Localize strings, speeds.

BUG=86527
TEST=run in non-English. For European languages, during a download the decimal separators should be commas (e.g. "0,0 MB"). (The speeds are in strings files and might take a little time to run through the translation machinery.)

Review URL: http://codereview.chromium.org/7189076

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@90092 0039d316-1c4b-4281-b951-d872f2087c98

7 files changed, 189 insertions, 151 deletions

diff --git a/base/base.gyp b/base/base.gyp
index 668171d..3b7b085 100644
--- a/base/base.gyp
+++ b/base/base.gyp
@@ -130,6 +130,7 @@
         'i18n/case_conversion_unittest.cc',
         'i18n/file_util_icu_unittest.cc',
         'i18n/icu_string_conversions_unittest.cc',
+        'i18n/number_formatting_unittest.cc',
         'i18n/rtl_unittest.cc',
         'i18n/time_formatting_unittest.cc',
         'json/json_reader_unittest.cc',
diff --git a/base/i18n/number_formatting.cc b/base/i18n/number_formatting.cc
index a529a84..b2eeb16 100644
--- a/base/i18n/number_formatting.cc
+++ b/base/i18n/number_formatting.cc
@@ -20,6 +20,10 @@ namespace {
 
 struct NumberFormatWrapper {
   NumberFormatWrapper() {
+    Reset();
+  }
+
+  void Reset() {
     // There's no ICU call to destroy a NumberFormat object other than
     // operator delete, so use the default Delete, which calls operator delete.
     // This can cause problems if a different allocator is used by this file
@@ -32,12 +36,14 @@ struct NumberFormatWrapper {
   scoped_ptr<icu::NumberFormat> number_format;
 };
 
-}  // namespace
+LazyInstance<NumberFormatWrapper> g_number_format_int(LINKER_INITIALIZED);
+LazyInstance<NumberFormatWrapper> g_number_format_float(LINKER_INITIALIZED);
 
-static LazyInstance<NumberFormatWrapper> g_number_format(LINKER_INITIALIZED);
+}  // namespace
 
 string16 FormatNumber(int64 number) {
-  icu::NumberFormat* number_format = g_number_format.Get().number_format.get();
+  icu::NumberFormat* number_format =
+      g_number_format_int.Get().number_format.get();
 
   if (!number_format) {
     // As a fallback, just return the raw number in a string.
@@ -49,4 +55,29 @@ string16 FormatNumber(int64 number) {
   return string16(ustr.getBuffer(), static_cast<size_t>(ustr.length()));
 }
 
+string16 FormatDouble(double number, int fractional_digits) {
+  icu::NumberFormat* number_format =
+      g_number_format_float.Get().number_format.get();
+
+  if (!number_format) {
+    // As a fallback, just return the raw number in a string.
+    return UTF8ToUTF16(StringPrintf("%f", number));
+  }
+  number_format->setMaximumFractionDigits(fractional_digits);
+  number_format->setMinimumFractionDigits(fractional_digits);
+  icu::UnicodeString ustr;
+  number_format->format(number, ustr);
+
+  return string16(ustr.getBuffer(), static_cast<size_t>(ustr.length()));
+}
+
+namespace testing {
+
+void ResetFormatters() {
+  g_number_format_int.Get().Reset();
+  g_number_format_float.Get().Reset();
+}
+
+}  // namespace testing
+
 }  // namespace base
diff --git a/base/i18n/number_formatting.h b/base/i18n/number_formatting.h
index f5ec083..5df7f17 100644
--- a/base/i18n/number_formatting.h
+++ b/base/i18n/number_formatting.h
@@ -1,4 +1,4 @@
-// Copyright (c) 2010 The Chromium Authors. All rights reserved.
+// 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.
 
