Fix the local directory listing, FTP directory listing and the local file handling (drag'n'drop and opening from the file list).

For the local file listing, use the OS file system encoding. 

For the FTP directory listing, use ICU's encoding detector.GetDirectoryListingEntry and GetDirectoryLisingHeader were changed to accept string16 for file/directory names. To the former, a new parameter (|raw_bytes|) was added. It can be used to make a FTP request to a file with a non-ASCII name encoded in a legacy encoding.

For the local file handling on Windows, get rid of the code for 'doubly converted' UTF-8 in FileURLToFilePath, which led to issue 4619 and add a few cases to NetUtil*.FileURLConversion* test.

In addition, add
CodepageToUTF16 and UTF16ToCodepage along with a new unittest (ConvertBetweenCodepageAndUTF16) that shares the same set of case as ConvertBetweenCodepageAndWide. The test cases were expanded and revised a bit. 


BUG=2939,13229,4619     
    http://crbug.com/2939  http://crbug.com/13229 http://crbug.com/4619

TEST=1. Pass URLRequest*.FTP* (net_unittests)
     2. Pass StringUtiltTest.ConvertBetweenCode*
     3. Pass NetUtil*.GetDirectoryLis* (net_unittests)     
     4. Open a local directory containing files with non-ASCII names and they're displayed correctly in the directory list. On Windows and Mac OS X, it should always work. On Linux, your locale encoding (as returned by nl_langinfo(CODESET)) should match the actual encoding used in your filename. 
     5a. Pass NetUtil*.FileURL* (net_unittests) with the default codepage set to 1252 and 932. 
     5b. Make a file named 'caf챕.txt' on Windows and see if it can be opened both by clicking in the directory listing page of Chrome and by drag'n'drop. Test this with the default OS code pages set to Windows-1252, Windows-1251 (Russian) and Windows-932 (Japanese). 
Review URL: http://codereview.chromium.org/151065

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@20331 0039d316-1c4b-4281-b951-d872f2087c98

4 files changed, 83 insertions, 56 deletions

diff --git a/net/base/net_util.cc b/net/base/net_util.cc
index 2e6292c..00beb4e 100644
--- a/net/base/net_util.cc
+++ b/net/base/net_util.cc
@@ -860,7 +860,7 @@ std::string CanonicalizeHost(const std::wstring& host,
   return CanonicalizeHost(converted_host, host_info);
 }
 
-std::string GetDirectoryListingHeader(const std::string& title) {
+std::string GetDirectoryListingHeader(const string16& title) {
   static const StringPiece header(NetModule::GetResource(IDR_DIR_HEADER_HTML));
   if (header.empty()) {
     NOTREACHED() << "expected resource not found";
@@ -874,15 +874,21 @@ std::string GetDirectoryListingHeader(const std::string& title) {
   return result;
 }
 
-std::string GetDirectoryListingEntry(const std::string& name,
+std::string GetDirectoryListingEntry(const string16& name,
+                                     const std::string& raw_bytes,
                                      bool is_dir,
                                      int64 size,
-                                     const Time& modified) {
+                                     Time modified) {
   std::string result;
   result.append("<script>addRow(");
   string_escape::JsonDoubleQuote(name, true, &result);
   result.append(",");
-  string_escape::JsonDoubleQuote(EscapePath(name), true, &result);
+  if (raw_bytes.empty()) {
+    string_escape::JsonDoubleQuote(EscapePath(UTF16ToUTF8(name)),
+                                   true, &result);
+  } else {
+    string_escape::JsonDoubleQuote(EscapePath(raw_bytes), true, &result);
+  }
   if (is_dir) {
     result.append(",1,");
   } else {
diff --git a/net/base/net_util.h b/net/base/net_util.h
index 40df770..4320e1c 100644
--- a/net/base/net_util.h
+++ b/net/base/net_util.h
@@ -14,6 +14,7 @@
 #include <string>
 
 #include "base/basictypes.h"
+#include "base/string16.h"
 #include "net/base/escape.h"
 
 struct addrinfo;
@@ -147,12 +148,24 @@ std::string CanonicalizeHost(const std::string& host,
 std::string CanonicalizeHost(const std::wstring& host,
                              url_canon::CanonHostInfo* host_info);
 
-// Call these functions to get the html for a directory listing.
-// They will pass non-7bit-ascii characters unescaped, allowing
-// the browser to interpret the encoding (utf8, etc).
-std::string GetDirectoryListingHeader(const std::string& title);
-std::string GetDirectoryListingEntry(const std::string& name, bool is_dir,
-                                     int64 size, const base::Time& modified);
+// Call these functions to get the html snippet for a directory listing.
+// The return values of both functions are in UTF-8.
+std::string GetDirectoryListingHeader(const string16& title);
+
+// Given the name of a file in a directory (ftp or local) and
+// other information (is_dir, size, modification time), it returns
+// the html snippet to add the entry for the file to the directory listing.
+// Currently, it's a script tag containing a call to a Javascript function
+// |addRow|.
+//
+// Its 1st parameter is derived from |name| and is the Javascript-string
+// escaped form of |name| (i.e \uXXXX). The 2nd parameter is the url-escaped
+// |raw_bytes| if it's not empty. If empty, the 2nd parameter is the
+// url-escaped |name| in UTF-8.
+std::string GetDirectoryListingEntry(const string16& name,
+                                     const std::string& raw_bytes,
+                                     bool is_dir, int64 size,
+                                     base::Time modified);
 
