Bug 196078 - Part 2: Support displaying arbitrary text/* MIME types as plain text, r=smaug,necko-reviewers,valentin

This patch refactors how we check for text formats when deciding how to handle
resources, such that more text MIME types will be rendered in-browser, rather
than downloaded.

This change requires us to move more away from using the Gecko-Content-Viewers
category in the category manager for this decision, as we need to handle an
unlimited number of MIME types behind the scenes.

Support for Gecko-Content-Viewers was left in for both the in-tree use for
application/http-index-format and dynamically determining whether image/avif
and image/jxl are supported, as well as for the message/rfc822 type used by
Thunderbird.

Differential Revision: https://phabricator.services.mozilla.com/D212078

docshell/base/nsIDocumentLoaderFactory.idl

@@ -17,10 +17,9 @@ webidl Document;
 
 /**
  * To get a component that implements nsIDocumentLoaderFactory
- * for a given mimetype, use nsICategoryManager to find an entry
+ * for a given mimetype, use nsContentUtils::FindInternalDocumentViewer.
- * with the mimetype as its name in the category "Gecko-Content-Viewers".
+ * This will look up the MIME type within the "Gecko-Content-Viewers" category,
- * The value of the entry is the contractid of the component.
+ * with additional handlers for other content types.
- * The component is a service, so use GetService, not CreateInstance to get it.
  */
 
 [scriptable, uuid(e795239e-9d3c-47c4-b063-9e600fb3b287)]

dom/base/Document.cpp

@@ -1081,21 +1081,13 @@ nsresult ExternalResourceMap::PendingLoad::SetupViewer(
       new LoadgroupCallbacks(callbacks);
   newLoadGroup->SetNotificationCallbacks(newCallbacks);
 
-  // This is some serious hackery cribbed from docshell
-  nsCOMPtr<nsICategoryManager> catMan =
-      do_GetService(NS_CATEGORYMANAGER_CONTRACTID);
-  NS_ENSURE_TRUE(catMan, NS_ERROR_NOT_AVAILABLE);
-  nsCString contractId;
-  nsresult rv =
-      catMan->GetCategoryEntry("Gecko-Content-Viewers", type, contractId);
-  NS_ENSURE_SUCCESS(rv, rv);
   nsCOMPtr<nsIDocumentLoaderFactory> docLoaderFactory =
-      do_GetService(contractId.get());
+      nsContentUtils::FindInternalDocumentViewer(type);
   NS_ENSURE_TRUE(docLoaderFactory, NS_ERROR_NOT_AVAILABLE);
 
   nsCOMPtr<nsIDocumentViewer> viewer;
   nsCOMPtr<nsIStreamListener> listener;
-  rv = docLoaderFactory->CreateInstance(
+  nsresult rv = docLoaderFactory->CreateInstance(
       "external-resource", chan, newLoadGroup, type, nullptr, nullptr,
       getter_AddRefs(listener), getter_AddRefs(viewer));
   NS_ENSURE_SUCCESS(rv, rv);

dom/base/nsContentUtils.cpp

@@ -4631,28 +4631,66 @@ bool nsContentUtils::IsChildOfSameType(Document* aDoc) {
   return false;
 }
 
