From 047b5984b6dd17266dc07c9b0ef4cee82f84e914 Mon Sep 17 00:00:00 2001
From: muji <freeformsystems@gmail.com>
Date: Sat, 23 Apr 2016 13:20:21 +0800
Subject: [PATCH 1/4] Support _htmlBlockType for html_inline.

Add support for the _htmlBlockType property on inline HTML elements;
this allows AST consumers to be able to know ahead of time the type for
inline HTML elements which is useful when parsing processing
instructions or comments.

Moved the reHtmlBlockOpen RegExp to the common module so that it may be
shared between the inline and block parsers.
---
 lib/blocks.js  | 11 +----------
 lib/common.js  | 12 ++++++++++++
 lib/inlines.js | 11 +++++++++++
 3 files changed, 24 insertions(+), 10 deletions(-)

diff --git a/lib/blocks.js b/lib/blocks.js
index a7f405f7..1a8419ab 100644
--- a/lib/blocks.js
+++ b/lib/blocks.js
@@ -16,16 +16,7 @@ var C_OPEN_BRACKET = 91;
 
 var InlineParser = require('./inlines');
 
-var reHtmlBlockOpen = [
-   /./, // dummy for 0
-   /^<(?:script|pre|style)(?:\s|>|$)/i,
-   /^<!--/,
-   /^<[?]/,
-   /^<![A-Z]/,
-   /^<!\[CDATA\[/,
-   /^<[/]?(?:address|article|aside|base|basefont|blockquote|body|caption|center|col|colgroup|dd|details|dialog|dir|div|dl|dt|fieldset|figcaption|figure|footer|form|frame|frameset|h1|head|header|hr|html|iframe|legend|li|link|main|menu|menuitem|meta|nav|noframes|ol|optgroup|option|p|param|section|source|title|summary|table|tbody|td|tfoot|th|thead|title|tr|track|ul)(?:\s|[/]?[>]|$)/i,
-    new RegExp('^(?:' + OPENTAG + '|' + CLOSETAG + ')\s*$', 'i')
-];
+var reHtmlBlockOpen = require('./common').reHtmlBlockOpen;
 
 var reHtmlBlockClose = [
    /./, // dummy for 0
diff --git a/lib/common.js b/lib/common.js
index 605a3bf7..65316e91 100644
--- a/lib/common.js
+++ b/lib/common.js
@@ -27,6 +27,17 @@ var HTMLTAG = "(?:" + OPENTAG + "|" + CLOSETAG + "|" + HTMLCOMMENT + "|" +
         PROCESSINGINSTRUCTION + "|" + DECLARATION + "|" + CDATA + ")";
 var reHtmlTag = new RegExp('^' + HTMLTAG, 'i');
 
+var reHtmlBlockOpen = [
+   /./, // dummy for 0
+   /^<(?:script|pre|style)(?:\s|>|$)/i,
+   /^<!--/,
+   /^<[?]/,
+   /^<![A-Z]/,
+   /^<!\[CDATA\[/,
+   /^<[/]?(?:address|article|aside|base|basefont|blockquote|body|caption|center|col|colgroup|dd|details|dialog|dir|div|dl|dt|fieldset|figcaption|figure|footer|form|frame|frameset|h1|head|header|hr|html|iframe|legend|li|link|main|menu|menuitem|meta|nav|noframes|ol|optgroup|option|p|param|section|source|title|summary|table|tbody|td|tfoot|th|thead|title|tr|track|ul)(?:\s|[/]?[>]|$)/i,
+    new RegExp('^(?:' + OPENTAG + '|' + CLOSETAG + ')\s*$', 'i')
+];
+
 var reBackslashOrAmp = /[\\&]/;
 
 var ESCAPABLE = '[!"#$%&\'()*+,./:;<=>?@[\\\\\\]^_`{|}~-]';
@@ -96,6 +107,7 @@ module.exports = { unescapeString: unescapeString,
                    normalizeURI: normalizeURI,
                    escapeXml: escapeXml,
                    reHtmlTag: reHtmlTag,
+                   reHtmlBlockOpen: reHtmlBlockOpen,
                    OPENTAG: OPENTAG,
                    CLOSETAG: CLOSETAG,
                    ENTITY: ENTITY,
diff --git a/lib/inlines.js b/lib/inlines.js
index 96954864..99f6dc4d 100644
--- a/lib/inlines.js
+++ b/lib/inlines.js
@@ -10,6 +10,8 @@ var fromCodePoint = require('./from-code-point.js');
 var decodeHTML = require('entities').decodeHTML;
 require('string.prototype.repeat'); // Polyfill for String.prototype.repeat
 
+var reHtmlBlockOpen = require('./common').reHtmlBlockOpen;
+
 // Constants for character codes:
 
 var C_NEWLINE = 10;
@@ -213,6 +215,15 @@ var parseHtmlTag = function(block) {
         var node = new Node('html_inline');
         node._literal = m;
         block.appendChild(node);
+
+        var blockType;
+        for (blockType = 1; blockType <= 7; blockType++) {
+            if(reHtmlBlockOpen[blockType].test(m) && blockType < 7) {
+              node._htmlBlockType = blockType;
+              break;
+            }
+        }
+
         return true;
     }
 };

From 9032bba0e35b440222be568d6d0719528ac37477 Mon Sep 17 00:00:00 2001
From: muji <freeformsystems@gmail.com>
Date: Sat, 23 Apr 2016 13:31:22 +0800
Subject: [PATCH 2/4] Simplify iteration.