 // If text starts with "www." it is removed, otherwise text is returned
 // unmodified.
diff --git a/net/base/net_util_unittest.cc b/net/base/net_util_unittest.cc
index 78f7ab9..f346e92 100644
--- a/net/base/net_util_unittest.cc
+++ b/net/base/net_util_unittest.cc
@@ -407,18 +407,32 @@ TEST(NetUtilTest, FileURLConversion) {
      "file://some%20computer/foo/bar.txt"}, // UNC
     {L"D:\\Name;with%some symbols*#",
      "file:///D:/Name%3Bwith%25some%20symbols*%23"},
+    // issue 14153: To be tested with the OS default codepage other than 1252.
+    {L"D:\\latin1\\caf\x00E9\x00DD.txt",
+     "file:///D:/latin1/caf%C3%A9%C3%9D.txt"},
+    {L"D:\\otherlatin\\caf\x0119.txt",
+     "file:///D:/otherlatin/caf%C4%99.txt"},
+    {L"D:\\greek\\\x03B1\x03B2\x03B3.txt",
+     "file:///D:/greek/%CE%B1%CE%B2%CE%B3.txt"},
     {L"D:\\Chinese\\\x6240\x6709\x4e2d\x6587\x7f51\x9875.doc",
      "file:///D:/Chinese/%E6%89%80%E6%9C%89%E4%B8%AD%E6%96%87%E7%BD%91"
          "%E9%A1%B5.doc"},
+    {L"D:\\plane1\\\xD835\xDC00\xD835\xDC01.txt",  // Math alphabet "AB"
+     "file:///D:/plane1/%F0%9D%90%80%F0%9D%90%81.txt"},
 #elif defined(OS_POSIX)
     {L"/foo/bar.txt", "file:///foo/bar.txt"},
     {L"/foo/BAR.txt", "file:///foo/BAR.txt"},
     {L"/C:/foo/bar.txt", "file:///C:/foo/bar.txt"},
     {L"/some computer/foo/bar.txt", "file:///some%20computer/foo/bar.txt"},
     {L"/Name;with%some symbols*#", "file:///Name%3Bwith%25some%20symbols*%23"},
+    {L"/latin1/caf\x00E9\x00DD.txt", "file:///latin1/caf%C3%A9%C3%9D.txt"},
+    {L"/otherlatin/caf\x0119.txt", "file:///otherlatin/caf%C4%99.txt"},
+    {L"/greek/\x03B1\x03B2\x03B3.txt", "file:///greek/%CE%B1%CE%B2%CE%B3.txt"},
     {L"/Chinese/\x6240\x6709\x4e2d\x6587\x7f51\x9875.doc",
      "file:///Chinese/%E6%89%80%E6%9C%89%E4%B8%AD%E6%96%87%E7%BD"
          "%91%E9%A1%B5.doc"},
+    {L"/plane1/\x1D400\x1D401.txt",  // Math alphabet "AB"
+     "file:///plane1/%F0%9D%90%80%F0%9D%90%81.txt"},
 #endif
   };
 
@@ -474,21 +488,6 @@ TEST(NetUtilTest, FileURLConversion) {
     EXPECT_EQ(url_cases[i].file, output.ToWStringHack());
   }
 
-  // Here, we test that UTF-8 encoded strings get decoded properly, even when
-  // they might be stored with wide characters.  On posix systems, just treat
-  // this as a stream of bytes.
-  const wchar_t utf8[] = L"file:///d:/Chinese/\xe6\x89\x80\xe6\x9c\x89\xe4\xb8"
-                         L"\xad\xe6\x96\x87\xe7\xbd\x91\xe9\xa1\xb5.doc";
-#if defined(OS_WIN)
-  const wchar_t wide[] =
-      L"D:\\Chinese\\\x6240\x6709\x4e2d\x6587\x7f51\x9875.doc";
-#elif defined(OS_POSIX)
-  const wchar_t wide[] = L"/d:/Chinese/\xe6\x89\x80\xe6\x9c\x89\xe4\xb8\xad\xe6"
-                         L"\x96\x87\xe7\xbd\x91\xe9\xa1\xb5.doc";
-#endif
-  EXPECT_TRUE(net::FileURLToFilePath(GURL(WideToUTF8(utf8)), &output));
-  EXPECT_EQ(wide, output.ToWStringHack());
-
   // Unfortunately, UTF8ToWide discards invalid UTF8 input.
 #ifdef BUG_878908_IS_FIXED
   // Test that no conversion happens if the UTF-8 input is invalid, and that
@@ -862,7 +861,8 @@ TEST(NetUtilTest, GetSuggestedFilename) {
 namespace {
 