+static bool IsJSONType(const nsACString& aContentType) {
+  return aContentType.EqualsLiteral(TEXT_JSON) ||
+         aContentType.EqualsLiteral(APPLICATION_JSON);
+}
+
+static bool IsNonPlainTextType(const nsACString& aContentType) {
+  // MIME type suffixes which should not be plain text.
+  static constexpr std::string_view kNonPlainTextTypes[] = {
+      "html",
+      "xml",
+      "xsl",
+      "calendar",
+      "x-calendar",
+      "x-vcalendar",
+      "vcalendar",
+      "vcard",
+      "x-vcard",
+      "directory",
+      "ldif",
+      "qif",
+      "x-qif",
+      "x-csv",
+      "x-vcf",
+      "rtf",
+      "comma-separated-values",
+      "csv",
+      "tab-separated-values",
+      "tsv",
+      "ofx",
+      "vnd.sun.j2me.app-descriptor",
+      "x-ms-iqy",
+      "x-ms-odc",
+      "x-ms-rqy",
+      "x-ms-contact"};
+
+  // Trim off the "text/" prefix for comparison.
+  MOZ_ASSERT(StringBeginsWith(aContentType, "text/"_ns));
+  std::string_view suffix = aContentType;
+  suffix.remove_prefix(5);
+
+  for (std::string_view type : kNonPlainTextTypes) {
+    if (type == suffix) {
+      return true;
+    }
+  }
+  return false;
+}
+
 bool nsContentUtils::IsPlainTextType(const nsACString& aContentType) {
-  // NOTE: if you add a type here, add it to the content_types array in
+  // All `text/*`, any JSON type and any JavaScript type are considered "plain
-  // layout/build/components.conf as well.
+  // text" types for the purposes of how to render them as a document.
-  return aContentType.EqualsLiteral(TEXT_PLAIN) ||
+  return (StringBeginsWith(aContentType, "text/"_ns) &&
-         aContentType.EqualsLiteral(TEXT_CSS) ||
+          !IsNonPlainTextType(aContentType)) ||
-         aContentType.EqualsLiteral(TEXT_CACHE_MANIFEST) ||
+         IsJSONType(aContentType) || IsJavascriptMIMEType(aContentType);
-         aContentType.EqualsLiteral(TEXT_VTT) ||
-         aContentType.EqualsLiteral(APPLICATION_JAVASCRIPT) ||
-         aContentType.EqualsLiteral(APPLICATION_XJAVASCRIPT) ||
-         aContentType.EqualsLiteral(TEXT_ECMASCRIPT) ||
-         aContentType.EqualsLiteral(APPLICATION_ECMASCRIPT) ||
-         aContentType.EqualsLiteral(TEXT_JAVASCRIPT) ||
-         aContentType.EqualsLiteral(APPLICATION_JSON) ||
-         aContentType.EqualsLiteral(TEXT_JSON) ||
-         aContentType.EqualsLiteral(TEXT_EVENT_STREAM);
 }
 
 bool nsContentUtils::IsUtf8OnlyPlainTextType(const nsACString& aContentType) {
   // NOTE: This must be a subset of the list in IsPlainTextType().
-  return aContentType.EqualsLiteral(TEXT_CACHE_MANIFEST) ||
+  return IsJSONType(aContentType) ||
-         aContentType.EqualsLiteral(APPLICATION_JSON) ||
+         aContentType.EqualsLiteral(TEXT_CACHE_MANIFEST) ||
-         aContentType.EqualsLiteral(TEXT_JSON) ||
          aContentType.EqualsLiteral(TEXT_VTT);
 }
 
@@ -7218,9 +7256,12 @@ nsContentUtils::FindInternalDocumentViewer(const nsACString& aType,
     return docFactory.forget();
   }
 
-  if (DecoderTraits::IsSupportedInVideoDocument(aType)) {
+  // If the type wasn't registered in `Gecko-Content-Viewers`, check if it's
-    docFactory =
+  // another type which we may dynamically support, such as `text/*` types or
-        do_GetService("@mozilla.org/content/document-loader-factory;1");
+  // video document types. These types are all backed by the nsContentDLF.
+  if (IsPlainTextType(aType) ||
+      DecoderTraits::IsSupportedInVideoDocument(aType)) {
+    docFactory = do_GetService(CONTENT_DLF_CONTRACTID);
     if (docFactory && aLoaderType) {
       *aLoaderType = TYPE_CONTENT;
     }
@@ -7930,32 +7971,40 @@ void nsContentUtils::DestroyMatchString(void* aData) {
   }
 }
 
