diff --git a/deps/icu-small/source/common/cmemory.h b/deps/icu-small/source/common/cmemory.h
index b5e0c2b7a17296..c9156f253cf1c7 100644
--- a/deps/icu-small/source/common/cmemory.h
+++ b/deps/icu-small/source/common/cmemory.h
@@ -725,9 +725,14 @@ class MemoryPool : public UMemory {
     }
 
     MemoryPool& operator=(MemoryPool&& other) U_NOEXCEPT {
-        fCount = other.fCount;
-        fPool = std::move(other.fPool);
-        other.fCount = 0;
+        // Since `this` may contain instances that need to be deleted, we can't
+        // just throw them away and replace them with `other`. The normal way of
+        // dealing with this in C++ is to swap `this` and `other`, rather than
+        // simply overwrite: the destruction of `other` can then take care of
+        // running MemoryPool::~MemoryPool() over the still-to-be-deallocated
+        // instances.
+        std::swap(fCount, other.fCount);
+        std::swap(fPool, other.fPool);
         return *this;
     }
 
@@ -796,9 +801,6 @@ class MemoryPool : public UMemory {
 template<typename T, int32_t stackCapacity = 8>
 class MaybeStackVector : protected MemoryPool<T, stackCapacity> {
 public:
-    using MemoryPool<T, stackCapacity>::MemoryPool;
-    using MemoryPool<T, stackCapacity>::operator=;
-
     template<typename... Args>
     T* emplaceBack(Args&&... args) {
         return this->create(args...);
diff --git a/deps/icu-small/source/common/locid.cpp b/deps/icu-small/source/common/locid.cpp
index 2804e36bf62cb1..874e4a70556f31 100644
--- a/deps/icu-small/source/common/locid.cpp
+++ b/deps/icu-small/source/common/locid.cpp
@@ -35,6 +35,7 @@
 
 #include "unicode/bytestream.h"
 #include "unicode/locid.h"
+#include "unicode/localebuilder.h"
 #include "unicode/strenum.h"
 #include "unicode/stringpiece.h"
 #include "unicode/uloc.h"
@@ -1028,7 +1029,7 @@ class AliasReplacer {
     // place the the replaced locale ID in out and return true.
     // Otherwise return false for no replacement or error.
     bool replace(
-        const Locale& locale, CharString& out, UErrorCode status);
+        const Locale& locale, CharString& out, UErrorCode& status);
 
 private:
     const char* language;
@@ -1336,10 +1337,13 @@ AliasReplacer::replaceTerritory(UVector& toBeFreed, UErrorCode& status)
         // Cannot use nullptr for language because that will construct
         // the default locale, in that case, use "und" to get the correct
         // locale.
-        Locale l(language == nullptr ? "und" : language, nullptr, script);
+        Locale l = LocaleBuilder()
+            .setLanguage(language == nullptr ? "und" : language)
+            .setScript(script)
+            .build(status);
         l.addLikelySubtags(status);
         const char* likelyRegion = l.getCountry();
-        CharString* item = nullptr;
+        LocalPointer<CharString> item;
         if (likelyRegion != nullptr && uprv_strlen(likelyRegion) > 0) {
             size_t len = uprv_strlen(likelyRegion);
             const char* foundInReplacement = uprv_strstr(replacement,
@@ -1351,20 +1355,22 @@ AliasReplacer::replaceTerritory(UVector& toBeFreed, UErrorCode& status)
                          *(foundInReplacement-1) == ' ');
                 U_ASSERT(foundInReplacement[len] == ' ' ||
                          foundInReplacement[len] == '\0');
-                item = new CharString(foundInReplacement, (int32_t)len, status);
+                item.adoptInsteadAndCheckErrorCode(
+                    new CharString(foundInReplacement, (int32_t)len, status), status);
             }
         }
-        if (item == nullptr) {
-            item = new CharString(replacement,
-                                  (int32_t)(firstSpace - replacement), status);
+        if (item.isNull() && U_SUCCESS(status)) {
+            item.adoptInsteadAndCheckErrorCode(
+                new CharString(replacement,
+                               (int32_t)(firstSpace - replacement), status), status);
         }
         if (U_FAILURE(status)) { return false; }
-        if (item == nullptr) {
+        if (item.isNull()) {
             status = U_MEMORY_ALLOCATION_ERROR;
             return false;
         }
         replacedRegion = item->data();
-        toBeFreed.addElement(item, status);
+        toBeFreed.addElement(item.orphan(), status);
     }
     U_ASSERT(!same(region, replacedRegion));
     region = replacedRegion;
@@ -1453,7 +1459,7 @@ AliasReplacer::outputToString(
         int32_t variantsStart = out.length();
         for (int32_t i = 0; i < variants.size(); i++) {
              out.append(SEP_CHAR, status)
-                 .append((const char*)((UVector*)variants.elementAt(i)),
+                 .append((const char*)(variants.elementAt(i)),
                          status);
         }
         T_CString_toUpperCase(out.data() + variantsStart);
@@ -1470,7 +1476,7 @@ AliasReplacer::outputToString(
 }
 