 struct GetDirectoryListingEntryCase {
-  const char* name;
+  const wchar_t* name;
+  const char* raw_bytes;
   bool is_dir;
   int64 filesize;
   base::Time time;
@@ -872,22 +872,50 @@ struct GetDirectoryListingEntryCase {
 }  // namespace
 TEST(NetUtilTest, GetDirectoryListingEntry) {
   const GetDirectoryListingEntryCase test_cases[] = {
-    {"Foo",
+    {L"Foo",
+     "",
      false,
      10000,
      base::Time(),
      "<script>addRow(\"Foo\",\"Foo\",0,\"9.8 kB\",\"\");</script>\n"},
-    {"quo\"tes",
+    {L"quo\"tes",
+     "",
+     false,
+     10000,
+     base::Time(),
+     "<script>addRow(\"quo\\\"tes\",\"quo%22tes\",0,\"9.8 kB\",\"\");</script>"
+         "\n"},
+    {L"quo\"tes",
+     "quo\"tes",
      false,
      10000,
      base::Time(),
      "<script>addRow(\"quo\\\"tes\",\"quo%22tes\",0,\"9.8 kB\",\"\");</script>"
          "\n"},
+    // U+D55C0 U+AE00. raw_bytes is empty (either a local file with
+    // UTF-8/UTF-16 encoding or a remote file on an ftp server using UTF-8
+    {L"\xD55C\xAE00.txt",
+     "",
+     false,
+     10000,
+     base::Time(),
+     "<script>addRow(\"\\uD55C\\uAE00.txt\",\"%ED%95%9C%EA%B8%80.txt\""
+         ",0,\"9.8 kB\",\"\");</script>\n"},
+    // U+D55C0 U+AE00. raw_bytes is the corresponding EUC-KR sequence:
+    // a local or remote file in EUC-KR.
+    {L"\xD55C\xAE00.txt",
+     "\xC7\xD1\xB1\xDB.txt",
+     false,
+     10000,
+     base::Time(),
+     "<script>addRow(\"\\uD55C\\uAE00.txt\",\"%C7%D1%B1%DB.txt\""
+         ",0,\"9.8 kB\",\"\");</script>\n"},
   };
 
   for (size_t i = 0; i < ARRAYSIZE_UNSAFE(test_cases); ++i) {
     const std::string results = net::GetDirectoryListingEntry(
-        test_cases[i].name,
+        WideToUTF16(test_cases[i].name),
+        test_cases[i].raw_bytes,
         test_cases[i].is_dir,
         test_cases[i].filesize,
         test_cases[i].time);
diff --git a/net/base/net_util_win.cc b/net/base/net_util_win.cc
index effb212..244f4ad 100644
--- a/net/base/net_util_win.cc
+++ b/net/base/net_util_win.cc
@@ -57,33 +57,13 @@ bool FileURLToFilePath(const GURL& url, FilePath* file_path) {
   }
   file_path_str.assign(UTF8ToWide(path));
 
-  // Now we have an unescaped filename, but are still not sure about its
-  // encoding. For example, each character could be part of a UTF-8 string.
-  if (file_path_str.empty() || !IsString8Bit(file_path_str)) {
-    // assume our 16-bit encoding is correct if it won't fit into an 8-bit
-    // string
-    return true;
-  }
-
-  // Convert our narrow string into the native wide path.
-  std::string narrow;
-  if (!WideToLatin1(file_path_str, &narrow)) {
-    NOTREACHED() << "Should have filtered out non-8-bit strings above.";
-    return false;
-  }
-  if (IsStringUTF8(narrow)) {
-    // Our string actually looks like it could be UTF-8, convert to 8-bit
-    // UTF-8 and then to the corresponding wide string.
-    file_path_str = UTF8ToWide(narrow);
-  } else {
-    // Our wide string contains only 8-bit characters and it's not UTF-8, so
-    // we assume it's in the native codepage.
-    file_path_str = base::SysNativeMBToWide(narrow);
-  }
-
-  // Fail if 8-bit -> wide conversion failed and gave us an empty string back
-  // (we already filtered out empty strings above).
-  return !file_path_str.empty();
+  // We used to try too hard and see if |path| made up entirely of
+  // the 1st 256 characters in the Unicode was a zero-extended UTF-16.
+  // If so, we converted it to 'Latin-1' and checked if the result was UTF-8.
+  // If the check passed, we converted the result to UTF-8.
+  // Otherwise, we treated the result as the native OS encoding.
+  // However, that led to http://crbug.com/4619 and http://crbug.com/14153
+  return true;
 }
 
 }  // namespace net