@@ -11,8 +11,24 @@
 
 namespace base {
 
+// Return a number formatted with separators in the user's locale.
+// Ex: FormatNumber(1234567)
+//         => "1,234,567" in English, "1.234.567" in German
 string16 FormatNumber(int64 number);
 
+// Return a number formatted with separators in the user's locale.
+// Ex: FormatDouble(1234567.8, 1)
+//         => "1,234,567.8" in English, "1.234.567,8" in German
+string16 FormatDouble(double number, int fractional_digits);
+
+namespace testing {
+
+// Causes cached formatters to be discarded and recreated. Only useful for
+// testing.
+void ResetFormatters();
+
+}  // namespace testing
+
 }  // namespace base
 
 #endif  // BASE_I18N_NUMBER_FORMATTING_H_
diff --git a/base/i18n/number_formatting_unittest.cc b/base/i18n/number_formatting_unittest.cc
new file mode 100644
index 0000000..e6a54b5
--- /dev/null
+++ b/base/i18n/number_formatting_unittest.cc
@@ -0,0 +1,92 @@
+// 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 <limits>
+
+#include "base/i18n/number_formatting.h"
+#include "base/utf_string_conversions.h"
+#include "testing/gtest/include/gtest/gtest.h"
+#include "unicode/locid.h"
+
+namespace {
+
+void SetICUDefaultLocale(const std::string& locale_string) {
+  icu::Locale locale(locale_string.c_str());
+  UErrorCode error_code = U_ZERO_ERROR;
+  icu::Locale::setDefault(locale, error_code);
+  EXPECT_TRUE(U_SUCCESS(error_code));
+}
+
+}  // namespace
+
+TEST(NumberFormattingTest, FormatNumber) {
+  static const struct {
+    int64 number;
+    const char* expected_english;
+    const char* expected_german;
+  } cases[] = {
+    {0, "0", "0"},
+    {1024, "1,024", "1.024"},
+    {std::numeric_limits<int64>::max(),
+        "9,223,372,036,854,775,807", "9.223.372.036.854.775.807"},
+    {std::numeric_limits<int64>::min(),
+        "-9,223,372,036,854,775,808", "-9.223.372.036.854.775.808"},
+    {-42, "-42", "-42"},
+  };
+
+  for (size_t i = 0; i < ARRAYSIZE_UNSAFE(cases); ++i) {
+    SetICUDefaultLocale("en");
+    base::testing::ResetFormatters();
+    EXPECT_EQ(cases[i].expected_english,
+              UTF16ToUTF8(base::FormatNumber(cases[i].number)));
+    SetICUDefaultLocale("de");
+    base::testing::ResetFormatters();
+    EXPECT_EQ(cases[i].expected_german,
+              UTF16ToUTF8(base::FormatNumber(cases[i].number)));
+  }
+}
+
+TEST(NumberFormattingTest, FormatDouble) {
+  static const struct {
+    double number;
+    int frac_digits;
+    const char* expected_english;
+    const char* expected_german;
+  } cases[] = {
+    {0.0, 0, "0", "0"},
+    {-0.0, 4, "-0.0000", "-0,0000"},
+    {1024.2, 0, "1,024", "1.024"},
+    {-1024.223, 2, "-1,024.22", "-1.024,22"},
+    {std::numeric_limits<double>::max(), 6,
+        "179,769,313,486,232,000,000,000,000,000,000,000,000,000,000,000,000,"
+        "000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,"
+        "000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,"
+        "000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,"
+        "000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,"
+        "000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,"
+        "000.000000",
+        "179.769.313.486.232.000.000.000.000.000.000.000.000.000.000.000.000."
+        "000.000.000.000.000.000.000.000.000.000.000.000.000.000.000.000.000."
+        "000.000.000.000.000.000.000.000.000.000.000.000.000.000.000.000.000."
+        "000.000.000.000.000.000.000.000.000.000.000.000.000.000.000.000.000."
+        "000.000.000.000.000.000.000.000.000.000.000.000.000.000.000.000.000."
+        "000.000.000.000.000.000.000.000.000.000.000.000.000.000.000.000.000."
+        "000,000000"},
+    {std::numeric_limits<double>::min(), 2, "0.00", "0,00"},
+    {-42.7, 3, "-42.700", "-42,700"},
+  };
+
+  for (size_t i = 0; i < ARRAYSIZE_UNSAFE(cases); ++i) {
+  SetICUDefaultLocale("en");
+  base::testing::ResetFormatters();
+  EXPECT_EQ(cases[i].expected_english,
+            UTF16ToUTF8(base::FormatDouble(cases[i].number,
+                                           cases[i].frac_digits)));
+  SetICUDefaultLocale("de");
+  base::testing::ResetFormatters();
+  EXPECT_EQ(cases[i].expected_german,
+            UTF16ToUTF8(base::FormatDouble(cases[i].number,
+                                           cases[i].frac_digits)));
+  }
+}
diff --git a/base/string_util.cc b/base/string_util.cc
index 81b9ee7..31c5e1a 100644
--- a/base/string_util.cc
+++ b/base/string_util.cc
@@ -604,85 +604,35 @@ bool EndsWith(const string16& str, const string16& search,
 }
 #endif
 