 bool
-AliasReplacer::replace(const Locale& locale, CharString& out, UErrorCode status)
+AliasReplacer::replace(const Locale& locale, CharString& out, UErrorCode& status)
 {
     data = AliasData::singleton(status);
     if (U_FAILURE(status)) {
@@ -2453,9 +2459,13 @@ Locale::setKeywordValue(const char* keywordName, const char* keywordValue, UErro
     if (U_FAILURE(status)) {
         return;
     }
+    if (status == U_STRING_NOT_TERMINATED_WARNING) {
+        status = U_ZERO_ERROR;
+    }
     int32_t bufferLength = uprv_max((int32_t)(uprv_strlen(fullName) + 1), ULOC_FULLNAME_CAPACITY);
     int32_t newLength = uloc_setKeywordValue(keywordName, keywordValue, fullName,
                                              bufferLength, &status) + 1;
+    U_ASSERT(status != U_STRING_NOT_TERMINATED_WARNING);
     /* Handle the case the current buffer is not enough to hold the new id */
     if (status == U_BUFFER_OVERFLOW_ERROR) {
         U_ASSERT(newLength > bufferLength);
@@ -2472,6 +2482,7 @@ Locale::setKeywordValue(const char* keywordName, const char* keywordValue, UErro
         fullName = newFullName;
         status = U_ZERO_ERROR;
         uloc_setKeywordValue(keywordName, keywordValue, fullName, newLength, &status);
+        U_ASSERT(status != U_STRING_NOT_TERMINATED_WARNING);
     } else {
         U_ASSERT(newLength <= bufferLength);
     }
diff --git a/deps/icu-small/source/common/rbbitblb.cpp b/deps/icu-small/source/common/rbbitblb.cpp
index 65b597c4a936b0..cbd8f315c252d8 100644
--- a/deps/icu-small/source/common/rbbitblb.cpp
+++ b/deps/icu-small/source/common/rbbitblb.cpp
@@ -1402,12 +1402,13 @@ void RBBITableBuilder::exportTable(void *where) {
             U_ASSERT (sd->fAccepting <= 255);
             U_ASSERT (sd->fLookAhead <= 255);
             U_ASSERT (0 <= sd->fTagsIdx && sd->fTagsIdx <= 255);
-            row->r8.fAccepting = sd->fAccepting;
-            row->r8.fLookAhead = sd->fLookAhead;
-            row->r8.fTagsIdx   = sd->fTagsIdx;
+            RBBIStateTableRow8 *r8 = (RBBIStateTableRow8*)row;
+            r8->fAccepting = sd->fAccepting;
+            r8->fLookAhead = sd->fLookAhead;
+            r8->fTagsIdx   = sd->fTagsIdx;
             for (col=0; col<catCount; col++) {
                 U_ASSERT (sd->fDtran->elementAti(col) <= kMaxStateFor8BitsTable);
-                row->r8.fNextState[col] = sd->fDtran->elementAti(col);
+                r8->fNextState[col] = sd->fDtran->elementAti(col);
             }
         } else {
             U_ASSERT (sd->fAccepting <= 0xffff);
@@ -1603,12 +1604,13 @@ void RBBITableBuilder::exportSafeTable(void *where) {
         UnicodeString *rowString = (UnicodeString *)fSafeTable->elementAt(state);
         RBBIStateTableRow   *row = (RBBIStateTableRow *)(table->fTableData + state*table->fRowLen);
         if (use8BitsForSafeTable()) {
-            row->r8.fAccepting = 0;
-            row->r8.fLookAhead = 0;
-            row->r8.fTagsIdx    = 0;
+            RBBIStateTableRow8 *r8 = (RBBIStateTableRow8*)row;
+            r8->fAccepting = 0;
+            r8->fLookAhead = 0;
+            r8->fTagsIdx    = 0;
             for (col=0; col<catCount; col++) {
                 U_ASSERT(rowString->charAt(col) <= kMaxStateFor8BitsTable);
-                row->r8.fNextState[col] = static_cast<uint8_t>(rowString->charAt(col));
+                r8->fNextState[col] = static_cast<uint8_t>(rowString->charAt(col));
             }
         } else {
             row->r16.fAccepting = 0;
diff --git a/deps/icu-small/source/common/uloc.cpp b/deps/icu-small/source/common/uloc.cpp
index 522f33dbe243a9..ebfbb506508c1b 100644
--- a/deps/icu-small/source/common/uloc.cpp
+++ b/deps/icu-small/source/common/uloc.cpp
@@ -877,6 +877,9 @@ uloc_setKeywordValue(const char* keywordName,
     if(U_FAILURE(*status)) {
         return -1;
     }
+    if (*status == U_STRING_NOT_TERMINATED_WARNING) {
+        *status = U_ZERO_ERROR;
+    }
     if (keywordName == NULL || keywordName[0] == 0 || bufferCapacity <= 1) {
         *status = U_ILLEGAL_ARGUMENT_ERROR;
         return 0;
@@ -914,6 +917,7 @@ uloc_setKeywordValue(const char* keywordName,
     startSearchHere = (char*)locale_getKeywordsStart(buffer);
     if(startSearchHere == NULL || (startSearchHere[1]==0)) {
         if(keywordValueLen == 0) { /* no keywords = nothing to remove */
+            U_ASSERT(*status != U_STRING_NOT_TERMINATED_WARNING);
             return bufLen;
         }
 
@@ -933,6 +937,7 @@ uloc_setKeywordValue(const char* keywordName,
         startSearchHere += keywordNameLen;
         *startSearchHere++ = '=';
         uprv_strcpy(startSearchHere, keywordValueBuffer);
+        U_ASSERT(*status != U_STRING_NOT_TERMINATED_WARNING);
         return needLen;
     } /* end shortcut - no @ */
 
@@ -1047,13 +1052,27 @@ uloc_setKeywordValue(const char* keywordName,
     if (!handledInputKeyAndValue || U_FAILURE(*status)) {
         /* if input key/value specified removal of a keyword not present in locale, or
          * there was an error in CharString.append, leave original locale alone. */
+        U_ASSERT(*status != U_STRING_NOT_TERMINATED_WARNING);
         return bufLen;
     }
 
     // needLen = length of the part before '@'
     needLen = (int32_t)(startSearchHere - buffer);
-    return needLen + updatedKeysAndValues.extract(
+    // Check to see can we fit the startSearchHere, if not, return
+    // U_BUFFER_OVERFLOW_ERROR without copy updatedKeysAndValues into it.
+    // We do this because this API function does not behave like most others:
+    // It promises never to set a U_STRING_NOT_TERMINATED_WARNING.
+    // When the contents fits but without the terminating NUL, in this case we need to not change
+    // the buffer contents and return with a buffer overflow error.
+    int32_t appendLength = updatedKeysAndValues.length();
+    if (appendLength >= bufferCapacity - needLen) {
+        *status = U_BUFFER_OVERFLOW_ERROR;
+        return needLen + appendLength;
+    }
+    needLen += updatedKeysAndValues.extract(
                          startSearchHere, bufferCapacity - needLen, *status);
+    U_ASSERT(*status != U_STRING_NOT_TERMINATED_WARNING);
+    return needLen;
 }
 
 /* ### ID parsing implementation **************************************************/
diff --git a/deps/icu-small/source/common/unicode/docmain.h b/deps/icu-small/source/common/unicode/docmain.h
index f09d7e1dc29cec..14491494c5ca7a 100644
--- a/deps/icu-small/source/common/unicode/docmain.h
+++ b/deps/icu-small/source/common/unicode/docmain.h
@@ -143,6 +143,11 @@
  *     <td>icu::MessageFormat</td>
  *   </tr>
  *   <tr>
+ *     <td>List Formatting</td>
+ *     <td>ulistformatter.h</td>
+ *     <td>icu::ListFormatter</td>
+ *   </tr>
+ *   <tr>
  *     <td>Number Formatting<br/>(includes currency and unit formatting)</td>
  *     <td>unumberformatter.h, unum.h</td>
  *     <td>icu::number::NumberFormatter (ICU 60+) or icu::NumberFormat (older versions)</td>
diff --git a/deps/icu-small/source/common/unicode/urename.h b/deps/icu-small/source/common/unicode/urename.h
index 20232cd209c2d6..fe59fdd893d940 100644
--- a/deps/icu-small/source/common/unicode/urename.h
+++ b/deps/icu-small/source/common/unicode/urename.h
@@ -1137,6 +1137,7 @@
 #define ulocimp_toLanguageTag U_ICU_ENTRY_POINT_RENAME(ulocimp_toLanguageTag)
 #define ulocimp_toLegacyKey U_ICU_ENTRY_POINT_RENAME(ulocimp_toLegacyKey)
 #define ulocimp_toLegacyType U_ICU_ENTRY_POINT_RENAME(ulocimp_toLegacyType)
+#define ultag_getTKeyStart U_ICU_ENTRY_POINT_RENAME(ultag_getTKeyStart)
 #define ultag_isExtensionSubtags U_ICU_ENTRY_POINT_RENAME(ultag_isExtensionSubtags)
 #define ultag_isLanguageSubtag U_ICU_ENTRY_POINT_RENAME(ultag_isLanguageSubtag)
 #define ultag_isPrivateuseValueSubtags U_ICU_ENTRY_POINT_RENAME(ultag_isPrivateuseValueSubtags)
diff --git a/deps/icu-small/source/common/unicode/uvernum.h b/deps/icu-small/source/common/unicode/uvernum.h
index a4cbb9e0fe8661..a46481a3fe610c 100644
--- a/deps/icu-small/source/common/unicode/uvernum.h
+++ b/deps/icu-small/source/common/unicode/uvernum.h
@@ -66,7 +66,7 @@
  *  This value will change in the subsequent releases of ICU
  *  @stable ICU 2.6
  */
-#define U_ICU_VERSION_MINOR_NUM 1
+#define U_ICU_VERSION_MINOR_NUM 2
 
