Implement mimetype sniffing for extensions.

abarth: can you review the changes to mime_sniffer.cc?
paul: everything else?

BUG=13296
TEST=Added unit tests
Review URL: http://codereview.chromium.org/159345

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@21612 0039d316-1c4b-4281-b951-d872f2087c98

2 files changed, 107 insertions, 1 deletions

diff --git a/net/base/mime_sniffer.cc b/net/base/mime_sniffer.cc
index 0a8bc57..07feb33 100644
--- a/net/base/mime_sniffer.cc
+++ b/net/base/mime_sniffer.cc
@@ -256,7 +256,7 @@ static bool CheckForMagicNumbers(const char* content, size_t size,
                                  Histogram* counter, std::string* result) {
   for (size_t i = 0; i < magic_len; ++i) {
     if (MatchMagicNumber(content, size, &(magic[i]), result)) {
-      counter->Add(static_cast<int>(i));
+      if (counter) counter->Add(static_cast<int>(i));
       return true;
     }
   }
@@ -438,6 +438,43 @@ static bool IsUnknownMimeType(const std::string& mime_type) {
   return false;
 }
 
+// Sniff a crx (chrome extension) file.
+static bool SniffCRX(const char* content, size_t content_size, const GURL& url,
+                     const std::string& type_hint, std::string* result) {
+  static SnifferHistogram counter("mime_sniffer.kSniffCRX", 3);
+
+  // Technically, the crx magic number is just Cr24, but the bytes after that
+  // are a version number which changes infrequently. Including it in the
+  // sniffing gives us less room for error. If the version number ever changes,
+  // we can just add an entry to this list.
+  //
+  // TODO(aa): If we ever have another magic number, we'll want to pass a
+  // histogram into CheckForMagicNumbers(), below, to see which one matched.
+  const struct MagicNumber kCRXMagicNumbers[] = {
+    MAGIC_NUMBER("application/x-chrome-extension", "Cr24\x02\x00\x00\x00")
+  };
+
+  // Only consider files that have the extension ".crx".
+  const char kCRXExtension[] = ".crx";
+  const int kExtensionLength = arraysize(kCRXExtension) - 1;  // ignore null
+  if (url.path().rfind(kCRXExtension, std::string::npos, kExtensionLength) ==
+      url.path().size() - kExtensionLength) {
+    counter.Add(1);
+  } else {
+    return false;
+  }
+
+  if (CheckForMagicNumbers(content, content_size,
+                           kCRXMagicNumbers, arraysize(kCRXMagicNumbers),
+                           NULL, result)) {
+    counter.Add(2);
+  } else {
+    return false;
+  }
+
+  return true;
+}
+
 bool ShouldSniffMimeType(const GURL& url, const std::string& mime_type) {
   static SnifferHistogram should_sniff_counter(
       "mime_sniffer.ShouldSniffMimeType2", 3);
@@ -535,6 +572,11 @@ bool SniffMimeType(const char* content, size_t content_size,
     return content_size >= kMaxBytesToSniff;
   }
 
+  // CRX files (chrome extensions) have a special sniffing algorithm. It is
+  // tighter than the others because we don't have to match legacy behavior.
+  if (SniffCRX(content, content_size, url, type_hint, result))
+    return true;
+
   // Now we look in our large table of magic numbers to see if we can find
   // anything that matches the content.
   if (SniffForMagicNumbers(content, content_size, result))
diff --git a/net/base/mime_sniffer_unittest.cc b/net/base/mime_sniffer_unittest.cc
index 96eb441..ed1634c 100644
--- a/net/base/mime_sniffer_unittest.cc
+++ b/net/base/mime_sniffer_unittest.cc
@@ -90,6 +90,70 @@ TEST(MimeSnifferTest, BasicSniffingTest) {
   TestArray(tests, arraysize(tests));
 }
 
+TEST(MimeSnifferTest, ChromeExtensionsTest) {
+  SnifferTest tests[] = {
+    // schemes
+    { "Cr24\x02\x00\x00\x00", sizeof("Cr24\x02\x00\x00\x00")-1,
+      "http://www.example.com/foo.crx",
+      "", "application/x-chrome-extension" },
+    { "Cr24\x02\x00\x00\x00", sizeof("Cr24\x02\x00\x00\x00")-1,
+      "https://www.example.com/foo.crx",
+      "", "application/x-chrome-extension" },
+    { "Cr24\x02\x00\x00\x00", sizeof("Cr24\x02\x00\x00\x00")-1,
+      "ftp://www.example.com/foo.crx",
+      "", "application/x-chrome-extension" },
+
+    // some other mimetypes that should get converted
+    { "Cr24\x02\x00\x00\x00", sizeof("Cr24\x02\x00\x00\x00")-1,
+      "http://www.example.com/foo.crx",
+      "text/plain", "application/x-chrome-extension" },
+    { "Cr24\x02\x00\x00\x00", sizeof("Cr24\x02\x00\x00\x00")-1,
+      "http://www.example.com/foo.crx",
+      "application/octet-stream", "application/x-chrome-extension" },
+
+    // success edge cases
+    { "Cr24\x02\x00\x00\x00", sizeof("Cr24\x02\x00\x00\x00")-1,
+      "http://www.example.com/foo.crx?query=string",
+      "", "application/x-chrome-extension" },
+    { "Cr24\x02\x00\x00\x00", sizeof("Cr24\x02\x00\x00\x00")-1,
+      "http://www.example.com/foo..crx",
+      "", "application/x-chrome-extension" },
+
+    // wrong file extension
+    { "Cr24\x02\x00\x00\x00", sizeof("Cr24\x02\x00\x00\x00")-1,
+      "http://www.example.com/foo.bin",
+      "", "application/octet-stream" },
+    { "Cr24\x02\x00\x00\x00", sizeof("Cr24\x02\x00\x00\x00")-1,
+      "http://www.example.com/foo.bin?monkey",
+      "", "application/octet-stream" },
+    { "Cr24\x02\x00\x00\x00", sizeof("Cr24\x02\x00\x00\x00")-1,
+      "invalid-url",
+      "", "application/octet-stream" },
+    { "Cr24\x02\x00\x00\x00", sizeof("Cr24\x02\x00\x00\x00")-1,
+      "http://www.example.com",
+      "", "application/octet-stream" },
+    { "Cr24\x02\x00\x00\x00", sizeof("Cr24\x02\x00\x00\x00")-1,
+      "http://www.example.com/",
+      "", "application/octet-stream" },
+    { "Cr24\x02\x00\x00\x00", sizeof("Cr24\x02\x00\x00\x00")-1,
+      "http://www.example.com/foo",
+      "", "application/octet-stream" },
+    { "Cr24\x02\x00\x00\x00", sizeof("Cr24\x02\x00\x00\x00")-1,
+      "http://www.example.com/foocrx",
+      "", "application/octet-stream" },
+    { "Cr24\x02\x00\x00\x00", sizeof("Cr24\x02\x00\x00\x00")-1,
+      "http://www.example.com/foo.crx.blech",
+      "", "application/octet-stream" },
+
+    // wrong magic
+    { "Cr24\x02\x00\x00\x01", sizeof("Cr24\x02\x00\x00\x01")-1,
+      "http://www.example.com/foo.crx?monkey",
+      "", "application/octet-stream" },
+  };
+
+  TestArray(tests, arraysize(tests));
+}
+
 TEST(MimeSnifferTest, MozillaCompatibleTest) {
   SnifferTest tests[] = {
     { " \n <hTmL>\n <hea", sizeof(" \n <hTmL>\n <hea")-1,