From 3aacb86e066fbc1dd5897b23f055e83a36118871 Mon Sep 17 00:00:00 2001
From: Henri Sivonen <hsivonen@iki.fi>
Date: Mon, 27 Feb 2012 13:57:48 +0200
Subject: [PATCH] Bug 650784 part 1 - Introduce a new API for converting HTML
 to plain text. r=smaug.

---
 content/base/public/nsContentUtils.h          | 24 ++++++++-
 content/base/src/nsContentUtils.cpp           | 44 ++++++++++++++-
 content/base/src/nsDOMParser.cpp              |  2 +-
 content/base/test/TestPlainTextSerializer.cpp | 28 ++--------
 parser/html/Makefile.in                       |  3 ++
 parser/html/nsHtml5StringParser.cpp           |  5 +-
 parser/html/nsHtml5StringParser.h             |  3 +-
 parser/html/nsIParserUtils.idl                | 25 +++++++++
 toolkit/components/build/nsToolkitCompsCID.h  |  6 +++
 .../components/build/nsToolkitCompsModule.cpp |  3 ++
 .../feeds/nsIScriptableUnescapeHTML.idl       | 19 +++++--
 .../feeds/nsScriptableUnescapeHTML.cpp        | 54 ++++++++-----------
 .../feeds/nsScriptableUnescapeHTML.h          |  5 +-
 widget/xpwidgets/nsHTMLFormatConverter.cpp    | 39 ++------------
 14 files changed, 160 insertions(+), 100 deletions(-)
 create mode 100644 parser/html/nsIParserUtils.idl

diff --git a/content/base/public/nsContentUtils.h b/content/base/public/nsContentUtils.h
index 16d119f81ca..009e337bcbc 100644
--- a/content/base/public/nsContentUtils.h
+++ b/content/base/public/nsContentUtils.h
@@ -1144,12 +1144,34 @@ public:
    * @param aSourceBuffer the string to parse as an HTML document
    * @param aTargetDocument the document object to parse into. Must not have
    *                        child nodes.
+   * @param aScriptingEnabledForNoscriptParsing whether <noscript> is parsed
+   *                                            as if scripting was enabled
    * @return NS_ERROR_DOM_INVALID_STATE_ERR if a re-entrant attempt to parse
    *         fragments is made, NS_ERROR_OUT_OF_MEMORY if aSourceBuffer is too
    *         long and NS_OK otherwise.
    */
   static nsresult ParseDocumentHTML(const nsAString& aSourceBuffer,
-                                    nsIDocument* aTargetDocument);
+                                    nsIDocument* aTargetDocument,
+                                    bool aScriptingEnabledForNoscriptParsing);
+
+  /**
+   * Converts HTML source to plain text by parsing the source and using the
+   * plain text serializer on the resulting tree.
+   *
+   * @param aSourceBuffer the string to parse as an HTML document
+   * @param aResultBuffer the string where the plain text result appears;
+   *                      may be the same string as aSourceBuffer
+   * @param aFlags Flags from nsIDocumentEncoder.
+   * @param aWrapCol Number of columns after which to line wrap; 0 for no
+   *                 auto-wrapping
+   * @return NS_ERROR_DOM_INVALID_STATE_ERR if a re-entrant attempt to parse
+   *         fragments is made, NS_ERROR_OUT_OF_MEMORY if aSourceBuffer is too
+   *         long and NS_OK otherwise.
+   */
+  static nsresult ConvertToPlainText(const nsAString& aSourceBuffer,
+                                     nsAString& aResultBuffer,
+                                     PRUint32 aFlags,
+                                     PRUint32 aWrapCol);
 
   /**
    * Creates a new XML document, which is marked to be loaded as data.
diff --git a/content/base/src/nsContentUtils.cpp b/content/base/src/nsContentUtils.cpp
index 7204cd9a535..3dfc604ae98 100644
--- a/content/base/src/nsContentUtils.cpp
+++ b/content/base/src/nsContentUtils.cpp
@@ -3927,7 +3927,8 @@ nsContentUtils::ParseFragmentHTML(const nsAString& aSourceBuffer,
 /* static */
 nsresult
 nsContentUtils::ParseDocumentHTML(const nsAString& aSourceBuffer,
-                                  nsIDocument* aTargetDocument)
+                                  nsIDocument* aTargetDocument,
+                                  bool aScriptingEnabledForNoscriptParsing)
 {
   if (nsContentUtils::sFragmentParsingActive) {
     NS_NOTREACHED("Re-entrant fragment parsing attempted.");
@@ -3941,7 +3942,8 @@ nsContentUtils::ParseDocumentHTML(const nsAString& aSourceBuffer,
   }
   nsresult rv =
     sHTMLFragmentParser->ParseDocument(aSourceBuffer,
-                                       aTargetDocument);
+                                       aTargetDocument,
+                                       aScriptingEnabledForNoscriptParsing);
   return rv;
 }
 