 /** The current ICU patchlevel version as an integer.
  *  This value will change in the subsequent releases of ICU
@@ -139,7 +139,7 @@
  *  This value will change in the subsequent releases of ICU
  *  @stable ICU 2.4
  */
-#define U_ICU_VERSION "68.1"
+#define U_ICU_VERSION "68.2"
 
 /**
  * The current ICU library major version number as a string, for library name suffixes.
@@ -158,7 +158,7 @@
 /** Data version in ICU4C.
  * @internal ICU 4.4 Internal Use Only
  **/
-#define U_ICU_DATA_VERSION "68.1"
+#define U_ICU_DATA_VERSION "68.2"
 #endif  /* U_HIDE_INTERNAL_API */
 
 /*===========================================================================
diff --git a/deps/icu-small/source/common/wintz.cpp b/deps/icu-small/source/common/wintz.cpp
index 047f4290f10d0e..6805297a38a161 100644
--- a/deps/icu-small/source/common/wintz.cpp
+++ b/deps/icu-small/source/common/wintz.cpp
@@ -36,17 +36,58 @@
 
 U_NAMESPACE_BEGIN
 
+// Note these constants and the struct are only used when dealing with the fallback path for RDP sesssions.
+
+// This is the location of the time zones in the registry on Vista+ systems.
+// See: https://docs.microsoft.com/windows/win32/api/timezoneapi/ns-timezoneapi-dynamic_time_zone_information
+#define WINDOWS_TIMEZONES_REG_KEY_PATH L"SOFTWARE\\Microsoft\\Windows NT\\CurrentVersion\\Time Zones"
+
+// Max length for a registry key is 255. +1 for null.
+// See: https://docs.microsoft.com/windows/win32/sysinfo/registry-element-size-limits
+#define WINDOWS_MAX_REG_KEY_LENGTH 256
+
+#if U_PLATFORM_HAS_WINUWP_API == 0
+
+// This is the layout of the TZI binary value in the registry.
+// See: https://docs.microsoft.com/windows/win32/api/timezoneapi/ns-timezoneapi-time_zone_information
+typedef struct _REG_TZI_FORMAT {
+    LONG Bias;
+    LONG StandardBias;
+    LONG DaylightBias;
+    SYSTEMTIME StandardDate;
+    SYSTEMTIME DaylightDate;
+} REG_TZI_FORMAT;
+
+#endif // U_PLATFORM_HAS_WINUWP_API
+
 /**
-* Main Windows time zone detection function.
-* Returns the Windows time zone converted to an ICU time zone as a heap-allocated buffer, or nullptr upon failure.
+* This is main Windows time zone detection function.
+*
+* It returns the Windows time zone converted to an ICU time zone as a heap-allocated buffer, or nullptr upon failure.
+*
+* We use the Win32 API GetDynamicTimeZoneInformation (which is available since Vista) to get the current time zone info,
+* as this API returns a non-localized time zone name which can be then mapped to an ICU time zone.
+*
+* However, in some RDP/terminal services situations, this struct isn't always fully complete, and the TimeZoneKeyName
+* field of the struct might be NULL. This can happen with some 3rd party RDP clients, and also when using older versions
+* of the RDP protocol, which don't send the newer TimeZoneKeyNamei information and only send the StandardName and DaylightName.
+*
+* Since these 3rd party clients and older RDP clients only send the pre-Vista time zone information to the server, this means that we
+* need to fallback on using the pre-Vista methods to determine the time zone. This unfortunately requires examining the registry directly
+* in order to try and determine the current time zone.
+*
+* Note that this can however still fail in some cases though if the client and server are using different languages, as the StandardName
+* that is sent by client is localized in the client's language. However, we must compare this to the names that are on the server, which
+* are localized in registry using the server's language. Despite that, this is the best we can do.
 *
-* Note: We use the Win32 API GetDynamicTimeZoneInformation (available since Vista+) to get the current time zone info.
-* This API returns a non-localized time zone name, which is mapped to an ICU time zone ID (~ Olsen ID).
+* Note: This fallback method won't work for the UWP version though, as we can't use the registry APIs in UWP.
+*
+* Once we have the current Windows time zone, then we can then map it to an ICU time zone ID (~ Olsen ID).
 */
 U_CAPI const char* U_EXPORT2
 uprv_detectWindowsTimeZone()
 {
-    // Obtain the DYNAMIC_TIME_ZONE_INFORMATION info to get the non-localized time zone name.
+    // We first try to obtain the time zone directly by using the TimeZoneKeyName field of the DYNAMIC_TIME_ZONE_INFORMATION struct.
     DYNAMIC_TIME_ZONE_INFORMATION dynamicTZI;
     uprv_memset(&dynamicTZI, 0, sizeof(dynamicTZI));
     SYSTEMTIME systemTimeAllZero;
@@ -86,22 +127,138 @@ uprv_detectWindowsTimeZone()
             // Note '-' before 'utcOffsetMin'. The timezone ID's sign convention
             // is that a timezone ahead of UTC is Etc/GMT-<offset> and a timezone
             // behind UTC is Etc/GMT+<offset>.
-            int ret = snprintf(gmtOffsetTz, UPRV_LENGTHOF(gmtOffsetTz), "Etc/GMT%+d", -utcOffsetMins / 60);
+            int ret = snprintf(gmtOffsetTz, UPRV_LENGTHOF(gmtOffsetTz), "Etc/GMT%+ld", -utcOffsetMins / 60);
             if (ret > 0 && ret < UPRV_LENGTHOF(gmtOffsetTz)) {
                 return uprv_strdup(gmtOffsetTz);
             }
         }
     }
 
