Move filename determination to net_util

net/base/net_util.cc

 // Copyright (c) 2011 The Chromium Authors. All rights reserved.
 // Use of this source code is governed by a BSD-style license that can be
 // found in the LICENSE file.
 
 #include "net/base/net_util.h"
 
 #include <unicode/regex.h>
 #include <unicode/ucnv.h>
 #include <unicode/uidna.h>
 #include <unicode/ulocdata.h>
@@ skipping 46 matching lines @@
 #include "base/time.h"
 #include "base/utf_offset_string_conversions.h"
 #include "base/utf_string_conversions.h"
 #include "googleurl/src/gurl.h"
 #include "googleurl/src/url_canon.h"
 #include "googleurl/src/url_canon_ip.h"
 #include "googleurl/src/url_parse.h"
 #include "grit/net_resources.h"
 #include "net/base/dns_util.h"
 #include "net/base/escape.h"
+#include "net/base/mime_util.h"
 #include "net/base/net_module.h"
 #if defined(OS_WIN)
 #include "net/base/winsock_init.h"
 #endif
 #include "unicode/datefmt.h"
 
 using base::Time;
 
 namespace net {
 
@@ skipping 837 matching lines @@
 }
 
 char* do_strdup(const char* src) {
 #if defined(OS_WIN)
   return _strdup(src);
 #else
   return strdup(src);
 #endif
 }
 
+#if defined(OS_WIN)
+// Returns whether the specified extension is automatically integrated into the
+// windows shell.
+bool IsShellIntegratedExtension(const string16& extension) {
+  string16 extension_lower = StringToLowerASCII(extension);
+
+  static const wchar_t* const integrated_extensions[] = {
+    // See <http://msdn.microsoft.com/en-us/library/ms811694.aspx>.
+    L"local",
+    // Right-clicking on shortcuts can be magical.
+    L"lnk",
+  };
+
+  for (int i = 0; i < arraysize(integrated_extensions); ++i) {
+    if (extension_lower == integrated_extensions[i])
+      return true;
+  }
+
+  // See <http://www.juniper.net/security/auto/vulnerabilities/vuln2612.html>.
+  // That vulnerability report is not exactly on point, but files become magical
+  // if their end in a CLSID.  Here we block extensions that look like CLSIDs.
+  if (!extension_lower.empty() && extension_lower[0] == L'{' &&
+      extension_lower[extension_lower.length() - 1] == L'}')
+    return true;
+
+  return false;
+}
+
+// Returns whether the specified file name is a reserved name on windows.
+// This includes names like "com2.zip" (which correspond to devices) and
+// desktop.ini and thumbs.db which have special meaning to the windows shell.
+bool IsReservedName(const string16& filename) {
+  // This list is taken from the MSDN article "Naming a file"
+  // http://msdn2.microsoft.com/en-us/library/aa365247(VS.85).aspx
+  // I also added clock$ because GetSaveFileName seems to consider it as a
+  // reserved name too.
+  static const wchar_t* const known_devices[] = {
+    L"con", L"prn", L"aux", L"nul", L"com1", L"com2", L"com3", L"com4", L"com5",
+    L"com6", L"com7", L"com8", L"com9", L"lpt1", L"lpt2", L"lpt3", L"lpt4",
+    L"lpt5", L"lpt6", L"lpt7", L"lpt8", L"lpt9", L"clock$"
+  };
+  string16 filename_lower = StringToLowerASCII(filename);
+
+  for (int i = 0; i < arraysize(known_devices); ++i) {
+    // Exact match.
+    if (filename_lower == known_devices[i])
+      return true;
+    // Starts with "DEVICE.".
+    if (filename_lower.find(string16(known_devices[i]) + L".") == 0)
+      return true;
+  }
+
+  static const wchar_t* const magic_names[] = {
+    // These file names are used by the "Customize folder" feature of the shell.
+    L"desktop.ini",
+    L"thumbs.db",
+  };
+
+  for (int i = 0; i < arraysize(magic_names); ++i) {
+    if (filename_lower == magic_names[i])
+      return true;
+  }
+
+  return false;
+}
+#endif  // OS_WIN
+
+void GenerateSafeExtension(const std::string& mime_type, FilePath* file_name) {
+  // We're worried about two things here:
+  //
+  // 1) Usability.  If the site fails to provide a file extension, we want to
+  //    guess a reasonable file extension based on the content type.
+  //
+  // 2) Shell integration.  Some file extensions automatically integrate with
+  //    the shell.  We block these extensions to prevent a malicious web site
+  //    from integrating with the user's shell.
+
+  // See if our file name already contains an extension.
+  FilePath::StringType extension = file_name->Extension();
+  if (!extension.empty())
+    extension.erase(extension.begin());  // Erase preceding '.'.
+
+#if defined(OS_WIN)
+  static const FilePath::CharType default_extension[] =
+      FILE_PATH_LITERAL("download");
+
+  // Rename shell-integrated extensions.
+  // TODO(asanka): Consider stripping out the bad extension and replacing it
+  // with the preferred extension for the MIME type if one is available.
+  if (IsShellIntegratedExtension(extension))
+    extension.assign(default_extension);
+#endif
+
+  if (extension.empty() && !mime_type.empty()) {
+    // The GetPreferredExtensionForMimeType call will end up going to disk.  Do
+    // this on another thread to avoid slowing the IO thread.
+    // http://crbug.com/61827
+    // TODO(asanka): Remove this ScopedAllowIO once all callers have switched
+    // over to IO safe threads.
+    base::ThreadRestrictions::ScopedAllowIO allow_io;
+    net::GetPreferredExtensionForMimeType(mime_type, &extension);
+  }
+
+  *file_name = file_name->ReplaceExtension(extension);
+}
+
 }  // namespace
 