-DataUnits GetByteDisplayUnits(int64 bytes) {
-  // The byte thresholds at which we display amounts.  A byte count is displayed
-  // in unit U when kUnitThresholds[U] <= bytes < kUnitThresholds[U+1].
-  // This must match the DataUnits enum.
-  static const int64 kUnitThresholds[] = {
-    0,              // DATA_UNITS_BYTE,
-    3*1024,         // DATA_UNITS_KIBIBYTE,
-    2*1024*1024,    // DATA_UNITS_MEBIBYTE,
-    1024*1024*1024  // DATA_UNITS_GIBIBYTE,
-  };
-
-  if (bytes < 0) {
-    NOTREACHED() << "Negative bytes value";
-    return DATA_UNITS_BYTE;
-  }
-
-  int unit_index = arraysize(kUnitThresholds);
-  while (--unit_index > 0) {
-    if (bytes >= kUnitThresholds[unit_index])
-      break;
-  }
-
-  DCHECK(unit_index >= DATA_UNITS_BYTE && unit_index <= DATA_UNITS_GIBIBYTE);
-  return DataUnits(unit_index);
-}
-
-// TODO(mpcomplete): deal with locale
-// Byte suffixes.  This must match the DataUnits enum.
-static const char* const kByteStrings[] = {
-  "B",
-  "kB",
-  "MB",
-  "GB"
+static const char* const kByteStringsUnlocalized[] = {
+  " B",
+  " kB",
+  " MB",
+  " GB",
+  " TB",
+  " PB"
 };
 
-static const char* const kSpeedStrings[] = {
-  "B/s",
-  "kB/s",
-  "MB/s",
-  "GB/s"
-};
-
-string16 FormatBytesInternal(int64 bytes,
-                             DataUnits units,
-                             bool show_units,
-                             const char* const* suffix) {
-  if (bytes < 0) {
-    NOTREACHED() << "Negative bytes value";
-    return string16();
-  }
-
-  DCHECK(units >= DATA_UNITS_BYTE && units <= DATA_UNITS_GIBIBYTE);
-
-  // Put the quantity in the right units.
+string16 FormatBytesUnlocalized(int64 bytes) {
   double unit_amount = static_cast<double>(bytes);
-  for (int i = 0; i < units; ++i)
-    unit_amount /= 1024.0;
+  size_t dimension = 0;
+  const int kKilo = 1024;
+  while (unit_amount >= kKilo &&
+         dimension < arraysize(kByteStringsUnlocalized) - 1) {
+    unit_amount /= kKilo;
+    dimension++;
+  }
 
   char buf[64];
-  if (bytes != 0 && units != DATA_UNITS_BYTE && unit_amount < 100)
-    base::snprintf(buf, arraysize(buf), "%.1lf", unit_amount);
-  else
-    base::snprintf(buf, arraysize(buf), "%.0lf", unit_amount);
-
-  std::string ret(buf);
-  if (show_units) {
-    ret += " ";
-    ret += suffix[units];
+  if (bytes != 0 && dimension > 0 && unit_amount < 100) {
+    base::snprintf(buf, arraysize(buf), "%.1lf%s", unit_amount,
+                   kByteStringsUnlocalized[dimension]);
+  } else {
+    base::snprintf(buf, arraysize(buf), "%.0lf%s", unit_amount,
+                   kByteStringsUnlocalized[dimension]);
   }
 