-    // If DST is NOT disabled, but we have an empty TimeZoneKeyName, then it is unclear
-    // what we should do as this should not happen.
+    // If DST is NOT disabled, but the TimeZoneKeyName field of the struct is NULL, then we may be dealing with a
+    // RDP/terminal services session where the 'Time Zone Redirection' feature is enabled. However, either the RDP
+    // client sent the server incomplete info (some 3rd party RDP clients only send the StandardName and  DaylightName,
+    // but do not send the important TimeZoneKeyName), or if the RDP server has not appropriately populated the struct correctly.
+    //
+    // In this case we unfortunately have no choice but to fallback to using the pre-Vista method of determining the
+    // time zone, which requires examining the registry directly.
+    //
+    // Note that this can however still fail though if the client and server are using different languages, as the StandardName
+    // that is sent by client is *localized* in the client's language. However, we must compare this to the names that are
+    // on the server, which are *localized* in registry using the server's language.
+    //
+    // One other note is that this fallback method doesn't work for the UWP version, as we can't use the registry APIs.
+
+    // windowsTimeZoneName will point at timezoneSubKeyName if we had to fallback to using the registry, and we found a match.
+    WCHAR timezoneSubKeyName[WINDOWS_MAX_REG_KEY_LENGTH];
+    WCHAR *windowsTimeZoneName = dynamicTZI.TimeZoneKeyName;
+
     if (dynamicTZI.TimeZoneKeyName[0] == 0) {
+
+// We can't use the registry APIs in the UWP version.
+#if U_PLATFORM_HAS_WINUWP_API == 1
+        (void)timezoneSubKeyName; // suppress unused variable warnings.
         return nullptr;
+#else
+        // Open the path to the time zones in the Windows registry.
+        LONG ret;
+        HKEY hKeyAllTimeZones = nullptr;
+        ret = RegOpenKeyExW(HKEY_LOCAL_MACHINE, WINDOWS_TIMEZONES_REG_KEY_PATH, 0, KEY_READ,
+                            reinterpret_cast<PHKEY>(&hKeyAllTimeZones));
+
+        if (ret != ERROR_SUCCESS) {
+            // If we can't open the key, then we can't do much, so fail.
+            return nullptr;
+        }
+
+        // Read the number of subkeys under the time zone registry path.
+        DWORD numTimeZoneSubKeys;
+        ret = RegQueryInfoKeyW(hKeyAllTimeZones, nullptr, nullptr, nullptr, &numTimeZoneSubKeys,
+                               nullptr, nullptr, nullptr, nullptr, nullptr, nullptr, nullptr);
+
+        if (ret != ERROR_SUCCESS) {
+            RegCloseKey(hKeyAllTimeZones);
+            return nullptr;
+        }
+
+        // Examine each of the subkeys to try and find a match for the localized standard name ("Std").
+        //
+        // Note: The name of the time zone subkey itself is not localized, but the "Std" name is localized. This means
+        // that we could fail to find a match if the RDP client and RDP server are using different languages, but unfortunately
+        // there isn't much we can do about it.
+        HKEY hKeyTimeZoneSubKey = nullptr;
+        ULONG registryValueType;
+        WCHAR registryStandardName[WINDOWS_MAX_REG_KEY_LENGTH];
+
+        for (DWORD i = 0; i < numTimeZoneSubKeys; i++) {
+            // Note: RegEnumKeyExW wants the size of the buffer in characters.
+            DWORD size = UPRV_LENGTHOF(timezoneSubKeyName);
+            ret = RegEnumKeyExW(hKeyAllTimeZones, i, timezoneSubKeyName, &size, nullptr, nullptr, nullptr, nullptr);
+
+            if (ret != ERROR_SUCCESS) {
+                RegCloseKey(hKeyAllTimeZones);
+                return nullptr;
+            }
+
+            ret = RegOpenKeyExW(hKeyAllTimeZones, timezoneSubKeyName, 0, KEY_READ,
+                                reinterpret_cast<PHKEY>(&hKeyTimeZoneSubKey));
+
+            if (ret != ERROR_SUCCESS) {
+                RegCloseKey(hKeyAllTimeZones);
+                return nullptr;
+            }
+
+            // Note: RegQueryValueExW wants the size of the buffer in bytes.
+            size = sizeof(registryStandardName);
+            ret = RegQueryValueExW(hKeyTimeZoneSubKey, L"Std", nullptr, &registryValueType,
+                                   reinterpret_cast<LPBYTE>(registryStandardName), &size);
+
+            if (ret != ERROR_SUCCESS || registryValueType != REG_SZ) {
+                RegCloseKey(hKeyTimeZoneSubKey);
+                RegCloseKey(hKeyAllTimeZones);
+                return nullptr;
+            }
+
+            // Note: wcscmp does an ordinal (byte) comparison.
+            if (wcscmp(reinterpret_cast<WCHAR *>(registryStandardName), dynamicTZI.StandardName) == 0) {
+                // Since we are comparing the *localized* time zone name, it's possible that some languages might use
+                // the same string for more than one time zone. Thus we need to examine the TZI data in the registry to
+                // compare the GMT offset (the bias), and the DST transition dates, to ensure it's the same time zone
+                // as the currently reported one.
+                REG_TZI_FORMAT registryTziValue;
+                uprv_memset(&registryTziValue, 0, sizeof(registryTziValue));
+
+                // Note: RegQueryValueExW wants the size of the buffer in bytes.
+                DWORD timezoneTziValueSize = sizeof(registryTziValue);
+                ret = RegQueryValueExW(hKeyTimeZoneSubKey, L"TZI", nullptr, &registryValueType,
+                                     reinterpret_cast<LPBYTE>(&registryTziValue), &timezoneTziValueSize);
+
+                if (ret == ERROR_SUCCESS) {
+                    if ((dynamicTZI.Bias == registryTziValue.Bias) &&
+                        (memcmp((const void *)&dynamicTZI.StandardDate, (const void *)&registryTziValue.StandardDate, sizeof(SYSTEMTIME)) == 0) &&
+                        (memcmp((const void *)&dynamicTZI.DaylightDate, (const void *)&registryTziValue.DaylightDate, sizeof(SYSTEMTIME)) == 0))
+                    {
+                        // We found a matching time zone.
+                        windowsTimeZoneName = timezoneSubKeyName;
+                        break;
+                    }
+                }
+            }
+            RegCloseKey(hKeyTimeZoneSubKey);
+            hKeyTimeZoneSubKey = nullptr;
+        }
+
+        if (hKeyTimeZoneSubKey != nullptr) {
+            RegCloseKey(hKeyTimeZoneSubKey);
+        }
+        if (hKeyAllTimeZones != nullptr) {
+            RegCloseKey(hKeyAllTimeZones);
+        }
+#endif // U_PLATFORM_HAS_WINUWP_API
     }
 
     CharString winTZ;
     UErrorCode status = U_ZERO_ERROR;
-    winTZ.appendInvariantChars(UnicodeString(TRUE, dynamicTZI.TimeZoneKeyName, -1), status);
+    winTZ.appendInvariantChars(UnicodeString(TRUE, windowsTimeZoneName, -1), status);
 
