Comment parsing: in the generated XML file, mark HTML that is safe to pass

through to the output even if the input comment comes from an untrusted source

Attribute filtering is currently based on a blacklist, which right now includes
all event handler attributes (they contain JavaScipt code).  It should be
switched to a whitelist, but going over all of the HTML5 spec requires a
significant amount of time.

git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@206882 91177308-0d34-0410-b5e6-96231b3b80d8

diff --git a/bindings/xml/comment-xml-schema.rng b/bindings/xml/comment-xml-schema.rng
index a8913a360b79d4fcb8bf46a55f012f4e7b2cf283..29a91bf674f65a1105845e3c07da0c34fb6d60db 100644 (file)
--- a/bindings/xml/comment-xml-schema.rng
+++ b/bindings/xml/comment-xml-schema.rng
@@ -580,6 +580,12 @@
         </data>
       </element>
       <element name="rawHTML">
+        <optional>
+          <!-- If not specified, the default value is 'false'. -->
+          <attribute name="isSafeToPassThrough">
+            <data type="boolean" />
+          </attribute>
+        </optional>
         <!-- Non-empty text content. -->
         <data type="string">
           <param name="pattern">.*\S.*</param>

diff --git a/include/clang/AST/Comment.h b/include/clang/AST/Comment.h
index 50e9196b1a9d62adec9d4d57fa2f23d834fba06d..90dfb2570a5f91f96b3d2a6849d64256a8f17ae6 100644 (file)
--- a/include/clang/AST/Comment.h
+++ b/include/clang/AST/Comment.h
@@ -100,16 +100,27 @@ protected:
   };
   enum { NumInlineCommandCommentBits = NumInlineContentCommentBits + 10 };
 
+  class HTMLTagCommentBitfields {
+    friend class HTMLTagComment;
+
+    unsigned : NumInlineContentCommentBits;
+
+    /// True if this tag is safe to pass through to HTML output even if the
+    /// comment comes from an untrusted source.
+    unsigned IsSafeToPassThrough : 1;
+  };
+  enum { NumHTMLTagCommentBits = NumInlineContentCommentBits + 1 };
+
   class HTMLStartTagCommentBitfields {
     friend class HTMLStartTagComment;
 
-    unsigned : NumInlineContentCommentBits;
+    unsigned : NumHTMLTagCommentBits;
 
     /// True if this tag is self-closing (e. g., <br />).  This is based on tag
     /// spelling in comment (plain <br> would not set this flag).
     unsigned IsSelfClosing : 1;
   };
-  enum { NumHTMLStartTagCommentBits = NumInlineContentCommentBits + 1 };
+  enum { NumHTMLStartTagCommentBits = NumHTMLTagCommentBits + 1 };
 
   class ParagraphCommentBitfields {
     friend class ParagraphComment;
@@ -155,6 +166,7 @@ protected:
     InlineContentCommentBitfields InlineContentCommentBits;
     TextCommentBitfields TextCommentBits;
     InlineCommandCommentBitfields InlineCommandCommentBits;
+    HTMLTagCommentBitfields HTMLTagCommentBits;
     HTMLStartTagCommentBitfields HTMLStartTagCommentBits;
     ParagraphCommentBitfields ParagraphCommentBits;
     BlockCommandCommentBitfields BlockCommandCommentBits;
@@ -360,8 +372,7 @@ public:
 };
 
 /// Abstract class for opening and closing HTML tags.  HTML tags are always
-/// treated as inline content (regardless HTML semantics); opening and closing
-/// tags are not matched.
+/// treated as inline content (regardless HTML semantics).
 class HTMLTagComment : public InlineContentComment {
 protected:
   StringRef TagName;
@@ -377,6 +388,7 @@ protected:
       TagName(TagName),
       TagNameRange(TagNameBegin, TagNameEnd) {
     setLocation(TagNameBegin);
+    HTMLTagCommentBits.IsSafeToPassThrough = 1;
   }
 
 public:
@@ -392,6 +404,14 @@ public:
     return SourceRange(L.getLocWithOffset(1),
                        L.getLocWithOffset(1 + TagName.size()));
   }
+
+  bool isSafeToPassThrough() const {
+    return HTMLTagCommentBits.IsSafeToPassThrough;
+  }
+
+  void setUnsafeToPassThrough() {
+    HTMLTagCommentBits.IsSafeToPassThrough = 0;
+  }
 };
 
 /// An opening HTML tag with attributes.

diff --git a/include/clang/AST/CommentHTMLTags.td b/include/clang/AST/CommentHTMLTags.td
index f98e32ddca4f830ca97b09435822762d13c99983..79951b80ee5ea0230f210bef250fabf9c0ba982a 100644 (file)
--- a/include/clang/AST/CommentHTMLTags.td
+++ b/include/clang/AST/CommentHTMLTags.td
@@ -52,3 +52,83 @@ def Tr      : Tag<"tr"> { let EndTagOptional = 1; }
 def Th      : Tag<"th"> { let EndTagOptional = 1; }
 def Td      : Tag<"td"> { let EndTagOptional = 1; }
 
+// Define a blacklist of attributes that are not safe to pass through to HTML
+// output if the input is untrusted.
+//
+// FIXME: this should be a whitelist.  When changing this to a whitelist, don't
+// forget to change the default in the TableGen backend.
+class Attribute<string spelling> {
+  string Spelling = spelling;
+  bit IsSafeToPassThrough = 1;
+}
+class EventHandlerContentAttribute<string spelling> : Attribute<spelling> {
+  let IsSafeToPassThrough = 0;
+}
+
+// This list is based on HTML5 draft as of 04 February 2014.
+//
+// The list is intentionally organized as one item per line to make it easier
+// to compare with the HTML spec.
+foreach AttrName = [
+    "onabort",
+    "onblur",
+    "oncancel",
+    "oncanplay",
+    "oncanplaythrough",
+    "onchange",
+    "onclick",
+    "onclose",
+    "oncuechange",
+    "ondblclick",
+    "ondrag",
+    "ondragend",
+    "ondragenter",
+    "ondragexit",
+    "ondragleave",
+    "ondragover",
+    "ondragstart",
+    "ondrop",
+    "ondurationchange",
+    "onemptied",
+    "onended",
+    "onerror",
+    "onfocus",
+    "oninput",
+    "oninvalid",
+    "onkeydown",
+    "onkeypress",
+    "onkeyup",
+    "onload",
+    "onloadeddata",
+    "onloadedmetadata",
+    "onloadstart",
+    "onmousedown",
+    "onmouseenter",
+    "onmouseleave",
+    "onmousemove",
+    "onmouseout",
+    "onmouseover",
+    "onmouseup",
+    "onmousewheel",
+    "onpause",
+    "onplay",
+    "onplaying",
+    "onprogress",
+    "onratechange",
+    "onreset",
+    "onresize",
+    "onscroll",
+    "onseeked",
+    "onseeking",
+    "onselect",
+    "onshow",
+    "onstalled",
+    "onsubmit",
+    "onsuspend",
+    "ontimeupdate",
+    "ontoggle",
+    "onvolumechange",
+    "onwaiting"
+  ] in {
+  def Attr#AttrName : EventHandlerContentAttribute<AttrName>;
+}

diff --git a/include/clang/Basic/DiagnosticCommentKinds.td b/include/clang/Basic/DiagnosticCommentKinds.td
index 49781fec9a9cf882b96aba4ac05194ce364289eb..49296f9b5e1ea63d47d964e808f66984dab8c745 100644 (file)
--- a/include/clang/Basic/DiagnosticCommentKinds.td
+++ b/include/clang/Basic/DiagnosticCommentKinds.td
@@ -41,6 +41,10 @@ def warn_doc_html_start_end_mismatch : Warning<
 def note_doc_html_end_tag : Note<
   "end tag">;
 
+def warn_doc_html_missing_end_tag : Warning<
+  "HTML tag '%0' requires an end tag">,
+  InGroup<DocumentationHTML>, DefaultIgnore;
+
 // Commands
 
 def warn_doc_block_command_empty_paragraph : Warning<

diff --git a/lib/AST/CommentSema.cpp b/lib/AST/CommentSema.cpp
index e51f8781c1b0ca67d6b7daf2c142c9dc6a7af5e7..5a0bf37739cc0a36c311482bbbb4bb9c98b736cc 100644 (file)
--- a/lib/AST/CommentSema.cpp
+++ b/lib/AST/CommentSema.cpp
@@ -467,6 +467,11 @@ void Sema::actOnHTMLStartTagFinish(
                               SourceLocation GreaterLoc,
                               bool IsSelfClosing) {
   Tag->setAttrs(Attrs);
+  for (const auto &Attr : Attrs) {
+    if (!isHTMLAttributeSafeToPassThrough(Attr.Name))
+      Tag->setUnsafeToPassThrough();
+  }
+
   Tag->setGreaterLoc(GreaterLoc);
   if (IsSelfClosing)
     Tag->setSelfClosing();
@@ -482,6 +487,7 @@ HTMLEndTagComment *Sema::actOnHTMLEndTag(SourceLocation LocBegin,
   if (isHTMLEndTagForbidden(TagName)) {
     Diag(HET->getLocation(), diag::warn_doc_html_end_forbidden)
       << TagName << HET->getSourceRange();
+    HET->setUnsafeToPassThrough();
     return HET;
   }
 
@@ -497,14 +503,19 @@ HTMLEndTagComment *Sema::actOnHTMLEndTag(SourceLocation LocBegin,
   if (!FoundOpen) {
     Diag(HET->getLocation(), diag::warn_doc_html_end_unbalanced)
       << HET->getSourceRange();
+    HET->setUnsafeToPassThrough();
     return HET;
   }
 
   while (!HTMLOpenTags.empty()) {
-    const HTMLStartTagComment *HST = HTMLOpenTags.pop_back_val();
+    HTMLStartTagComment *HST = HTMLOpenTags.pop_back_val();
     StringRef LastNotClosedTagName = HST->getTagName();
-    if (LastNotClosedTagName == TagName)
+    if (LastNotClosedTagName == TagName) {
+      // If the start tag is unsafe, end tag is unsafe as well.
+      if (!HST->isSafeToPassThrough())
+        HET->setUnsafeToPassThrough();
       break;
+    }
 
     if (isHTMLEndTagOptional(LastNotClosedTagName))
       continue;
@@ -518,16 +529,18 @@ HTMLEndTagComment *Sema::actOnHTMLEndTag(SourceLocation LocBegin,
                                                 HET->getLocation(),
                                                 &CloseLineInvalid);
 
-    if (OpenLineInvalid || CloseLineInvalid || OpenLine == CloseLine)
+    if (OpenLineInvalid || CloseLineInvalid || OpenLine == CloseLine) {
       Diag(HST->getLocation(), diag::warn_doc_html_start_end_mismatch)
         << HST->getTagName() << HET->getTagName()
         << HST->getSourceRange() << HET->getSourceRange();
-    else {
+      HST->setUnsafeToPassThrough();
+    } else {
       Diag(HST->getLocation(), diag::warn_doc_html_start_end_mismatch)
         << HST->getTagName() << HET->getTagName()
         << HST->getSourceRange();
       Diag(HET->getLocation(), diag::note_doc_html_end_tag)
         << HET->getSourceRange();
+      HST->setUnsafeToPassThrough();
     }
   }
 
@@ -538,6 +551,18 @@ FullComment *Sema::actOnFullComment(
                               ArrayRef<BlockContentComment *> Blocks) {
   FullComment *FC = new (Allocator) FullComment(Blocks, ThisDeclInfo);
   resolveParamCommandIndexes(FC);
+
+  // Complain about HTML tags that are not closed.
+  while (!HTMLOpenTags.empty()) {
+    HTMLStartTagComment *HST = HTMLOpenTags.pop_back_val();
+    if (isHTMLEndTagOptional(HST->getTagName()))
+      continue;
+
+    Diag(HST->getLocation(), diag::warn_doc_html_missing_end_tag)
+      << HST->getTagName() << HST->getSourceRange();
+    HST->setUnsafeToPassThrough();
+  }
+
   return FC;
 }
 

diff --git a/lib/Index/CommentToXML.cpp b/lib/Index/CommentToXML.cpp
index 43c423274daac70ba18b70d623dc4495a8401882..377440f81d58ca61cbf37e217752998f5a94e35c 100644 (file)
--- a/lib/Index/CommentToXML.cpp
+++ b/lib/Index/CommentToXML.cpp
@@ -667,14 +667,20 @@ void CommentASTToXMLConverter::visitInlineCommandComment(
 
 void CommentASTToXMLConverter::visitHTMLStartTagComment(
     const HTMLStartTagComment *C) {
-  Result << "<rawHTML><![CDATA[";
+  Result << "<rawHTML";
+  if (C->isSafeToPassThrough())
+    Result << " isSafeToPassThrough=\"1\"";
+  Result << "><![CDATA[";
   printHTMLStartTagComment(C, Result);
   Result << "]]></rawHTML>";
 }
 
 void
 CommentASTToXMLConverter::visitHTMLEndTagComment(const HTMLEndTagComment *C) {
-  Result << "<rawHTML>&lt;/" << C->getTagName() << "&gt;</rawHTML>";
+  Result << "<rawHTML";
+  if (C->isSafeToPassThrough())
+    Result << " isSafeToPassThrough=\"1\"";
+  Result << ">&lt;/" << C->getTagName() << "&gt;</rawHTML>";
 }
 
 void

diff --git a/test/Index/Inputs/CommentXML/valid-function-02.xml b/test/Index/Inputs/CommentXML/valid-function-02.xml
index 989d6a7c143d2845c9ef7a6325a226252c7b022b..98e4fd1c1edbe673a3f7c40ed22ea249c941b5b6 100644 (file)
--- a/test/Index/Inputs/CommentXML/valid-function-02.xml
+++ b/test/Index/Inputs/CommentXML/valid-function-02.xml
@@ -1,5 +1,14 @@
 <?xml version="1.0" encoding="utf-8"?>
 <Function>
 <Name>aaa</Name>
-<Abstract><Para>Aaa <bold>bbb</bold> <monospaced>ccc</monospaced> <emphasized>ddd</emphasized>.</Para></Abstract>
+<Abstract>
+  <Para>Aaa
+    <bold>bbb</bold>
+    <monospaced>ccc</monospaced>
+    <emphasized>ddd</emphasized>
+    <rawHTML>&lt;eee&gt;</rawHTML>
+    <rawHTML isSafeToPassThrough="0">&lt;fff&gt;</rawHTML>
+    <rawHTML isSafeToPassThrough="1">&lt;ggg&gt;</rawHTML>.
+  </Para>
+</Abstract>
 </Function>

diff --git a/test/Index/comment-to-html-xml-conversion.cpp b/test/Index/comment-to-html-xml-conversion.cpp
index 327fa6448327e489142a8ee629f8af29f5640f9d..590e18719741546a6197732f5cf687d580e71149 100644 (file)
--- a/test/Index/comment-to-html-xml-conversion.cpp
+++ b/test/Index/comment-to-html-xml-conversion.cpp
@@ -472,7 +472,7 @@ void test_full_comment_1(int x1, int x2);
 /// <br><a href="http://example.com/">Aaa</a>
 void comment_to_html_conversion_24();
 
-// CHECK: comment-to-html-xml-conversion.cpp:[[@LINE-2]]:6: FunctionDecl=comment_to_html_conversion_24:{{.*}} FullCommentAsHTML=[<p class="para-brief"> <br><a href="http://example.com/">Aaa</a></p>] FullCommentAsXML=[<Function file="{{[^"]+}}comment-to-html-xml-conversion.cpp" line="[[@LINE-2]]" column="6"><Name>comment_to_html_conversion_24</Name><USR>c:@F@comment_to_html_conversion_24#</USR><Declaration>void comment_to_html_conversion_24()</Declaration><Abstract><Para> <rawHTML><![CDATA[<br>]]></rawHTML><rawHTML><![CDATA[<a href="http://example.com/">]]></rawHTML>Aaa<rawHTML>&lt;/a&gt;</rawHTML></Para></Abstract></Function>]
+// CHECK: comment-to-html-xml-conversion.cpp:[[@LINE-2]]:6: FunctionDecl=comment_to_html_conversion_24:{{.*}} FullCommentAsHTML=[<p class="para-brief"> <br><a href="http://example.com/">Aaa</a></p>] FullCommentAsXML=[<Function file="{{[^"]+}}comment-to-html-xml-conversion.cpp" line="[[@LINE-2]]" column="6"><Name>comment_to_html_conversion_24</Name><USR>c:@F@comment_to_html_conversion_24#</USR><Declaration>void comment_to_html_conversion_24()</Declaration><Abstract><Para> <rawHTML isSafeToPassThrough="1"><![CDATA[<br>]]></rawHTML><rawHTML isSafeToPassThrough="1"><![CDATA[<a href="http://example.com/">]]></rawHTML>Aaa<rawHTML isSafeToPassThrough="1">&lt;/a&gt;</rawHTML></Para></Abstract></Function>]
 // CHECK-NEXT:  CommentAST=[
 // CHECK-NEXT:    (CXComment_FullComment
 // CHECK-NEXT:       (CXComment_Paragraph
@@ -678,7 +678,7 @@ void comment_to_html_conversion_33();
 /// <em>0&lt;i</em>
 void comment_to_html_conversion_34();
 
-// CHECK: comment-to-html-xml-conversion.cpp:[[@LINE-2]]:6: FunctionDecl=comment_to_html_conversion_34:{{.*}} FullCommentAsHTML=[<p class="para-brief"> <em>0&lt;i</em></p>] FullCommentAsXML=[<Function file="{{[^"]+}}comment-to-html-xml-conversion.cpp" line="[[@LINE-2]]" column="6"><Name>comment_to_html_conversion_34</Name><USR>c:@F@comment_to_html_conversion_34#</USR><Declaration>void comment_to_html_conversion_34()</Declaration><Abstract><Para> <rawHTML><![CDATA[<em>]]></rawHTML>0&lt;i<rawHTML>&lt;/em&gt;</rawHTML></Para></Abstract></Function>]
+// CHECK: comment-to-html-xml-conversion.cpp:[[@LINE-2]]:6: FunctionDecl=comment_to_html_conversion_34:{{.*}} FullCommentAsHTML=[<p class="para-brief"> <em>0&lt;i</em></p>] FullCommentAsXML=[<Function file="{{[^"]+}}comment-to-html-xml-conversion.cpp" line="[[@LINE-2]]" column="6"><Name>comment_to_html_conversion_34</Name><USR>c:@F@comment_to_html_conversion_34#</USR><Declaration>void comment_to_html_conversion_34()</Declaration><Abstract><Para> <rawHTML isSafeToPassThrough="1"><![CDATA[<em>]]></rawHTML>0&lt;i<rawHTML isSafeToPassThrough="1">&lt;/em&gt;</rawHTML></Para></Abstract></Function>]
 // CHECK-NEXT:  CommentAST=[
 // CHECK-NEXT:    (CXComment_FullComment
 // CHECK-NEXT:       (CXComment_Paragraph
@@ -853,6 +853,34 @@ enum class comment_to_xml_conversion_17 {
 // CHECK: comment-to-html-xml-conversion.cpp:[[@LINE-1]]:3: EnumConstantDecl=comment_to_xml_conversion_18:{{.*}} FullCommentAsXML=[<Variable file="{{[^"]+}}comment-to-html-xml-conversion.cpp" line="[[@LINE-1]]" column="3"><Name>comment_to_xml_conversion_18</Name><USR>c:@E@comment_to_xml_conversion_17@comment_to_xml_conversion_18</USR><Declaration>comment_to_xml_conversion_18</Declaration><Abstract><Para> Aaa.</Para></Abstract></Variable>]
 };
 