@@ -3992,6 +3994,44 @@ nsContentUtils::ParseFragmentXML(const nsAString& aSourceBuffer,
   return rv;
 }
 
+/* static */
+nsresult
+nsContentUtils::ConvertToPlainText(const nsAString& aSourceBuffer,
+                                   nsAString& aResultBuffer,
+                                   PRUint32 aFlags,
+                                   PRUint32 aWrapCol)
+{
+  nsCOMPtr<nsIURI> uri;
+  NS_NewURI(getter_AddRefs(uri), "about:blank");
+  nsCOMPtr<nsIPrincipal> principal =
+    do_CreateInstance("@mozilla.org/nullprincipal;1");
+  nsCOMPtr<nsIDOMDocument> domDocument;
+  nsresult rv = nsContentUtils::CreateDocument(EmptyString(),
+                                               EmptyString(),
+                                               nsnull,
+                                               uri,
+                                               uri,
+                                               principal,
+                                               nsnull,
+                                               DocumentFlavorHTML,
+                                               getter_AddRefs(domDocument));
+  NS_ENSURE_SUCCESS(rv, rv);
+
+  nsCOMPtr<nsIDocument> document = do_QueryInterface(domDocument);
+  rv = nsContentUtils::ParseDocumentHTML(aSourceBuffer, document,
+    !(aFlags & nsIDocumentEncoder::OutputNoScriptContent));
+  NS_ENSURE_SUCCESS(rv, rv);
+
+  nsCOMPtr<nsIDocumentEncoder> encoder = do_CreateInstance(
+    "@mozilla.org/layout/documentEncoder;1?type=text/plain");
+
+  rv = encoder->Init(domDocument, NS_LITERAL_STRING("text/plain"), aFlags);
+  NS_ENSURE_SUCCESS(rv, rv);
+
+  encoder->SetWrapColumn(aWrapCol);
+
+  return encoder->EncodeToString(aResultBuffer);
+}
 
 /* static */
 nsresult
diff --git a/content/base/src/nsDOMParser.cpp b/content/base/src/nsDOMParser.cpp
index e74cfacf33c..29a5023bb96 100644
--- a/content/base/src/nsDOMParser.cpp
+++ b/content/base/src/nsDOMParser.cpp
@@ -102,7 +102,7 @@ nsDOMParser::ParseFromString(const PRUnichar *str,
     NS_ENSURE_SUCCESS(rv, rv);
     nsCOMPtr<nsIDocument> document = do_QueryInterface(domDocument);
     nsDependentString sourceBuffer(str);
-    rv = nsContentUtils::ParseDocumentHTML(sourceBuffer, document);
+    rv = nsContentUtils::ParseDocumentHTML(sourceBuffer, document, false);
     NS_ENSURE_SUCCESS(rv, rv);
 
     // Keep the XULXBL state, base URL and principal setting in sync with the
diff --git a/content/base/test/TestPlainTextSerializer.cpp b/content/base/test/TestPlainTextSerializer.cpp
index abec0d829aa..638aa4459b6 100644
--- a/content/base/test/TestPlainTextSerializer.cpp
+++ b/content/base/test/TestPlainTextSerializer.cpp
@@ -37,37 +37,19 @@
 
 #include "TestHarness.h"
 
-#include "nsIParser.h"
-#include "nsIHTMLToTextSink.h"
-#include "nsIParser.h"
-#include "nsIContentSink.h"
-#include "nsIParserService.h"
 #include "nsServiceManagerUtils.h"
 #include "nsStringGlue.h"
-#include "nsParserCIID.h"
 #include "nsIDocumentEncoder.h"
 #include "nsCRT.h"
-
-static NS_DEFINE_CID(kCParserCID, NS_PARSER_CID);
+#include "nsIParserUtils.h"
+#include "nsToolkitCompsCID.h"
 
 void
 ConvertBufToPlainText(nsString &aConBuf, int aFlag)
 {
-  nsCOMPtr<nsIParser> parser = do_CreateInstance(kCParserCID);
-  if (parser) {
-    nsCOMPtr<nsIContentSink> sink;
-    sink = do_CreateInstance(NS_PLAINTEXTSINK_CONTRACTID);
-    if (sink) {
-      nsCOMPtr<nsIHTMLToTextSink> textSink(do_QueryInterface(sink));
-      if (textSink) {
-        nsAutoString convertedText;
-        textSink->Initialize(&convertedText, aFlag, 72);
-        parser->SetContentSink(sink);
-        parser->Parse(aConBuf, 0, NS_LITERAL_CSTRING("text/html"), true);
-        aConBuf = convertedText;
-      }
-    }
-  }
+  nsCOMPtr<nsIParserUtils> utils =
+    do_GetService(NS_PARSERUTILS_CONTRACTID);
+  utils->ConvertToPlainText(aConBuf, aFlag, 72, aConBuf);
 }
 
 // Test for ASCII with format=flowed; delsp=yes
diff --git a/parser/html/Makefile.in b/parser/html/Makefile.in
index 105a6d74add..4ea25fac738 100644
--- a/parser/html/Makefile.in
+++ b/parser/html/Makefile.in
@@ -45,6 +45,9 @@ MODULE		= html5
 LIBRARY_NAME	= html5p_s
 LIBXUL_LIBRARY	= 1
 
+XPIDLSRCS = \
+  nsIParserUtils.idl \
+  $(NULL)
 
 EXPORTS		= \
 		jArray.h \
diff --git a/parser/html/nsHtml5StringParser.cpp b/parser/html/nsHtml5StringParser.cpp
index 91ffcabb572..dcf134f577f 100644
--- a/parser/html/nsHtml5StringParser.cpp
+++ b/parser/html/nsHtml5StringParser.cpp
@@ -102,7 +102,8 @@ nsHtml5StringParser::ParseFragment(const nsAString& aSourceBuffer,
 
 nsresult
 nsHtml5StringParser::ParseDocument(const nsAString& aSourceBuffer,
-                                   nsIDocument* aTargetDoc)
+                                   nsIDocument* aTargetDoc,
+                                   bool aScriptingEnabledForNoscriptParsing)
 {
   MOZ_ASSERT(!aTargetDoc->GetFirstChild());
 
@@ -116,7 +117,7 @@ nsHtml5StringParser::ParseDocument(const nsAString& aSourceBuffer,
 
   mExecutor->PreventScriptExecution();
 
-  Tokenize(aSourceBuffer, aTargetDoc, false);
+  Tokenize(aSourceBuffer, aTargetDoc, aScriptingEnabledForNoscriptParsing);
   return NS_OK;
 }
 
diff --git a/parser/html/nsHtml5StringParser.h b/parser/html/nsHtml5StringParser.h
index 4634ff30151..c186297104e 100644
--- a/parser/html/nsHtml5StringParser.h
+++ b/parser/html/nsHtml5StringParser.h
@@ -86,7 +86,8 @@ class nsHtml5StringParser : public nsParserBase
      *
      */
     nsresult ParseDocument(const nsAString& aSourceBuffer,
-                           nsIDocument* aTargetDoc);
+                           nsIDocument* aTargetDoc,
+                           bool aScriptingEnabledForNoscriptParsing);
 
   private:
 
diff --git a/parser/html/nsIParserUtils.idl b/parser/html/nsIParserUtils.idl
new file mode 100644
index 00000000000..53d310ffec6
--- /dev/null
+++ b/parser/html/nsIParserUtils.idl
@@ -0,0 +1,25 @@
+/* This Source Code Form is subject to the terms of the Mozilla Public
+ * License, v. 2.0. If a copy of the MPL was not distributed with this file,
+ * You can obtain one at http://mozilla.org/MPL/2.0/. */
+
+#include "nsISupports.idl"
+
+/**
+ * Non-Web HTML parser functionality to Firefox extensions and XULRunner apps. 
+ * Don't use this from within Gecko--use nsContentUtils directly instead.
+ */
+[scriptable, uuid(290f49bb-0619-4bda-8006-ab31bec7231a)]
+interface nsIParserUtils : nsISupports
+{
+  /**
+   * Convert HTML to plain text.
+   *
+   * @param src the HTML source to parse (C++ callers are allowed but not
+   *            required to use the same string for the return value.)
+   * @param flags conversion option flags defined in nsIDocumentEncoder
+   * @param wrapCol number of characters per line; 0 for no auto-wrapping
+   */
+  AString convertToPlainText(in AString src,
+                             in unsigned long flags,
+                             in unsigned long wrapCol);
+};
diff --git a/toolkit/components/build/nsToolkitCompsCID.h b/toolkit/components/build/nsToolkitCompsCID.h
index c24332cf7e2..d61661ace70 100644
--- a/toolkit/components/build/nsToolkitCompsCID.h
+++ b/toolkit/components/build/nsToolkitCompsCID.h
@@ -96,6 +96,9 @@
 #define NS_URLCLASSIFIERHASHCOMPLETER_CONTRACTID \
     "@mozilla.org/url-classifier/hashcompleter;1"
 
+#define NS_PARSERUTILS_CONTRACTID \
+    "@mozilla.org/parserutils;1"
+
 #define NS_SCRIPTABLEUNESCAPEHTML_CONTRACTID "@mozilla.org/feed-unescapehtml;1"
 
 #define NS_NAVHISTORYSERVICE_CONTRACTID \
@@ -179,6 +182,9 @@
 #define NS_URLCLASSIFIERUTILS_CID \
 { 0xb7b2ccec, 0x7912, 0x4ea6, { 0xa5, 0x48, 0xb0, 0x38, 0x44, 0x70, 0x04, 0xbd} }
 
+#define NS_PARSERUTILS_CID  \
+{ 0xaf7b24cb, 0x893f, 0x41bb, { 0x96, 0x1f, 0x5a, 0x69, 0x38, 0x8e, 0x27, 0xc3 } }
+
 // {10f2f5f0-f103-4901-980f-ba11bd70d60d}
 #define NS_SCRIPTABLEUNESCAPEHTML_CID  \
 { 0x10f2f5f0, 0xf103, 0x4901, { 0x98, 0x0f, 0xba, 0x11, 0xbd, 0x70, 0xd6, 0x0d} }
diff --git a/toolkit/components/build/nsToolkitCompsModule.cpp b/toolkit/components/build/nsToolkitCompsModule.cpp
index 9bbbc6e1bba..d0e9078e006 100644
--- a/toolkit/components/build/nsToolkitCompsModule.cpp
+++ b/toolkit/components/build/nsToolkitCompsModule.cpp
@@ -140,6 +140,7 @@ NS_DEFINE_NAMED_CID(NS_URLCLASSIFIERSTREAMUPDATER_CID);
 NS_DEFINE_NAMED_CID(NS_URLCLASSIFIERUTILS_CID);
 #endif
 #ifdef MOZ_FEEDS
+NS_DEFINE_NAMED_CID(NS_PARSERUTILS_CID);
 NS_DEFINE_NAMED_CID(NS_SCRIPTABLEUNESCAPEHTML_CID);
 #endif
 NS_DEFINE_NAMED_CID(NS_BROWSERSTATUSFILTER_CID);
@@ -165,6 +166,7 @@ static const mozilla::Module::CIDEntry kToolkitCIDs[] = {
   { &kNS_URLCLASSIFIERUTILS_CID, false, NULL, nsUrlClassifierUtilsConstructor },
 #endif
 #ifdef MOZ_FEEDS
+  { &kNS_PARSERUTILS_CID, false, NULL, nsScriptableUnescapeHTMLConstructor },
   { &kNS_SCRIPTABLEUNESCAPEHTML_CID, false, NULL, nsScriptableUnescapeHTMLConstructor },
 #endif
   { &kNS_BROWSERSTATUSFILTER_CID, false, NULL, nsBrowserStatusFilterConstructor },
@@ -193,6 +195,7 @@ static const mozilla::Module::ContractIDEntry kToolkitContracts[] = {
   { NS_URLCLASSIFIERUTILS_CONTRACTID, &kNS_URLCLASSIFIERUTILS_CID },
 #endif
 #ifdef MOZ_FEEDS
+  { NS_PARSERUTILS_CONTRACTID, &kNS_PARSERUTILS_CID },
   { NS_SCRIPTABLEUNESCAPEHTML_CONTRACTID, &kNS_SCRIPTABLEUNESCAPEHTML_CID },
 #endif
   { NS_BROWSERSTATUSFILTER_CONTRACTID, &kNS_BROWSERSTATUSFILTER_CID },
diff --git a/toolkit/components/feeds/nsIScriptableUnescapeHTML.idl b/toolkit/components/feeds/nsIScriptableUnescapeHTML.idl
index 59ee8b4bf8f..942dc7327ca 100644
--- a/toolkit/components/feeds/nsIScriptableUnescapeHTML.idl
+++ b/toolkit/components/feeds/nsIScriptableUnescapeHTML.idl
@@ -41,20 +41,31 @@ interface nsIDOMDocumentFragment;
 interface nsIURI;
 
 /**
- * A utility class that unescapes HTML strings.
+ * A utility class for HTML parsing in the feed processor.
  */
 [scriptable, uuid(3ab244a9-f09d-44da-9e3f-ee4d67367f2d)]
 interface nsIScriptableUnescapeHTML : nsISupports 
 {
   /** 
-   * Converts all entities to Unicode.
+   * Converts HTML to plain text. This is equivalent to calling
+   * nsIParserUtils::convertToPlainText(src, 
+   *   nsIDocumentEncoder::OutputSelectionOnly |
+   *   nsIDocumentEncoder::OutputAbsoluteLinks, 0).
    *
-   * @param src The HTML string to escape.
+   * You should most likely call nsIParserUtils::convertToPlainText()
+   * instead of calling this method.
+   *
+   * @param src The HTML string to convert to plain text.
    */ 
   AString unescape(in AString src);
         
   /**
-   * Appends the text to the element.
+   * Parses markup into a sanitized document fragment.
+   *
+   * @param fragment the input markup
+   * @param isXML true if |fragment| is XML and false if HTML
+   * @param baseURI the base URL for this fragment
+   * @param element the context node for the fragment parsing algorithm
    */
   nsIDOMDocumentFragment parseFragment(in AString fragment,
                                        in boolean isXML,
diff --git a/toolkit/components/feeds/nsScriptableUnescapeHTML.cpp b/toolkit/components/feeds/nsScriptableUnescapeHTML.cpp
index db7ef2646d4..1fa5fe210a3 100644
--- a/toolkit/components/feeds/nsScriptableUnescapeHTML.cpp
+++ b/toolkit/components/feeds/nsScriptableUnescapeHTML.cpp
@@ -49,7 +49,6 @@
 #include "nsParserCIID.h"
 #include "nsContentUtils.h"
 #include "nsIContentSink.h"
-#include "nsIHTMLToTextSink.h"
 #include "nsIDocumentEncoder.h"
 #include "nsIDOMDocumentFragment.h"
 #include "nsIFragmentContentSink.h"
@@ -70,42 +69,35 @@
 
 #define XHTML_DIV_TAG "div xmlns=\"http://www.w3.org/1999/xhtml\""
 
-NS_IMPL_ISUPPORTS1(nsScriptableUnescapeHTML, nsIScriptableUnescapeHTML)
+NS_IMPL_ISUPPORTS2(nsScriptableUnescapeHTML,
+                   nsIScriptableUnescapeHTML,
+                   nsIParserUtils)
 
 static NS_DEFINE_CID(kCParserCID, NS_PARSER_CID);
 
-// From /widget/HTMLConverter
-//
-// Takes HTML and converts it to plain text but in unicode.
-//
+
+
 NS_IMETHODIMP
-nsScriptableUnescapeHTML::Unescape(const nsAString & aFromStr, 
+nsScriptableUnescapeHTML::ConvertToPlainText(const nsAString & aFromStr,
+                                             PRUint32 aFlags,
+                                             PRUint32 aWrapCol,
+                                             nsAString & aToStr)
+{
+  return nsContentUtils::ConvertToPlainText(aFromStr,
+    aToStr,
+    aFlags,
+    aWrapCol);
+}
+
+NS_IMETHODIMP
+nsScriptableUnescapeHTML::Unescape(const nsAString & aFromStr,
                                    nsAString & aToStr)
 {
-  // create the parser to do the conversion.
-  aToStr.SetLength(0);
-  nsresult rv;
-  nsCOMPtr<nsIParser> parser = do_CreateInstance(kCParserCID, &rv);
-  if (NS_FAILED(rv)) return rv;
-
-  // convert it!
-  nsCOMPtr<nsIContentSink> sink;
-
-  sink = do_CreateInstance(NS_PLAINTEXTSINK_CONTRACTID);
-  NS_ENSURE_TRUE(sink, NS_ERROR_FAILURE);
-
-  nsCOMPtr<nsIHTMLToTextSink> textSink(do_QueryInterface(sink));
-  NS_ENSURE_TRUE(textSink, NS_ERROR_FAILURE);
-
-  textSink->Initialize(&aToStr, nsIDocumentEncoder::OutputSelectionOnly
-                       | nsIDocumentEncoder::OutputAbsoluteLinks, 0);
-
-  parser->SetContentSink(sink);
-
-  parser->Parse(aFromStr, 0, NS_LITERAL_CSTRING("text/html"),
-                true, eDTDMode_fragment);
-
-  return NS_OK;
+  return nsContentUtils::ConvertToPlainText(aFromStr,
+    aToStr,
+    nsIDocumentEncoder::OutputSelectionOnly |
+    nsIDocumentEncoder::OutputAbsoluteLinks,
+    0);
 }
 
 // The feed version of nsContentUtils::CreateContextualFragment It
diff --git a/toolkit/components/feeds/nsScriptableUnescapeHTML.h b/toolkit/components/feeds/nsScriptableUnescapeHTML.h
index bea14c5a2e4..6e08cf7bf5b 100644
--- a/toolkit/components/feeds/nsScriptableUnescapeHTML.h
+++ b/toolkit/components/feeds/nsScriptableUnescapeHTML.h
@@ -38,12 +38,15 @@
 #define nsScriptableHTMLUnescape_h__
 
 #include "nsIScriptableUnescapeHTML.h"
+#include "nsIParserUtils.h"
 
-class nsScriptableUnescapeHTML : public nsIScriptableUnescapeHTML
+class nsScriptableUnescapeHTML : public nsIScriptableUnescapeHTML,
+                                 public nsIParserUtils
 {
 public:
   NS_DECL_ISUPPORTS
   NS_DECL_NSISCRIPTABLEUNESCAPEHTML
+  NS_DECL_NSIPARSERUTILS
 };
 
 #endif // nsScriptableHTMLUnescape_h__
diff --git a/widget/xpwidgets/nsHTMLFormatConverter.cpp b/widget/xpwidgets/nsHTMLFormatConverter.cpp
index c4215675e0e..2a24cdd88d7 100644
--- a/widget/xpwidgets/nsHTMLFormatConverter.cpp
+++ b/widget/xpwidgets/nsHTMLFormatConverter.cpp
@@ -49,15 +49,9 @@
 #include "nsITransferable.h" // for mime defs, this is BAD
 
 // HTML convertor stuff
-#include "nsIParser.h"
-#include "nsIDTD.h"
-#include "nsParserCIID.h"
-#include "nsIContentSink.h"
 #include "nsPrimitiveHelpers.h"
 #include "nsIDocumentEncoder.h"
-#include "nsIHTMLToTextSink.h"
-
-static NS_DEFINE_CID(kCParserCID, NS_PARSER_CID);
+#include "nsContentUtils.h"
 
 nsHTMLFormatConverter::nsHTMLFormatConverter()
 {
@@ -272,36 +266,13 @@ nsHTMLFormatConverter::Convert(const char *aFromDataFlavor, nsISupports *aFromDa
 NS_IMETHODIMP
 nsHTMLFormatConverter::ConvertFromHTMLToUnicode(const nsAutoString & aFromStr, nsAutoString & aToStr)
 {
-  // create the parser to do the conversion.
-  aToStr.SetLength(0);
-  nsresult rv;
-  nsCOMPtr<nsIParser> parser = do_CreateInstance(kCParserCID, &rv);
-  if ( !parser )
-    return rv;
-
-  // convert it!
-  nsCOMPtr<nsIContentSink> sink;
-
-  sink = do_CreateInstance(NS_PLAINTEXTSINK_CONTRACTID);
-  NS_ENSURE_TRUE(sink, NS_ERROR_FAILURE);
-
-  nsCOMPtr<nsIHTMLToTextSink> textSink(do_QueryInterface(sink));
-  NS_ENSURE_TRUE(textSink, NS_ERROR_FAILURE);
-
-  // We set OutputNoScriptContent and OutputNoFramesContent unconditionally
-  // here because |aFromStr| is already filtered based on user preferences.
-  PRUint32 flags =
+  return nsContentUtils::ConvertToPlainText(aFromStr,
+    aToStr,
     nsIDocumentEncoder::OutputSelectionOnly |
     nsIDocumentEncoder::OutputAbsoluteLinks |
     nsIDocumentEncoder::OutputNoScriptContent |
-    nsIDocumentEncoder::OutputNoFramesContent;
-  textSink->Initialize(&aToStr, flags, 0);
-
-  parser->SetContentSink(sink);
-
-  parser->Parse(aFromStr, 0, NS_LITERAL_CSTRING("text/html"), true, eDTDMode_fragment);
-  
-  return NS_OK;
+    nsIDocumentEncoder::OutputNoFramesContent,
+    0);
 } // ConvertFromHTMLToUnicode