     // Map Windows Timezone name (non-localized) to ICU timezone ID (~ Olson timezone id).
     StackUResourceBundle winTZBundle;
@@ -123,18 +280,29 @@ uprv_detectWindowsTimeZone()
     int regionCodeLen = GetGeoInfoW(geoId, GEO_ISO2, regionCodeW, UPRV_LENGTHOF(regionCodeW), 0);
 
     const UChar *icuTZ16 = nullptr;
-    int32_t tzLen;
+    int32_t tzListLen = 0;
 
     if (regionCodeLen != 0) {
         for (int i = 0; i < UPRV_LENGTHOF(regionCodeW); i++) {
             regionCode[i] = static_cast<char>(regionCodeW[i]);
         }
-        icuTZ16 = ures_getStringByKey(winTZBundle.getAlias(), regionCode, &tzLen, &status);
+        icuTZ16 = ures_getStringByKey(winTZBundle.getAlias(), regionCode, &tzListLen, &status);
     }
     if (regionCodeLen == 0 || U_FAILURE(status)) {
         // fallback to default "001" (world)
         status = U_ZERO_ERROR;
-        icuTZ16 = ures_getStringByKey(winTZBundle.getAlias(), "001", &tzLen, &status);
+        icuTZ16 = ures_getStringByKey(winTZBundle.getAlias(), "001", &tzListLen, &status);
+    }
+
+    // Note: We want the first entry in the string returned by ures_getStringByKey.
+    // However this string can be a space delimited list of timezones:
+    //  Ex: "America/New_York America/Detroit America/Indiana/Petersburg ..."
+    // We need to stop at the first space, so we pass tzLen (instead of tzListLen) to appendInvariantChars below.
+    int32_t tzLen = 0;
+    if (tzListLen > 0) {
+        while (!(icuTZ16[tzLen] == u'\0' || icuTZ16[tzLen] == u' ')) {
+            tzLen++;
+        }
     }
 
     // Note: cloneData returns nullptr if the status is a failure, so this
diff --git a/deps/icu-small/source/data/in/icudt68l.dat.bz2 b/deps/icu-small/source/data/in/icudt68l.dat.bz2
index fcee4ecfa17efe..8fd32b7471d648 100644
Binary files a/deps/icu-small/source/data/in/icudt68l.dat.bz2 and b/deps/icu-small/source/data/in/icudt68l.dat.bz2 differ
diff --git a/deps/icu-small/source/i18n/dtitvfmt.cpp b/deps/icu-small/source/i18n/dtitvfmt.cpp
index a913dc43c808d6..d6ec501af88701 100644
--- a/deps/icu-small/source/i18n/dtitvfmt.cpp
+++ b/deps/icu-small/source/i18n/dtitvfmt.cpp
@@ -1422,7 +1422,11 @@ DateIntervalFormat::setIntervalPattern(UCalendarDateFields field,
         if ( field == UCAL_AM_PM ) {
             fInfo->getIntervalPattern(*bestSkeleton, UCAL_HOUR, pattern,status);
             if ( !pattern.isEmpty() ) {
-                setIntervalPattern(field, pattern);
+                UBool suppressDayPeriodField = fSkeleton.indexOf(CAP_J) != -1;
+                UnicodeString adjustIntervalPattern;
+                adjustFieldWidth(*skeleton, *bestSkeleton, pattern, differenceInfo,
+                                 suppressDayPeriodField, adjustIntervalPattern);
+                setIntervalPattern(field, adjustIntervalPattern);
             }
             return false;
         }
@@ -1694,27 +1698,23 @@ DateIntervalFormat::adjustFieldWidth(const UnicodeString& inputSkeleton,
     DateIntervalInfo::parseSkeleton(inputSkeleton, inputSkeletonFieldWidth);
     DateIntervalInfo::parseSkeleton(bestMatchSkeleton, bestMatchSkeletonFieldWidth);
     if (suppressDayPeriodField) {
-        adjustedPtn.findAndReplace(UnicodeString(LOW_A), UnicodeString());
-        adjustedPtn.findAndReplace(UnicodeString("  "), UnicodeString(" "));
+        findReplaceInPattern(adjustedPtn, UnicodeString(LOW_A), UnicodeString());
+        findReplaceInPattern(adjustedPtn, UnicodeString("  "), UnicodeString(" "));
         adjustedPtn.trim();
     }
     if ( differenceInfo == 2 ) {
         if (inputSkeleton.indexOf(LOW_Z) != -1) {
-            adjustedPtn.findAndReplace(UnicodeString(LOW_V),
-                                       UnicodeString(LOW_Z));
-        }
-        if (inputSkeleton.indexOf(CAP_K) != -1) {
-            adjustedPtn.findAndReplace(UnicodeString(LOW_H),
-                                       UnicodeString(CAP_K));
-        }
-        if (inputSkeleton.indexOf(LOW_K) != -1) {
-            adjustedPtn.findAndReplace(UnicodeString(CAP_H),
-                                       UnicodeString(LOW_K));
-        }
-        if (inputSkeleton.indexOf(LOW_B) != -1) {
-            adjustedPtn.findAndReplace(UnicodeString(LOW_A),
-                                       UnicodeString(LOW_B));
-        }
+             findReplaceInPattern(adjustedPtn, UnicodeString(LOW_V), UnicodeString(LOW_Z));
+         }
+         if (inputSkeleton.indexOf(CAP_K) != -1) {
+             findReplaceInPattern(adjustedPtn, UnicodeString(LOW_H), UnicodeString(CAP_K));
+         }
+         if (inputSkeleton.indexOf(LOW_K) != -1) {
+             findReplaceInPattern(adjustedPtn, UnicodeString(CAP_H), UnicodeString(LOW_K));
+         }
+         if (inputSkeleton.indexOf(LOW_B) != -1) {
+             findReplaceInPattern(adjustedPtn, UnicodeString(LOW_A), UnicodeString(LOW_B));
+         }
     }
     if (adjustedPtn.indexOf(LOW_A) != -1 && bestMatchSkeletonFieldWidth[LOW_A - PATTERN_CHAR_BASE] == 0) {
         bestMatchSkeletonFieldWidth[LOW_A - PATTERN_CHAR_BASE] = 1;
@@ -1792,6 +1792,39 @@ DateIntervalFormat::adjustFieldWidth(const UnicodeString& inputSkeleton,
     }
 }
 
+void
+DateIntervalFormat::findReplaceInPattern(UnicodeString& targetString,
+                                         const UnicodeString& strToReplace,
+                                         const UnicodeString& strToReplaceWith) {
+    int32_t firstQuoteIndex = targetString.indexOf(u'\'');
+    if (firstQuoteIndex == -1) {
+        targetString.findAndReplace(strToReplace, strToReplaceWith);
+    } else {
+        UnicodeString result;
+        UnicodeString source = targetString;
+
+        while (firstQuoteIndex >= 0) {
+            int32_t secondQuoteIndex = source.indexOf(u'\'', firstQuoteIndex + 1);
+            if (secondQuoteIndex == -1) {
+                secondQuoteIndex = source.length() - 1;
+            }
+
+            UnicodeString unquotedText(source, 0, firstQuoteIndex);
+            UnicodeString quotedText(source, firstQuoteIndex, secondQuoteIndex - firstQuoteIndex + 1);
+
+            unquotedText.findAndReplace(strToReplace, strToReplaceWith);
+            result += unquotedText;
+            result += quotedText;
+
+            source.remove(0, secondQuoteIndex + 1);
+            firstQuoteIndex = source.indexOf(u'\'');
+        }
+        source.findAndReplace(strToReplace, strToReplaceWith);
+        result += source;
+        targetString = result;
+    }
+}
+
 
 
 void
diff --git a/deps/icu-small/source/i18n/formatted_string_builder.cpp b/deps/icu-small/source/i18n/formatted_string_builder.cpp
index 5aabc31cc4391b..b370f14f2ac4ff 100644
--- a/deps/icu-small/source/i18n/formatted_string_builder.cpp
+++ b/deps/icu-small/source/i18n/formatted_string_builder.cpp
@@ -276,6 +276,11 @@ int32_t FormattedStringBuilder::prepareForInsertHelper(int32_t index, int32_t co
     char16_t *oldChars = getCharPtr();
     Field *oldFields = getFieldPtr();
     if (fLength + count > oldCapacity) {
+        if ((fLength + count) > INT32_MAX / 2) {
+            // If we continue, then newCapacity will overlow int32_t in the next line.
+            status = U_INPUT_TOO_LONG_ERROR;
+            return -1;
+        }
         int32_t newCapacity = (fLength + count) * 2;
         int32_t newZero = newCapacity / 2 - (fLength + count) / 2;
 
@@ -330,12 +335,14 @@ int32_t FormattedStringBuilder::prepareForInsertHelper(int32_t index, int32_t co
         fZero = newZero;
         fLength += count;
     }
+    U_ASSERT((fZero + index) >= 0);
     return fZero + index;
 }
 
 int32_t FormattedStringBuilder::remove(int32_t index, int32_t count) {
     // TODO: Reset the heap here?  (If the string after removal can fit on stack?)
     int32_t position = index + fZero;
+    U_ASSERT(position >= 0);
     uprv_memmove2(getCharPtr() + position,
             getCharPtr() + position + count,
             sizeof(char16_t) * (fLength - index - count));
diff --git a/deps/icu-small/source/i18n/formattedval_impl.h b/deps/icu-small/source/i18n/formattedval_impl.h
index 8005b0abb4d848..1e6eb1e639f809 100644
--- a/deps/icu-small/source/i18n/formattedval_impl.h
+++ b/deps/icu-small/source/i18n/formattedval_impl.h
@@ -117,6 +117,12 @@ class FormattedValueFieldPositionIteratorImpl : public UMemory, public Formatted
 };
 
 
+// Internal struct that must be exported for MSVC
+struct U_I18N_API SpanInfo {
+    int32_t spanValue;
+    int32_t length;
+};
+
 // Export an explicit template instantiation of the MaybeStackArray that
 //    is used as a data member of CEBuffer.
 //
@@ -126,7 +132,7 @@ class FormattedValueFieldPositionIteratorImpl : public UMemory, public Formatted
 // See digitlst.h, pluralaffix.h, datefmt.h, and others for similar examples.
 //
 #if U_PF_WINDOWS <= U_PLATFORM && U_PLATFORM <= U_PF_CYGWIN
-template class U_I18N_API MaybeStackArray<int32_t, 8>;
+template class U_I18N_API MaybeStackArray<SpanInfo, 8>;
 #endif
 
 /**
@@ -162,13 +168,19 @@ class U_I18N_API FormattedValueStringBuilderImpl : public UMemory, public Format
         return fString;
     }
 
-    void appendSpanIndex(int32_t index);
-    void prependSpanIndex(int32_t index);
+    /**
+     * Adds additional metadata used for span fields.
+     *
+     * spanValue: the index of the list item, for example.
+     * length: the length of the span, used to split adjacent fields.
+     */
+    void appendSpanInfo(int32_t spanValue, int32_t length, UErrorCode& status);
+    void prependSpanInfo(int32_t spanValue, int32_t length, UErrorCode& status);
 
 private:
     FormattedStringBuilder fString;
     FormattedStringBuilder::Field fNumericField;
-    MaybeStackArray<int32_t, 8> spanIndices;
+    MaybeStackArray<SpanInfo, 8> spanIndices;
 
     bool nextPositionImpl(ConstrainedFieldPosition& cfpos, FormattedStringBuilder::Field numericField, UErrorCode& status) const;
     static bool isIntOrGroup(FormattedStringBuilder::Field field);
diff --git a/deps/icu-small/source/i18n/formattedval_sbimpl.cpp b/deps/icu-small/source/i18n/formattedval_sbimpl.cpp
index b2ae4c34c0a66d..84c2d00666c2be 100644
--- a/deps/icu-small/source/i18n/formattedval_sbimpl.cpp
+++ b/deps/icu-small/source/i18n/formattedval_sbimpl.cpp
@@ -46,19 +46,19 @@ Appendable& FormattedValueStringBuilderImpl::appendTo(Appendable& appendable, UE
 
 UBool FormattedValueStringBuilderImpl::nextPosition(ConstrainedFieldPosition& cfpos, UErrorCode& status) const {
     // NOTE: MSVC sometimes complains when implicitly converting between bool and UBool
-    return nextPositionImpl(cfpos, fNumericField, status) ? TRUE : FALSE;
+    return nextPositionImpl(cfpos, fNumericField, status) ? true : false;
 }
 
 UBool FormattedValueStringBuilderImpl::nextFieldPosition(FieldPosition& fp, UErrorCode& status) const {
     int32_t rawField = fp.getField();
 
     if (rawField == FieldPosition::DONT_CARE) {
-        return FALSE;
+        return false;
     }
 
     if (rawField < 0 || rawField >= UNUM_FIELD_COUNT) {
         status = U_ILLEGAL_ARGUMENT_ERROR;
-        return FALSE;
+        return false;
     }
 
     ConstrainedFieldPosition cfpos;
@@ -67,7 +67,7 @@ UBool FormattedValueStringBuilderImpl::nextFieldPosition(FieldPosition& fp, UErr
     if (nextPositionImpl(cfpos, kUndefinedField, status)) {
         fp.setBeginIndex(cfpos.getStart());
         fp.setEndIndex(cfpos.getLimit());
-        return TRUE;
+        return true;
     }
 
     // Special case: fraction should start after integer if fraction is not present
@@ -85,7 +85,7 @@ UBool FormattedValueStringBuilderImpl::nextFieldPosition(FieldPosition& fp, UErr
         fp.setEndIndex(i - fString.fZero);
     }
 