+/// <a href="http://example.org/">
+void comment_to_xml_conversion_unsafe_html_01();
+// CHECK: comment-to-html-xml-conversion.cpp:[[@LINE-1]]:6: FunctionDecl=comment_to_xml_conversion_unsafe_html_01:{{.*}} FullCommentAsXML=[<Function file="{{[^"]+}}comment-to-html-xml-conversion.cpp" line="[[@LINE-1]]" column="6"><Name>comment_to_xml_conversion_unsafe_html_01</Name><USR>c:@F@comment_to_xml_conversion_unsafe_html_01#</USR><Declaration>void comment_to_xml_conversion_unsafe_html_01()</Declaration><Abstract><Para> <rawHTML><![CDATA[<a href="http://example.org/">]]></rawHTML></Para></Abstract></Function>]
+
+/// <a href="http://example.org/"><em>Aaa</em>
+void comment_to_xml_conversion_unsafe_html_02();
+// CHECK: comment-to-html-xml-conversion.cpp:[[@LINE-1]]:6: FunctionDecl=comment_to_xml_conversion_unsafe_html_02:{{.*}} FullCommentAsXML=[<Function file="{{[^"]+}}comment-to-html-xml-conversion.cpp" line="[[@LINE-1]]" column="6"><Name>comment_to_xml_conversion_unsafe_html_02</Name><USR>c:@F@comment_to_xml_conversion_unsafe_html_02#</USR><Declaration>void comment_to_xml_conversion_unsafe_html_02()</Declaration><Abstract><Para> <rawHTML><![CDATA[<a href="http://example.org/">]]></rawHTML><rawHTML isSafeToPassThrough="1"><![CDATA[<em>]]></rawHTML>Aaa<rawHTML isSafeToPassThrough="1">&lt;/em&gt;</rawHTML></Para></Abstract></Function>]
+
+/// <em>Aaa
+void comment_to_xml_conversion_unsafe_html_03();
+// CHECK: comment-to-html-xml-conversion.cpp:[[@LINE-1]]:6: FunctionDecl=comment_to_xml_conversion_unsafe_html_03:{{.*}} FullCommentAsXML=[<Function file="{{[^"]+}}comment-to-html-xml-conversion.cpp" line="[[@LINE-1]]" column="6"><Name>comment_to_xml_conversion_unsafe_html_03</Name><USR>c:@F@comment_to_xml_conversion_unsafe_html_03#</USR><Declaration>void comment_to_xml_conversion_unsafe_html_03()</Declaration><Abstract><Para> <rawHTML><![CDATA[<em>]]></rawHTML>Aaa</Para></Abstract></Function>]
+
+/// <em>Aaa</b></em>
+void comment_to_xml_conversion_unsafe_html_04();
+// CHECK: comment-to-html-xml-conversion.cpp:[[@LINE-1]]:6: FunctionDecl=comment_to_xml_conversion_unsafe_html_04:{{.*}} FullCommentAsXML=[<Function file="{{[^"]+}}comment-to-html-xml-conversion.cpp" line="[[@LINE-1]]" column="6"><Name>comment_to_xml_conversion_unsafe_html_04</Name><USR>c:@F@comment_to_xml_conversion_unsafe_html_04#</USR><Declaration>void comment_to_xml_conversion_unsafe_html_04()</Declaration><Abstract><Para> <rawHTML isSafeToPassThrough="1"><![CDATA[<em>]]></rawHTML>Aaa<rawHTML>&lt;/b&gt;</rawHTML><rawHTML isSafeToPassThrough="1">&lt;/em&gt;</rawHTML></Para></Abstract></Function>]
+
+/// <em>Aaa</em></b>
+void comment_to_xml_conversion_unsafe_html_05();
+// CHECK: comment-to-html-xml-conversion.cpp:[[@LINE-1]]:6: FunctionDecl=comment_to_xml_conversion_unsafe_html_05:{{.*}} FullCommentAsXML=[<Function file="{{[^"]+}}comment-to-html-xml-conversion.cpp" line="[[@LINE-1]]" column="6"><Name>comment_to_xml_conversion_unsafe_html_05</Name><USR>c:@F@comment_to_xml_conversion_unsafe_html_05#</USR><Declaration>void comment_to_xml_conversion_unsafe_html_05()</Declaration><Abstract><Para> <rawHTML isSafeToPassThrough="1"><![CDATA[<em>]]></rawHTML>Aaa<rawHTML isSafeToPassThrough="1">&lt;/em&gt;</rawHTML><rawHTML>&lt;/b&gt;</rawHTML></Para></Abstract></Function>]
+
+/// </table>
+void comment_to_xml_conversion_unsafe_html_06();
+// CHECK: comment-to-html-xml-conversion.cpp:[[@LINE-1]]:6: FunctionDecl=comment_to_xml_conversion_unsafe_html_06:{{.*}} FullCommentAsXML=[<Function file="{{[^"]+}}comment-to-html-xml-conversion.cpp" line="[[@LINE-1]]" column="6"><Name>comment_to_xml_conversion_unsafe_html_06</Name><USR>c:@F@comment_to_xml_conversion_unsafe_html_06#</USR><Declaration>void comment_to_xml_conversion_unsafe_html_06()</Declaration><Abstract><Para> <rawHTML>&lt;/table&gt;</rawHTML></Para></Abstract></Function>]
+
+/// <div onclick="alert('meow');">Aaa</div>
+void comment_to_xml_conversion_unsafe_html_07();
+// CHECK: comment-to-html-xml-conversion.cpp:[[@LINE-1]]:6: FunctionDecl=comment_to_xml_conversion_unsafe_html_07:{{.*}} FullCommentAsXML=[<Function file="{{[^"]+}}comment-to-html-xml-conversion.cpp" line="[[@LINE-1]]" column="6"><Name>comment_to_xml_conversion_unsafe_html_07</Name><USR>c:@F@comment_to_xml_conversion_unsafe_html_07#</USR><Declaration>void comment_to_xml_conversion_unsafe_html_07()</Declaration><Abstract><Para> <rawHTML><![CDATA[<div onclick="alert('meow');">]]></rawHTML>Aaa<rawHTML>&lt;/div&gt;</rawHTML></Para></Abstract></Function>]
+
 //===---
 // Check that we attach comments from the base class to derived classes if they don't have a comment.
 // rdar://13647476

diff --git a/test/Sema/warn-documentation.cpp b/test/Sema/warn-documentation.cpp
index ed25d949f5594b3bad4b2d8be42b576c4a96f7d5..4375cfcf560a31a1fb2ef8ce0f297f3fcf025d83 100644 (file)
--- a/test/Sema/warn-documentation.cpp
+++ b/test/Sema/warn-documentation.cpp
@@ -4,35 +4,43 @@
 // RUN: c-index-test -test-load-source all -comments-xml-schema=%S/../../bindings/xml/comment-xml-schema.rng %s | FileCheck %s -check-prefix=WRONG
 // WRONG-NOT: CommentXMLInvalid
 
+// expected-warning@+2 {{HTML tag 'a' requires an end tag}}
 // expected-warning@+1 {{expected quoted string after equals sign}}
 /// <a href=>
 int test_html1(int);
 
+// expected-warning@+2 {{HTML tag 'a' requires an end tag}}
 // expected-warning@+1 {{expected quoted string after equals sign}}
 /// <a href==>
 int test_html2(int);
 
+// expected-warning@+3 {{HTML tag 'a' requires an end tag}}
 // expected-warning@+2 {{expected quoted string after equals sign}}
 // expected-warning@+1 {{HTML start tag prematurely ended, expected attribute name or '>'}}
 /// <a href= blah
 int test_html3(int);
 
+// expected-warning@+2 {{HTML tag 'a' requires an end tag}}
 // expected-warning@+1 {{HTML start tag prematurely ended, expected attribute name or '>'}}
 /// <a =>
 int test_html4(int);
 
+// expected-warning@+2 {{HTML tag 'a' requires an end tag}}
 // expected-warning@+1 {{HTML start tag prematurely ended, expected attribute name or '>'}}
 /// <a "aaa">
 int test_html5(int);
 
+// expected-warning@+2 {{HTML tag 'a' requires an end tag}}
 // expected-warning@+1 {{HTML start tag prematurely ended, expected attribute name or '>'}}
 /// <a a="b" =>
 int test_html6(int);
 
+// expected-warning@+2 {{HTML tag 'a' requires an end tag}}
 // expected-warning@+1 {{HTML start tag prematurely ended, expected attribute name or '>'}}
 /// <a a="b" "aaa">
 int test_html7(int);
 
+// expected-warning@+2 {{HTML tag 'a' requires an end tag}}
 // expected-warning@+1 {{HTML start tag prematurely ended, expected attribute name or '>'}}
 /// <a a="b" =
 int test_html8(int);
@@ -67,6 +75,8 @@ int test_html_nesting3(int);
 /// Bbb</p>
 int test_html_nesting4(int);
 
+// expected-warning@+3 {{HTML tag 'b' requires an end tag}}
+// expected-warning@+2 {{HTML tag 'i' requires an end tag}}
 // expected-warning@+1 {{HTML end tag does not match any start tag}}
 /// <b><i>Meow</a>
 int test_html_nesting5(int);
@@ -81,6 +91,9 @@ int test_html_nesting6(int);
 /// <b><i>Meow</b></i>
 int test_html_nesting7(int);
 
+// expected-warning@+1 {{HTML tag 'b' requires an end tag}}
+/// <b>Meow
+int test_html_nesting8(int);
 
 // expected-warning@+1 {{empty paragraph passed to '\brief' command}}
 /// \brief\returns Aaa

diff --git a/utils/TableGen/ClangCommentHTMLTagsEmitter.cpp b/utils/TableGen/ClangCommentHTMLTagsEmitter.cpp
index bfcd2cfd15a438532deffc1e232775cafcc6f93f..2a9110f8eba7e78dfc4ad8a4dbd873ad735de46f 100644 (file)
--- a/utils/TableGen/ClangCommentHTMLTagsEmitter.cpp
+++ b/utils/TableGen/ClangCommentHTMLTagsEmitter.cpp
@@ -12,6 +12,7 @@
 //
 //===----------------------------------------------------------------------===//
 
+#include "TableGenBackends.h"
 #include "llvm/TableGen/Record.h"
 #include "llvm/TableGen/StringMatcher.h"
 #include "llvm/TableGen/TableGenBackend.h"
@@ -19,14 +20,11 @@
 
 using namespace llvm;
 
-namespace clang {
-void EmitClangCommentHTMLTags(RecordKeeper &Records, raw_ostream &OS) {
+void clang::EmitClangCommentHTMLTags(RecordKeeper &Records, raw_ostream &OS) {
   std::vector<Record *> Tags = Records.getAllDerivedDefinitions("Tag");
   std::vector<StringMatcher::StringPair> Matches;
-  for (std::vector<Record *>::iterator I = Tags.begin(), E = Tags.end();
-       I != E; ++I) {
-    Record &Tag = **I;
-    std::string Spelling = Tag.getValueAsString("Spelling");
+  for (Record *Tag : Tags) {
+    std::string Spelling = Tag->getValueAsString("Spelling");
     Matches.push_back(StringMatcher::StringPair(Spelling, "return true;"));
   }
 
@@ -38,19 +36,17 @@ void EmitClangCommentHTMLTags(RecordKeeper &Records, raw_ostream &OS) {
      << "}\n\n";
 }
 
-void EmitClangCommentHTMLTagsProperties(RecordKeeper &Records,
-                                        raw_ostream &OS) {
+void clang::EmitClangCommentHTMLTagsProperties(RecordKeeper &Records,
+                                               raw_ostream &OS) {
   std::vector<Record *> Tags = Records.getAllDerivedDefinitions("Tag");
   std::vector<StringMatcher::StringPair> MatchesEndTagOptional;
   std::vector<StringMatcher::StringPair> MatchesEndTagForbidden;
-  for (std::vector<Record *>::iterator I = Tags.begin(), E = Tags.end();
-       I != E; ++I) {
-    Record &Tag = **I;
-    std::string Spelling = Tag.getValueAsString("Spelling");
+  for (Record *Tag : Tags) {
+    std::string Spelling = Tag->getValueAsString("Spelling");
     StringMatcher::StringPair Match(Spelling, "return true;");
-    if (Tag.getValueAsBit("EndTagOptional"))
+    if (Tag->getValueAsBit("EndTagOptional"))
       MatchesEndTagOptional.push_back(Match);
-    if (Tag.getValueAsBit("EndTagForbidden"))
+    if (Tag->getValueAsBit("EndTagForbidden"))
       MatchesEndTagForbidden.push_back(Match);
   }
 
@@ -65,6 +61,21 @@ void EmitClangCommentHTMLTagsProperties(RecordKeeper &Records,
   StringMatcher("Name", MatchesEndTagForbidden, OS).Emit();
   OS << "  return false;\n"
      << "}\n\n";
+
+  std::vector<Record *> Attributes =
+      Records.getAllDerivedDefinitions("Attribute");
+  std::vector<StringMatcher::StringPair> Matches;
+  for (Record *Attribute : Attributes) {
+    std::string Spelling = Attribute->getValueAsString("Spelling");
+    if (!Attribute->getValueAsBit("IsSafeToPassThrough"))
+      Matches.push_back(StringMatcher::StringPair(Spelling, "return false;"));
+  }
+
+  emitSourceFileHeader("HTML attribute name matcher", OS);
+
+  OS << "bool isHTMLAttributeSafeToPassThrough(StringRef Name) {\n";
+  StringMatcher("Name", Matches, OS).Emit();
+  OS << "  return true;\n"
+     << "}\n\n";
 }
-} // end namespace clang