---
 lib/inlines.js | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/lib/inlines.js b/lib/inlines.js
index 99f6dc4d..0e5219df 100644
--- a/lib/inlines.js
+++ b/lib/inlines.js
@@ -217,8 +217,8 @@ var parseHtmlTag = function(block) {
         block.appendChild(node);
 
         var blockType;
-        for (blockType = 1; blockType <= 7; blockType++) {
-            if(reHtmlBlockOpen[blockType].test(m) && blockType < 7) {
+        for (blockType = 1; blockType < 7; blockType++) {
+            if(reHtmlBlockOpen[blockType].test(m)) {
               node._htmlBlockType = blockType;
               break;
             }

From ec01a508199400c44903c7819be52ee81812ba62 Mon Sep 17 00:00:00 2001
From: muji <freeformsystems@gmail.com>
Date: Sat, 23 Apr 2016 14:16:44 +0800
Subject: [PATCH 3/4] Support lowercase doctype declarations.

HTML5 allows the lowercase version `<!doctype html>`.
---
 lib/common.js | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/lib/common.js b/lib/common.js
index 65316e91..095dd569 100644
--- a/lib/common.js
+++ b/lib/common.js
@@ -21,7 +21,7 @@ var OPENTAG = "<" + TAGNAME + ATTRIBUTE + "*" + "\\s*/?>";
 var CLOSETAG = "</" + TAGNAME + "\\s*[>]";
 var HTMLCOMMENT = "<!---->|<!--(?:-?[^>-])(?:-?[^-])*-->";
 var PROCESSINGINSTRUCTION = "[<][?].*?[?][>]";
-var DECLARATION = "<![A-Z]+" + "\\s+[^>]*>";
+var DECLARATION = "<![A-Za-z]+" + "\\s+[^>]*>";
 var CDATA = "<!\\[CDATA\\[[\\s\\S]*?\\]\\]>";
 var HTMLTAG = "(?:" + OPENTAG + "|" + CLOSETAG + "|" + HTMLCOMMENT + "|" +
         PROCESSINGINSTRUCTION + "|" + DECLARATION + "|" + CDATA + ")";
@@ -32,7 +32,7 @@ var reHtmlBlockOpen = [
    /^<(?:script|pre|style)(?:\s|>|$)/i,
    /^<!--/,
    /^<[?]/,
-   /^<![A-Z]/,
+   /^<![A-Za-z]/,
    /^<!\[CDATA\[/,
    /^<[/]?(?:address|article|aside|base|basefont|blockquote|body|caption|center|col|colgroup|dd|details|dialog|dir|div|dl|dt|fieldset|figcaption|figure|footer|form|frame|frameset|h1|head|header|hr|html|iframe|legend|li|link|main|menu|menuitem|meta|nav|noframes|ol|optgroup|option|p|param|section|source|title|summary|table|tbody|td|tfoot|th|thead|title|tr|track|ul)(?:\s|[/]?[>]|$)/i,
     new RegExp('^(?:' + OPENTAG + '|' + CLOSETAG + ')\s*$', 'i')

From 42223ceea081b9d969e92c6e7396f2f2d695c33f Mon Sep 17 00:00:00 2001
From: muji <freeformsystems@gmail.com>
Date: Sat, 23 Apr 2016 14:43:33 +0800
Subject: [PATCH 4/4] Match on inline element names.

When determining _htmlBlockType for html_inline node types.
---
 lib/inlines.js | 6 +++++-
 1 file changed, 5 insertions(+), 1 deletion(-)

diff --git a/lib/inlines.js b/lib/inlines.js
index 0e5219df..001ddbca 100644
--- a/lib/inlines.js
+++ b/lib/inlines.js
@@ -10,7 +10,11 @@ var fromCodePoint = require('./from-code-point.js');
 var decodeHTML = require('entities').decodeHTML;
 require('string.prototype.repeat'); // Polyfill for String.prototype.repeat
 
-var reHtmlBlockOpen = require('./common').reHtmlBlockOpen;
+var reHtmlBlockOpen = require('./common').reHtmlBlockOpen.slice(0);
+
+// match on inline element names
+reHtmlBlockOpen[6] = 
+   /^<[/]?(?:big|small|tt|abbr|acronym|cite|code|dfn|em|kbd|strong|samp|time|var|bdo|br|img|map|object|script|span|sub|sup|button|input|label|select|textarea|b|i|a|q)(?:\s|[/]?[>]|$)/i;
 
 // Constants for character codes: