Bug 1399911 - preserve sourceURL comment directive on style sheets; r=bz,heycam

In addition to the sourceMappingURL comment, there is a second special
comment, "sourceURL", that can be used to set the "display name" of a
style sheet for developer tools.  This name is also used as the base
URL for the source-map URL resolution algorithm.  sourceURL is
described here:
https://blog.getfirebug.com/2009/08/11/give-your-eval-a-name-with-sourceurl/

This patch changes Firefox to record this URL, if specified, and to
expose it (chrome-only) vai StyleSheet.webidl.

MozReview-Commit-ID: 7NwXsOf7nbY

--HG--
extra : rebase_source : bd5d25b4d44f5f220a4624db346edbc4236c9886

dom/webidl/StyleSheet.webidl

@@ -39,4 +39,10 @@ interface StyleSheet {
   // then this is an empty string.
   [ChromeOnly, Pure]
   readonly attribute DOMString sourceMapURL;
+  // The source URL for this style sheet.  If the style sheet has the
+  // special "# sourceURL=" comment, then this is the URL specified
+  // there.  If no such comment is found, then this is the empty
+  // string.
+  [ChromeOnly, Pure]
+  readonly attribute DOMString sourceURL;
 };

layout/style/ServoBindingList.h

@@ -68,6 +68,8 @@ SERVO_BINDING_FUNC(Servo_StyleSheet_SizeOfIncludingThis, size_t,
                    RawServoStyleSheetContentsBorrowed sheet)
 SERVO_BINDING_FUNC(Servo_StyleSheet_GetSourceMapURL, void,
                    RawServoStyleSheetContentsBorrowed sheet, nsAString* result)
+SERVO_BINDING_FUNC(Servo_StyleSheet_GetSourceURL, void,
+                   RawServoStyleSheetContentsBorrowed sheet, nsAString* result)
 // We'd like to return `OriginFlags` here, but bindgen bitfield enums don't
 // work as return values with the Linux 32-bit ABI at the moment because
 // they wrap the value in a struct.

layout/style/ServoStyleSheet.cpp

@@ -225,6 +225,10 @@ ServoStyleSheet::ParseSheet(css::Loader* aLoader,
   Servo_StyleSheet_GetSourceMapURL(Inner()->mContents, &sourceMapURL);
   SetSourceMapURLFromComment(sourceMapURL);
 
+  nsString sourceURL;
+  Servo_StyleSheet_GetSourceURL(Inner()->mContents, &sourceURL);
+  SetSourceURL(sourceURL);
+
   Inner()->mURLData = extraData.forget();
   return NS_OK;
 }

layout/style/StyleSheet.cpp

@@ -261,6 +261,7 @@ StyleSheetInfo::StyleSheetInfo(StyleSheetInfo& aCopy,
                    // without children.
   , mSourceMapURL(aCopy.mSourceMapURL)
   , mSourceMapURLFromComment(aCopy.mSourceMapURLFromComment)
+  , mSourceURL(aCopy.mSourceURL)
 #ifdef DEBUG
   , mPrincipalSet(aCopy.mPrincipalSet)
 #endif
@@ -530,6 +531,18 @@ StyleSheet::SetSourceMapURLFromComment(const nsAString& aSourceMapURLFromComment
   mInner->mSourceMapURLFromComment = aSourceMapURLFromComment;
 }
 
+void
+StyleSheet::GetSourceURL(nsAString& aSourceURL)
+{
+  aSourceURL = mInner->mSourceURL;
+}
+
+void
+StyleSheet::SetSourceURL(const nsAString& aSourceURL)
+{
+  mInner->mSourceURL = aSourceURL;
+}
+
 css::Rule*
 StyleSheet::GetDOMOwnerRule() const
 {

layout/style/StyleSheet.h

@@ -216,6 +216,8 @@ public:
   void GetSourceMapURL(nsAString& aTitle);
   void SetSourceMapURL(const nsAString& aSourceMapURL);
   void SetSourceMapURLFromComment(const nsAString& aSourceMapURLFromComment);
+  void GetSourceURL(nsAString& aSourceURL);
+  void SetSourceURL(const nsAString& aSourceURL);
 
   // WebIDL CSSStyleSheet API
   // Can't be inline because we can't include ImportRule here.  And can't be

layout/style/StyleSheetInfo.h

@@ -70,6 +70,9 @@ struct StyleSheetInfo
   // so that the value does not overwrite any value that might have
   // come from a response header.
   nsString mSourceMapURLFromComment;
+  // This stores any source URL that might have been seen in a comment
+  // in the style sheet.
+  nsString mSourceURL;
 
 #ifdef DEBUG
   bool                   mPrincipalSet;

layout/style/nsCSSParser.cpp

@@ -1648,6 +1648,7 @@ CSSParserImpl::ParseSheet(const nsAString& aInput,
   }
 
   mSheet->SetSourceMapURLFromComment(scanner.GetSourceMapURL());
+  mSheet->SetSourceURL(scanner.GetSourceURL());
   ReleaseScanner();
 
   mParsingMode = css::eAuthorSheetFeatures;

layout/style/nsCSSScanner.cpp

@@ -542,29 +542,52 @@ nsCSSScanner::SkipWhitespace()
   }
 }
 
+/**
+ * If the given text appears at the current offset in the buffer,
+ * advance over the text and return true.  Otherwise, return false.
+ * mLength is the number of characters in mDirective.
+ */
+bool
+nsCSSScanner::CheckCommentDirective(const nsAString& aDirective)
+{
+  nsDependentSubstring text(&mBuffer[mOffset], &mBuffer[mCount]);
+
+  if (StringBeginsWith(text, aDirective)) {
+    Advance(aDirective.Length());
+    return true;
+  }
+  return false;
+}
+
 /**
  * Skip over one CSS comment starting at the current read position.
  */
 void
 nsCSSScanner::SkipComment()
 {
-  static const char sourceMappingURLDirective[] = "# sourceMappingURL=";
+  // Note that these do not start with "#" or "@" -- that is handled
+  // separately, below.
+  static NS_NAMED_LITERAL_STRING(kSourceMappingURLDirective, " sourceMappingURL=");
+  static NS_NAMED_LITERAL_STRING(kSourceURLDirective, " sourceURL=");
 
   MOZ_ASSERT(Peek() == '/' && Peek(1) == '*', "should not have been called");
   Advance(2);
-  // Look in each comment for a source map directive; using a simple
-  // state machine.  The states are:
-  // * sourceMapIndex >= 0 means that we're still looking for the
-  //   directive and expect the next character to be at that index of
-  //   sourceMappingURLDirective.
-  //   As a special case, when sourceMapIndex == 0, '@' is also recognized.
-  // * sourceMapIndex < 0 means that we don't need to look for the
-  //   directive any more -- whether it was found or not.
-  // * copying == true means that the directive was found and we're
-  //   copying characters into mSourceMapURL.  This stops at the first
-  //   whitespace, or at the end of the comment.
-  int sourceMapIndex = 0;
-  bool copying = false;
+
+  // If we saw one of the directives, this will be non-NULL and will
+  // point to the string into which the URL will be written.
+  nsString* directive = nullptr;
+  if (Peek() == '#' || Peek() == '@') {
+    // Check for the comment directives.
+    Advance();
+    if (CheckCommentDirective(kSourceMappingURLDirective)) {
+      mSourceMapURL.Truncate();
+      directive = &mSourceMapURL;
+    } else if (CheckCommentDirective(kSourceURLDirective)) {
+      mSourceURL.Truncate();
+      directive = &mSourceURL;
+    }
+  }
+
   for (;;) {
     int32_t ch = Peek();
     if (ch < 0) {
@@ -573,22 +596,6 @@ nsCSSScanner::SkipComment()
       SetEOFCharacters(eEOFCharacters_Asterisk | eEOFCharacters_Slash);
       return;
     }
-    if (sourceMapIndex >= 0) {
-      if ((sourceMapIndex == 0 && ch == '@') || ch == sourceMappingURLDirective[sourceMapIndex]) {
-        ++sourceMapIndex;
-        if (sourceMappingURLDirective[sourceMapIndex] == '\0') {
-          sourceMapIndex = -1;
-          mSourceMapURL.Truncate();
-          copying = true;
-          Advance();
-          // Make sure we don't copy out the '=' by falling through.
-          continue;
-        }
-      } else {
-        // Did not see the directive.
-        sourceMapIndex = -1;
-      }
-    }
 
     if (ch == '*') {
       Advance();
@@ -605,20 +612,20 @@ nsCSSScanner::SkipComment()
         Advance();
         return;
       }
-      if (copying) {
-        mSourceMapURL.Append('*');
+      if (directive != nullptr) {
+        directive->Append('*');
       }
     } else if (IsVertSpace(ch)) {
       AdvanceLine();
       // Done with the directive, so stop copying.
-      copying = false;
+      directive = nullptr;
     } else if (IsWhitespace(ch)) {
       Advance();
       // Done with the directive, so stop copying.
-      copying = false;
+      directive = nullptr;
     } else {
-      if (copying) {
-        mSourceMapURL.Append(ch);
+      if (directive != nullptr) {
+        directive->Append(ch);
       }
       Advance();
     }

layout/style/nsCSSScanner.h

@@ -235,6 +235,9 @@ class nsCSSScanner {
   const nsAString& GetSourceMapURL() const
   { return mSourceMapURL; }
 
+  const nsAString& GetSourceURL() const
+  { return mSourceURL; }
+
   // Get the text of the line containing the first character of
   // the most recently processed token.
   nsDependentSubstring GetCurrentLine() const;
@@ -330,6 +333,7 @@ protected:
   void AdvanceLine();
 
   void SkipWhitespace();
+  bool CheckCommentDirective(const nsAString& aDirective);
   void SkipComment();
 
   bool GatherEscape(nsString& aOutput, bool aInString);
@@ -366,6 +370,7 @@ protected:
   bool mSeenVariableReference;
 
   nsString mSourceMapURL;
+  nsString mSourceURL;
 };
 
 // Token for the grid-template-areas micro-syntax

layout/style/test/browser.ini

@@ -14,3 +14,4 @@ support-files =
 skip-if = stylo # Gecko-specific test
 [browser_sourcemap.js]
 [browser_sourcemap_comment.js]
+[browser_sourceurl_comment.js]

layout/style/test/browser_sourceurl_comment.js

@@ -0,0 +1,40 @@
+add_task(async function() {
+  // Test text and expected results.
+  let test_cases = [
+    ["/*# sourceURL=here*/", "here"],
+    ["/*# sourceURL=here  */", "here"],
+    ["/*@ sourceURL=here*/", "here"],
+    ["/*@ sourceURL=there*/ /*# sourceURL=here*/", "here"],
+    ["/*# sourceURL=here there  */", "here"],
+
+    ["/*# sourceURL=  here  */", ""],
+    ["/*# sourceURL=*/", ""],
+    ["/*# sourceUR=here  */", ""],
+    ["/*! sourceURL=here  */", ""],
+    ["/*# sourceURL = here  */", ""],
+    ["/*   # sourceURL=here   */", ""],
+  ];
+
+  let page = "<!DOCTYPE HTML>\n<html>\n<head>\n";
+  for (let i = 0; i < test_cases.length; ++i) {
+    page += `<style type="text/css"> #x${i} { color: red; }${test_cases[i][0]}</style>\n`;
+  }
+  page += "</head><body>some text</body></html>";
+
+  let uri = "data:text/html;base64," + btoa(page);
+  info(`URI is ${uri}`);
+
+  await BrowserTestUtils.withNewTab({
+    gBrowser,
+    url: uri
+  }, async function(browser) {
+    await ContentTask.spawn(browser, test_cases, function* (tests) {
+      for (let i = 0; i < content.document.styleSheets.length; ++i) {
+        let sheet = content.document.styleSheets[i];
+
+        info(`Checking sheet #${i}`);
+        is(sheet.sourceURL, tests[i][1], `correct source URL for sheet ${i}`);
+      }
+    });
+  });
+});