-bool nsContentUtils::IsJavascriptMIMEType(const nsAString& aMIMEType) {
+// Table ordered from most to least likely JS MIME types.
-  // Table ordered from most to least likely JS MIME types.
+static constexpr std::string_view kJavascriptMIMETypes[] = {
-  static const char* jsTypes[] = {"text/javascript",
+    "text/javascript",
-                                  "text/ecmascript",
+    "text/ecmascript",
-                                  "application/javascript",
+    "application/javascript",
-                                  "application/ecmascript",
+    "application/ecmascript",
-                                  "application/x-javascript",
+    "application/x-javascript",
-                                  "application/x-ecmascript",
+    "application/x-ecmascript",
-                                  "text/javascript1.0",
+    "text/javascript1.0",
-                                  "text/javascript1.1",
+    "text/javascript1.1",
-                                  "text/javascript1.2",
+    "text/javascript1.2",
-                                  "text/javascript1.3",
+    "text/javascript1.3",
-                                  "text/javascript1.4",
+    "text/javascript1.4",
-                                  "text/javascript1.5",
+    "text/javascript1.5",
-                                  "text/jscript",
+    "text/jscript",
-                                  "text/livescript",
+    "text/livescript",
-                                  "text/x-ecmascript",
+    "text/x-ecmascript",
-                                  "text/x-javascript",
+    "text/x-javascript"};
-                                  nullptr};
 
-  for (uint32_t i = 0; jsTypes[i]; ++i) {
+bool nsContentUtils::IsJavascriptMIMEType(const nsAString& aMIMEType) {
-    if (aMIMEType.LowerCaseEqualsASCII(jsTypes[i])) {
+  for (std::string_view type : kJavascriptMIMETypes) {
+    if (aMIMEType.LowerCaseEqualsASCII(type.data(), type.length())) {
       return true;
     }
   }
+  return false;
+}
 
+bool nsContentUtils::IsJavascriptMIMEType(const nsACString& aMIMEType) {
+  for (std::string_view type : kJavascriptMIMETypes) {
+    if (aMIMEType.LowerCaseEqualsASCII(type.data(), type.length())) {
+      return true;
+    }
+  }
   return false;
 }
 

dom/base/nsContentUtils.h

@@ -2769,6 +2769,7 @@ class nsContentUtils {
   static bool IsJavaScriptLanguage(const nsString& aName);
 
   static bool IsJavascriptMIMEType(const nsAString& aMIMEType);
+  static bool IsJavascriptMIMEType(const nsACString& aMIMEType);
 
   static void SplitMimeType(const nsAString& aValue, nsString& aType,
                             nsString& aParams);

gfx/thebes/gfxSVGGlyphs.cpp

@@ -131,20 +131,13 @@ gfxSVGGlyphsDocument* gfxSVGGlyphs::FindOrCreateGlyphsDocument(
 }
 
 nsresult gfxSVGGlyphsDocument::SetupPresentation() {
-  nsCOMPtr<nsICategoryManager> catMan =
-      do_GetService(NS_CATEGORYMANAGER_CONTRACTID);
-  nsCString contractId;
-  nsresult rv = catMan->GetCategoryEntry("Gecko-Content-Viewers",
-                                         "image/svg+xml", contractId);
-  NS_ENSURE_SUCCESS(rv, rv);
-
   nsCOMPtr<nsIDocumentLoaderFactory> docLoaderFactory =
-      do_GetService(contractId.get());
+      nsContentUtils::FindInternalDocumentViewer("image/svg+xml"_ns);
   NS_ASSERTION(docLoaderFactory, "Couldn't get DocumentLoaderFactory");
 
   nsCOMPtr<nsIDocumentViewer> viewer;
-  rv = docLoaderFactory->CreateInstanceForDocument(nullptr, mDocument, nullptr,
+  nsresult rv = docLoaderFactory->CreateInstanceForDocument(
-                                                   getter_AddRefs(viewer));
+      nullptr, mDocument, nullptr, getter_AddRefs(viewer));
   NS_ENSURE_SUCCESS(rv, rv);
 