-    return FALSE;
+    return false;
 }
 
 void FormattedValueStringBuilderImpl::getAllFieldPositions(FieldPositionIteratorHandler& fpih,
@@ -103,23 +103,12 @@ static constexpr Field kEndField = Field(0xf, 0xf);
 bool FormattedValueStringBuilderImpl::nextPositionImpl(ConstrainedFieldPosition& cfpos, Field numericField, UErrorCode& /*status*/) const {
     int32_t fieldStart = -1;
     Field currField = kUndefinedField;
-    UFieldCategory spanCategory = UFIELD_CATEGORY_UNDEFINED;
-    int32_t spanValue;
     for (int32_t i = fString.fZero + cfpos.getLimit(); i <= fString.fZero + fString.fLength; i++) {
         Field _field = (i < fString.fZero + fString.fLength) ? fString.getFieldPtr()[i] : kEndField;
         // Case 1: currently scanning a field.
         if (currField != kUndefinedField) {
             if (currField != _field) {
                 int32_t end = i - fString.fZero;
-                // Handle span fields; don't trim them
-                if (spanCategory != UFIELD_CATEGORY_UNDEFINED) {
-                    cfpos.setState(
-                        spanCategory,
-                        spanValue,
-                        fieldStart,
-                        end);
-                    return true;
-                }
                 // Grouping separators can be whitespace; don't throw them out!
                 if (isTrimmable(currField)) {
                     end = trimBack(i - fString.fZero);
@@ -182,13 +171,11 @@ bool FormattedValueStringBuilderImpl::nextPositionImpl(ConstrainedFieldPosition&
             if (elementField == Field(UFIELD_CATEGORY_LIST, ULISTFMT_ELEMENT_FIELD)
                     && cfpos.matchesField(elementField.getCategory(), elementField.getField())
                     && (cfpos.getLimit() < i - fString.fZero || cfpos.getCategory() != elementField.getCategory())) {
-                // Re-wind to the beginning of the field and then emit it
-                int32_t j = i - 1;
-                for (; j >= fString.fZero && fString.getFieldPtr()[j] == fString.getFieldPtr()[i-1]; j--) {}
+                int64_t si = cfpos.getInt64IterationContext() - 1;
                 cfpos.setState(
                     elementField.getCategory(),
                     elementField.getField(),
-                    j - fString.fZero + 1,
+                    i - fString.fZero - spanIndices[si].length,
                     i - fString.fZero);
                 return true;
             }
@@ -203,22 +190,28 @@ bool FormattedValueStringBuilderImpl::nextPositionImpl(ConstrainedFieldPosition&
         }
         // Case 3: check for field starting at this position
         // Case 3a: Need to add a SpanField
-        if (_field == Field(UFIELD_CATEGORY_LIST, ULISTFMT_ELEMENT_FIELD)
-                // don't return the same field twice in a row:
-                && (i == fString.fZero
-                    || fString.getFieldPtr()[i-1].getCategory() != UFIELD_CATEGORY_LIST
-                    || fString.getFieldPtr()[i-1].getField() != ULISTFMT_ELEMENT_FIELD)) {
+        if (_field == Field(UFIELD_CATEGORY_LIST, ULISTFMT_ELEMENT_FIELD)) {
             int64_t si = cfpos.getInt64IterationContext();
-            spanValue = spanIndices[si];
+            int32_t spanValue = spanIndices[si].spanValue;
+            int32_t length = spanIndices[si].length;
             cfpos.setInt64IterationContext(si + 1);
             if (cfpos.matchesField(UFIELD_CATEGORY_LIST_SPAN, spanValue)) {
-                spanCategory = UFIELD_CATEGORY_LIST_SPAN;
+                UFieldCategory spanCategory = UFIELD_CATEGORY_LIST_SPAN;
                 fieldStart = i - fString.fZero;
-                currField = _field;
+                int32_t end = fieldStart + length;
+                cfpos.setState(
+                    spanCategory,
+                    spanValue,
+                    fieldStart,
+                    end);
+                return true;
+            } else {
+                // Failed to match; jump ahead
+                i += length - 1;
                 continue;
             }
         }
-        // Case 3b: No SpanField or SpanField did not match
+        // Case 3b: No SpanField
         if (cfpos.matchesField(_field.getCategory(), _field.getField())) {
             fieldStart = i - fString.fZero;
             currField = _field;
@@ -226,24 +219,40 @@ bool FormattedValueStringBuilderImpl::nextPositionImpl(ConstrainedFieldPosition&
     }
 
     U_ASSERT(currField == kUndefinedField);
+    // Always set the position to the end so that we don't revisit previous sections
+    cfpos.setState(
+        cfpos.getCategory(),
+        cfpos.getField(),
+        fString.fLength,
+        fString.fLength);
     return false;
 }
 
-void FormattedValueStringBuilderImpl::appendSpanIndex(int32_t position) {
-    if (spanIndices.getCapacity() <= position) {
-        spanIndices.resize(position * 2);
+void FormattedValueStringBuilderImpl::appendSpanInfo(int32_t spanValue, int32_t length, UErrorCode& status) {
+    if (U_FAILURE(status)) { return; }
+    U_ASSERT(spanIndices.getCapacity() >= spanValue);
+    if (spanIndices.getCapacity() == spanValue) {
+        if (!spanIndices.resize(spanValue * 2, spanValue)) {
+            status = U_MEMORY_ALLOCATION_ERROR;
+            return;
+        }
     }
-    spanIndices[position] = position;
+    spanIndices[spanValue] = {spanValue, length};
 }
 
-void FormattedValueStringBuilderImpl::prependSpanIndex(int32_t position) {
-    if (spanIndices.getCapacity() <= position) {
-        spanIndices.resize(position * 2);
+void FormattedValueStringBuilderImpl::prependSpanInfo(int32_t spanValue, int32_t length, UErrorCode& status) {
+    if (U_FAILURE(status)) { return; }
+    U_ASSERT(spanIndices.getCapacity() >= spanValue);
+    if (spanIndices.getCapacity() == spanValue) {
+        if (!spanIndices.resize(spanValue * 2, spanValue)) {
+            status = U_MEMORY_ALLOCATION_ERROR;
+            return;
+        }
     }
-    for (int32_t i = 0; i < position; i++) {
+    for (int32_t i = spanValue - 1; i >= 0; i--) {
         spanIndices[i+1] = spanIndices[i];
     }
-    spanIndices[0] = position;
+    spanIndices[0] = {spanValue, length};
 }
 
 bool FormattedValueStringBuilderImpl::isIntOrGroup(Field field) {
diff --git a/deps/icu-small/source/i18n/listformatter.cpp b/deps/icu-small/source/i18n/listformatter.cpp
index ab04ac9080a5ab..be0d16bc7f52b3 100644
--- a/deps/icu-small/source/i18n/listformatter.cpp
+++ b/deps/icu-small/source/i18n/listformatter.cpp
@@ -16,6 +16,10 @@
 *   created by: Umesh P. Nair
 */
 