 const FormatUrlType kFormatUrlOmitNothing                     = 0;
 const FormatUrlType kFormatUrlOmitUsernamePassword            = 1 << 0;
 const FormatUrlType kFormatUrlOmitHTTP                        = 1 << 1;
 const FormatUrlType kFormatUrlOmitTrailingSlashOnBareHostname = 1 << 2;
 const FormatUrlType kFormatUrlOmitAll = kFormatUrlOmitUsernamePassword |
     kFormatUrlOmitHTTP | kFormatUrlOmitTrailingSlashOnBareHostname;
 
 // TODO(viettrungluu): We don't want non-POD globals; change this.
@@ skipping 27 matching lines @@
 
 #if defined(OS_POSIX)
   ReplaceSubstringsAfterOffset(&url_string, 0,
       FILE_PATH_LITERAL("\\"), FILE_PATH_LITERAL("%5C"));
 #endif
 
   return GURL(url_string);
 }
 
 std::string GetSpecificHeader(const std::string& headers,
-                               const std::string& name) {
+                              const std::string& name) {
   // We want to grab the Value from the "Key: Value" pairs in the headers,
   // which should look like this (no leading spaces, \n-separated) (we format
   // them this way in url_request_inet.cc):
   //    HTTP/1.1 200 OK\n
   //    ETag: "6d0b8-947-24f35ec0"\n
   //    Content-Length: 2375\n
   //    Content-Type: text/html; charset=UTF-8\n
   //    Last-Modified: Sun, 03 Sep 2006 04:34:43 GMT\n
   if (headers.empty())
     return std::string();
@@ skipping 259 matching lines @@
   result.append(");</script>\n");
 
   return result;
 }
 
 string16 StripWWW(const string16& text) {
   const string16 www(ASCIIToUTF16("www."));
   return StartsWith(text, www, true) ? text.substr(www.length()) : text;
 }
 
+void GenerateSafeFileName(const std::string& mime_type, FilePath* file_path) {
+  // Make sure we get the right file extension
+  GenerateSafeExtension(mime_type, file_path);
+
+#if defined(OS_WIN)
+  // Prepend "_" to the file name if it's a reserved name
+  FilePath::StringType leaf_name = file_path->BaseName().value();
+  DCHECK(!leaf_name.empty());
+  if (IsReservedName(leaf_name)) {
+    leaf_name = FilePath::StringType(FILE_PATH_LITERAL("_")) + leaf_name;
+    *file_path = file_path->DirName();
+    if (file_path->value() == FilePath::kCurrentDirectory) {
+      *file_path = FilePath(leaf_name);
+    } else {
+      *file_path = file_path->Append(leaf_name);
+    }
+  }
+#endif
+}
+
 string16 GetSuggestedFilename(const GURL& url,
                               const std::string& content_disposition,
                               const std::string& referrer_charset,
                               const std::string& suggested_name,
                               const string16& default_name) {
   // TODO: this function to be updated to match the httpbis recommendations.
   // Talk to abarth for the latest news.
 
   // We don't translate this fallback string, "download". If localization is
   // needed, the caller should provide localized fallback default_name.
@@ skipping 80 matching lines @@
   string16 path = UTF8ToUTF16(filename);
   file_util::ReplaceIllegalCharactersInPath(&path, '-');
   return path;
 #else
   std::string path = filename;
   file_util::ReplaceIllegalCharactersInPath(&path, '-');
   return UTF8ToUTF16(path);
 #endif
 }
 
+FilePath GenerateFileName(const GURL& url,
+                          const std::string& content_disposition,
+                          const std::string& referrer_charset,
+                          const std::string& suggested_name,
+                          const std::string& mime_type,
+                          const string16& default_file_name) {
+  string16 new_name = GetSuggestedFilename(GURL(url),
+                                           content_disposition,
+                                           referrer_charset,
+                                           suggested_name,
+                                           default_file_name);
+
+  // TODO(evan): this code is totally wrong -- we should just generate
+  // Unicode filenames and do all this encoding switching at the end.
+  // However, I'm just shuffling wrong code around, at least not adding
+  // to it.
+#if defined(OS_WIN)
+  FilePath generated_name = FilePath(new_name);
+#else
+  FilePath generated_name = FilePath(
+      base::SysWideToNativeMB(UTF16ToWide(new_name)));
+#endif
+
+  DCHECK(!generated_name.empty());
+
+  GenerateSafeFileName(mime_type, &generated_name);
+  return generated_name;
+}
+
 bool IsPortAllowedByDefault(int port) {
   int array_size = arraysize(kRestrictedPorts);
   for (int i = 0; i < array_size; i++) {
     if (kRestrictedPorts[i] == port) {
       return false;
     }
   }
   return true;
 }
 
@@ skipping 886 matching lines @@
 
 NetworkInterface::NetworkInterface(const std::string& name,
                                    const IPAddressNumber& address)
     : name(name), address(address) {
 }
 
 NetworkInterface::~NetworkInterface() {
 }
 
 }  // namespace net