   auto upem = mOwner->FontEntry()->UnitsPerEm();

image/SVGDocumentWrapper.cpp

@@ -283,20 +283,14 @@ nsresult SVGDocumentWrapper::SetupViewer(nsIRequest* aRequest,
   NS_ENSURE_TRUE(newLoadGroup, NS_ERROR_OUT_OF_MEMORY);
   newLoadGroup->SetLoadGroup(loadGroup);
 
-  nsCOMPtr<nsICategoryManager> catMan =
-      do_GetService(NS_CATEGORYMANAGER_CONTRACTID);
-  NS_ENSURE_TRUE(catMan, NS_ERROR_NOT_AVAILABLE);
-  nsCString contractId;
-  nsresult rv = catMan->GetCategoryEntry("Gecko-Content-Viewers", IMAGE_SVG_XML,
-                                         contractId);
-  NS_ENSURE_SUCCESS(rv, rv);
   nsCOMPtr<nsIDocumentLoaderFactory> docLoaderFactory =
-      do_GetService(contractId.get());
+      nsContentUtils::FindInternalDocumentViewer(
+          nsLiteralCString(IMAGE_SVG_XML));
   NS_ENSURE_TRUE(docLoaderFactory, NS_ERROR_NOT_AVAILABLE);
 
   nsCOMPtr<nsIDocumentViewer> viewer;
   nsCOMPtr<nsIStreamListener> listener;
-  rv = docLoaderFactory->CreateInstance(
+  nsresult rv = docLoaderFactory->CreateInstance(
       "external-resource", chan, newLoadGroup, nsLiteralCString(IMAGE_SVG_XML),
       nullptr, nullptr, getter_AddRefs(listener), getter_AddRefs(viewer));
   NS_ENSURE_SUCCESS(rv, rv);

layout/build/components.conf

@@ -14,24 +14,14 @@ UnloadFunc = 'nsLayoutModuleDtor'
 Priority = 100
 
 content_types = [
-    'application/ecmascript',
-    'application/javascript',
-    'application/json',
     'application/mathml+xml',
     'application/rdf+xml',
     'application/vnd.wap.xhtml+xml',
-    'application/x-javascript',
     'application/x-view-source',
     'application/xhtml+xml',
     'application/xml',
     'image/svg+xml',
-    'text/cache-manifest',
-    'text/css',
-    'text/ecmascript',
-    'text/event-stream',
     'text/html',
-    'text/javascript',
-    'text/json',
     'text/plain',
     'text/rdf',
     'text/vtt',

netwerk/test/browser/browser_display_plaintext_types.js

@@ -61,4 +61,7 @@ add_task(async function test_display_plaintext_type() {
   await expectOutcome("application/json", "jsonviewer");
   // NOTE: text/json does not load JSON viewer?
   await expectOutcome("text/json", "text");
+
+  // Unknown text/ types should be loadable as plain text documents.
+  await expectOutcome("text/unknown-type", "text");
 });

toolkit/components/utils/SimpleServices.sys.mjs

@@ -166,11 +166,12 @@ HttpIndexViewer.prototype = {
 
     aChannel.contentType = contentType;
 
-    let contract = Services.catMan.getCategoryEntry(
+    // NOTE: This assumes that both text/html and text/plain will continue to be
-      "Gecko-Content-Viewers",
+    // handled by nsContentDLF. If this ever changes this logic will need to be
-      contentType
+    // updated.
-    );
+    let factory = Cc[
-    let factory = Cc[contract].getService(Ci.nsIDocumentLoaderFactory);
+      "@mozilla.org/content/document-loader-factory;1"
+    ].getService(Ci.nsIDocumentLoaderFactory);
 
     let listener = {};
     let res = factory.createInstance(