-  return ASCIIToUTF16(ret);
-}
-
-string16 FormatBytes(int64 bytes, DataUnits units, bool show_units) {
-  return FormatBytesInternal(bytes, units, show_units, kByteStrings);
-}
-
-string16 FormatSpeed(int64 bytes, DataUnits units, bool show_units) {
-  return FormatBytesInternal(bytes, units, show_units, kSpeedStrings);
+  return ASCIIToUTF16(buf);
 }
 
 template<class StringType>
diff --git a/base/string_util.h b/base/string_util.h
index 0cb439c..f731c34 100644
--- a/base/string_util.h
+++ b/base/string_util.h
@@ -407,31 +407,11 @@ inline bool IsWhitespace(wchar_t c) {
   return wcschr(kWhitespaceWide, c) != NULL;
 }
 
-enum DataUnits {
-  DATA_UNITS_BYTE = 0,
-  DATA_UNITS_KIBIBYTE,
-  DATA_UNITS_MEBIBYTE,
-  DATA_UNITS_GIBIBYTE,
-};
-
-// Return the unit type that is appropriate for displaying the amount of bytes
-// passed in.
-BASE_API DataUnits GetByteDisplayUnits(int64 bytes);
-
-// Return a byte string in human-readable format, displayed in units appropriate
-// specified by 'units', with an optional unit suffix.
-// Ex: FormatBytes(512, DATA_UNITS_KIBIBYTE, true) => "0.5 KB"
-// Ex: FormatBytes(10*1024, DATA_UNITS_MEBIBYTE, false) => "0.1"
-BASE_API string16 FormatBytes(int64 bytes, DataUnits units, bool show_units);
-
-// As above, but with "/s" units.
-// Ex: FormatSpeed(512, DATA_UNITS_KIBIBYTE, true) => "0.5 KB/s"
-// Ex: FormatSpeed(10*1024, DATA_UNITS_MEBIBYTE, false) => "0.1"
-BASE_API string16 FormatSpeed(int64 bytes, DataUnits units, bool show_units);
-
-// Return a number formated with separators in the user's locale way.
-// Ex: FormatNumber(1234567) => 1,234,567
-BASE_API string16 FormatNumber(int64 number);
+// Return a byte string in human-readable format with a unit suffix. Not
+// appropriate for use in any UI; use of FormatBytes and friends in ui/base is
+// highly recommended instead. TODO(avi): Figure out how to get callers to use
+// FormatBytes instead; remove this.
+BASE_API string16 FormatBytesUnlocalized(int64 bytes);
 
 // Starting at |start_offset| (usually 0), replace the first instance of
 // |find_this| with |replace_with|.
diff --git a/base/string_util_unittest.cc b/base/string_util_unittest.cc
index 8bdeaf5..0911aae 100644
--- a/base/string_util_unittest.cc
+++ b/base/string_util_unittest.cc
@@ -491,32 +491,10 @@ TEST(StringUtilTest, LowerCaseEqualsASCII) {
   }
 }
 
-TEST(StringUtilTest, GetByteDisplayUnits) {
+TEST(StringUtilTest, FormatBytesUnlocalized) {
   static const struct {
     int64 bytes;
-    DataUnits expected;
-  } cases[] = {
-    {0, DATA_UNITS_BYTE},
-    {512, DATA_UNITS_BYTE},
-    {10*1024, DATA_UNITS_KIBIBYTE},
-    {10*1024*1024, DATA_UNITS_MEBIBYTE},
-    {10LL*1024*1024*1024, DATA_UNITS_GIBIBYTE},
-    {~(1LL<<63), DATA_UNITS_GIBIBYTE},
-#ifdef NDEBUG
-    {-1, DATA_UNITS_BYTE},
-#endif
-  };
-
-  for (size_t i = 0; i < ARRAYSIZE_UNSAFE(cases); ++i)
-    EXPECT_EQ(cases[i].expected, GetByteDisplayUnits(cases[i].bytes));
-}
-
-TEST(StringUtilTest, FormatBytes) {
-  static const struct {
-    int64 bytes;
-    DataUnits units;
     const char* expected;
-    const char* expected_with_units;
   } cases[] = {
     // Expected behavior: we show one post-decimal digit when we have
     // under two pre-decimal digits, except in cases where it makes no
@@ -524,39 +502,29 @@ TEST(StringUtilTest, FormatBytes) {
     // Since we switch units once we cross the 1000 mark, this keeps
     // the display of file sizes or bytes consistently around three
     // digits.
-    {0, DATA_UNITS_BYTE, "0", "0 B"},
-    {512, DATA_UNITS_BYTE, "512", "512 B"},
-    {512, DATA_UNITS_KIBIBYTE, "0.5", "0.5 kB"},
-    {1024*1024, DATA_UNITS_KIBIBYTE, "1024", "1024 kB"},
-    {1024*1024, DATA_UNITS_MEBIBYTE, "1.0", "1.0 MB"},
-    {1024*1024*1024, DATA_UNITS_GIBIBYTE, "1.0", "1.0 GB"},
-    {10LL*1024*1024*1024, DATA_UNITS_GIBIBYTE, "10.0", "10.0 GB"},
-    {99LL*1024*1024*1024, DATA_UNITS_GIBIBYTE, "99.0", "99.0 GB"},
-    {105LL*1024*1024*1024, DATA_UNITS_GIBIBYTE, "105", "105 GB"},
-    {105LL*1024*1024*1024 + 500LL*1024*1024, DATA_UNITS_GIBIBYTE,
-     "105", "105 GB"},
-    {~(1LL<<63), DATA_UNITS_GIBIBYTE, "8589934592", "8589934592 GB"},
-
-    {99*1024 + 103, DATA_UNITS_KIBIBYTE, "99.1", "99.1 kB"},
-    {1024*1024 + 103, DATA_UNITS_KIBIBYTE, "1024", "1024 kB"},
-    {1024*1024 + 205 * 1024, DATA_UNITS_MEBIBYTE, "1.2", "1.2 MB"},
-    {1024*1024*1024 + (927 * 1024*1024), DATA_UNITS_GIBIBYTE,
-     "1.9", "1.9 GB"},
-    {10LL*1024*1024*1024, DATA_UNITS_GIBIBYTE, "10.0", "10.0 GB"},
-    {100LL*1024*1024*1024, DATA_UNITS_GIBIBYTE, "100", "100 GB"},
-#ifdef NDEBUG
-    {-1, DATA_UNITS_BYTE, "", ""},
-#endif
+    {0, "0 B"},
+    {512, "512 B"},
+    {1024*1024, "1.0 MB"},
+    {1024*1024*1024, "1.0 GB"},
+    {10LL*1024*1024*1024, "10.0 GB"},
+    {99LL*1024*1024*1024, "99.0 GB"},
+    {105LL*1024*1024*1024, "105 GB"},
+    {105LL*1024*1024*1024 + 500LL*1024*1024, "105 GB"},
+    {~(1LL<<63), "8192 PB"},
+
+    {99*1024 + 103, "99.1 kB"},
+    {1024*1024 + 103, "1.0 MB"},
+    {1024*1024 + 205 * 1024, "1.2 MB"},
+    {1024*1024*1024 + (927 * 1024*1024), "1.9 GB"},
+    {10LL*1024*1024*1024, "10.0 GB"},
+    {100LL*1024*1024*1024, "100 GB"},
   };
 
   for (size_t i = 0; i < ARRAYSIZE_UNSAFE(cases); ++i) {
     EXPECT_EQ(ASCIIToUTF16(cases[i].expected),
-              FormatBytes(cases[i].bytes, cases[i].units, false));
-    EXPECT_EQ(ASCIIToUTF16(cases[i].expected_with_units),
-              FormatBytes(cases[i].bytes, cases[i].units, true));
+              FormatBytesUnlocalized(cases[i].bytes));
   }
 }
-
 TEST(StringUtilTest, ReplaceSubstringsAfterOffset) {
   static const struct {
     const char* str;