><p
- >Here’s a <a href="http://example.com/?foo=1&bar=2"
+ >Here’s a <a href="http://example.com/?foo=1&bar=2"
>link with an ampersand in the URL</a
>.</p
><p
- >Here’s a link with an amersand in the link text: <a href="http://att.com/" title="AT&T"
+ >Here’s a link with an amersand in the link text: <a href="http://att.com/" title="AT&T"
>AT&T</a
>.</p
><p
- >Here’s an <a href="/script?foo=1&bar=2"
+ >Here’s an <a href="/script?foo=1&bar=2"
>inline link</a
>.</p
><p
- >Here’s an <a href="/script?foo=1&bar=2"
+ >Here’s an <a href="/script?foo=1&bar=2"
>inline link in pointy braces</a
>.</p
></div
@@ -1155,7 +1155,7 @@ document.write('<a h'+'ref'+'="ma'+'ilto'+':'+e+'">'+'<code>'+e+'</code>'+'<\/'+
><h1
>Images</h1
><p
- >From “Voyage dans la Lune” by Georges Melies (1902):</p
+ >From “Voyage dans la Lune” by Georges Melies (1902):</p
><p
><img src="lalune.jpg" title="Voyage dans la Lune" alt="lalune"
/></p
@@ -1207,11 +1207,11 @@ document.write('<a h'+'ref'+'="ma'+'ilto'+':'+e+'">'+'<code>'+e+'</code>'+'<\/'+
/><ol
><li id="fn1"
><p
- >Here is the footnote. It can go anywhere after the footnote reference. It need not be placed at the end of the document. <a href="#fnref1" class="footnoteBackLink" title="Jump back to footnote 1">↩</a></p
+ >Here is the footnote. It can go anywhere after the footnote reference. It need not be placed at the end of the document. <a href="#fnref1" class="footnoteBackLink" title="Jump back to footnote 1">↩</a></p
></li
><li id="fn2"
><p
- >Here’s the long note. This one contains multiple blocks.</p
+ >Here’s the long note. This one contains multiple blocks.</p
><p
>Subsequent blocks are indented to show that they belong to the footnote (as with list items).</p
><pre
@@ -1220,7 +1220,7 @@ document.write('<a h'+'ref'+'="ma'+'ilto'+':'+e+'">'+'<code>'+e+'</code>'+'<\/'+
</code
></pre
><p
- >If you want, you can indent every line, but you can also be lazy and just indent the first line of each block. <a href="#fnref2" class="footnoteBackLink" title="Jump back to footnote 2">↩</a></p
+ >If you want, you can indent every line, but you can also be lazy and just indent the first line of each block. <a href="#fnref2" class="footnoteBackLink" title="Jump back to footnote 2">↩</a></p
></li
><li id="fn3"
><p
@@ -1230,15 +1230,15 @@ document.write('<a h'+'ref'+'="ma'+'ilto'+':'+e+'">'+'<code>'+e+'</code>'+'<\/'+
>links</a
> and <code
>]</code
- > verbatim characters, as well as [bracketed text]. <a href="#fnref3" class="footnoteBackLink" title="Jump back to footnote 3">↩</a></p
+ > verbatim characters, as well as [bracketed text]. <a href="#fnref3" class="footnoteBackLink" title="Jump back to footnote 3">↩</a></p
></li
><li id="fn4"
><p
- >In quote. <a href="#fnref4" class="footnoteBackLink" title="Jump back to footnote 4">↩</a></p
+ >In quote. <a href="#fnref4" class="footnoteBackLink" title="Jump back to footnote 4">↩</a></p
></li
><li id="fn5"
><p
- >In list. <a href="#fnref5" class="footnoteBackLink" title="Jump back to footnote 5">↩</a></p
+ >In list. <a href="#fnref5" class="footnoteBackLink" title="Jump back to footnote 5">↩</a></p
></li
></ol
></div