+#include "unicode/utypes.h"
+
+#if !UCONFIG_NO_FORMATTING
+
 #include "cmemory.h"
 #include "unicode/fpositer.h"  // FieldPositionIterator
 #include "unicode/listformatter.h"
@@ -171,21 +175,21 @@ PatternHandler* createPatternHandler(
     UErrorCode& status) {
     if (uprv_strcmp(lang, "es") == 0) {
         // Spanish
-        UnicodeString spanishYStr(TRUE, spanishY, -1);
+        UnicodeString spanishYStr(true, spanishY, -1);
         bool twoIsY = two == spanishYStr;
         bool endIsY = end == spanishYStr;
         if (twoIsY || endIsY) {
-            UnicodeString replacement(TRUE, spanishE, -1);
+            UnicodeString replacement(true, spanishE, -1);
             return new ContextualHandler(
                 shouldChangeToE,
                 twoIsY ? replacement : two, two,
                 endIsY ? replacement : end, end, status);
         }
-        UnicodeString spanishOStr(TRUE, spanishO, -1);
+        UnicodeString spanishOStr(true, spanishO, -1);
         bool twoIsO = two == spanishOStr;
         bool endIsO = end == spanishOStr;
         if (twoIsO || endIsO) {
-            UnicodeString replacement(TRUE, spanishU, -1);
+            UnicodeString replacement(true, spanishU, -1);
             return new ContextualHandler(
                 shouldChangeToU,
                 twoIsO ? replacement : two, two,
@@ -193,11 +197,11 @@ PatternHandler* createPatternHandler(
         }
     } else if (uprv_strcmp(lang, "he") == 0 || uprv_strcmp(lang, "iw") == 0) {
         // Hebrew
-        UnicodeString hebrewVavStr(TRUE, hebrewVav, -1);
+        UnicodeString hebrewVavStr(true, hebrewVav, -1);
         bool twoIsVav = two == hebrewVavStr;
         bool endIsVav = end == hebrewVavStr;
         if (twoIsVav || endIsVav) {
-            UnicodeString replacement(TRUE, hebrewVavDash, -1);
+            UnicodeString replacement(true, hebrewVavDash, -1);
             return new ContextualHandler(
                 shouldChangeToVavDash,
                 twoIsVav ? replacement : two, two,
@@ -238,7 +242,6 @@ ListFormatInternal(const ListFormatInternal &other) :
 };
 
 
-#if !UCONFIG_NO_FORMATTING
 class FormattedListData : public FormattedValueStringBuilderImpl {
 public:
     FormattedListData(UErrorCode&) : FormattedValueStringBuilderImpl(kUndefinedField) {}
@@ -248,7 +251,6 @@ class FormattedListData : public FormattedValueStringBuilderImpl {
 FormattedListData::~FormattedListData() = default;
 
 UPRV_FORMATTED_VALUE_SUBCLASS_AUTO_IMPL(FormattedList)
-#endif
 
 
 static Hashtable* listPatternHash = nullptr;
@@ -257,7 +259,7 @@ U_CDECL_BEGIN
 static UBool U_CALLCONV uprv_listformatter_cleanup() {
     delete listPatternHash;
     listPatternHash = nullptr;
-    return TRUE;
+    return true;
 }
 
 static void U_CALLCONV
@@ -350,7 +352,6 @@ const ListFormatInternal* ListFormatter::getListFormatInternal(
     return result;
 }
 
-#if !UCONFIG_NO_FORMATTING
 static const char* typeWidthToStyleString(UListFormatterType type, UListFormatterWidth width) {
     switch (type) {
         case ULISTFMT_TYPE_AND:
@@ -394,7 +395,6 @@ static const char* typeWidthToStyleString(UListFormatterType type, UListFormatte
 
     return nullptr;
 }
-#endif
 
 static const UChar solidus = 0x2F;
 static const UChar aliasPrefix[] = { 0x6C,0x69,0x73,0x74,0x50,0x61,0x74,0x74,0x65,0x72,0x6E,0x2F }; // "listPattern/"
@@ -515,14 +515,9 @@ ListFormatter* ListFormatter::createInstance(UErrorCode& errorCode) {
 }
 
 ListFormatter* ListFormatter::createInstance(const Locale& locale, UErrorCode& errorCode) {
-#if !UCONFIG_NO_FORMATTING
     return createInstance(locale, ULISTFMT_TYPE_AND, ULISTFMT_WIDTH_WIDE, errorCode);
-#else
-    return createInstance(locale, "standard", errorCode);
-#endif
 }
 
-#if !UCONFIG_NO_FORMATTING
 ListFormatter* ListFormatter::createInstance(
         const Locale& locale, UListFormatterType type, UListFormatterWidth width, UErrorCode& errorCode) {
     const char* style = typeWidthToStyleString(type, width);
@@ -532,7 +527,6 @@ ListFormatter* ListFormatter::createInstance(
     }
     return createInstance(locale, style, errorCode);
 }
-#endif
 
 ListFormatter* ListFormatter::createInstance(const Locale& locale, const char *style, UErrorCode& errorCode) {
     const ListFormatInternal* listFormatInternal = getListFormatInternal(locale, style, errorCode);
@@ -573,7 +567,7 @@ class FormattedListBuilder {
                 start,
                 {UFIELD_CATEGORY_LIST, ULISTFMT_ELEMENT_FIELD},
                 status);
-            data->appendSpanIndex(0);
+            data->appendSpanInfo(0, start.length(), status);
         }
     }
 
@@ -609,7 +603,7 @@ class FormattedListBuilder {
                 next,
                 {UFIELD_CATEGORY_LIST, ULISTFMT_ELEMENT_FIELD},
                 status);
-            data->appendSpanIndex(position);
+            data->appendSpanInfo(position, next.length(), status);
             data->getStringRef().append(
                 temp.tempSubString(offsets[1]),
                 {UFIELD_CATEGORY_LIST, ULISTFMT_LITERAL_FIELD},
@@ -628,7 +622,7 @@ class FormattedListBuilder {
                 next,
                 {UFIELD_CATEGORY_LIST, ULISTFMT_ELEMENT_FIELD},
                 status);
-            data->prependSpanIndex(position);
+            data->prependSpanInfo(position, next.length(), status);
             data->getStringRef().insert(
                 0,
                 temp.tempSubStringBetween(0, offsets[1]),
@@ -660,7 +654,6 @@ UnicodeString& ListFormatter::format(
         int32_t index,
         int32_t &offset,
         UErrorCode& errorCode) const {
-#if !UCONFIG_NO_FORMATTING
     int32_t initialOffset = appendTo.length();
     auto result = formatStringsToValue(items, nItems, errorCode);
     UnicodeStringAppendable appendable(appendTo);
@@ -671,11 +664,9 @@ UnicodeString& ListFormatter::format(
         result.nextPosition(cfpos, errorCode);
         offset = initialOffset + cfpos.getStart();
     }
-#endif
     return appendTo;
 }
 
-#if !UCONFIG_NO_FORMATTING
 FormattedList ListFormatter::formatStringsToValue(
         const UnicodeString items[],
         int32_t nItems,
@@ -741,7 +732,8 @@ FormattedList ListFormatter::formatStringsToValue(
         return FormattedList(result.data.orphan());
     }
 }
-#endif
 
 
 U_NAMESPACE_END
+
+#endif /* #if !UCONFIG_NO_FORMATTING */
diff --git a/deps/icu-small/source/i18n/measunit.cpp b/deps/icu-small/source/i18n/measunit.cpp
index dab3abb5e21ff6..ece83177625513 100644
--- a/deps/icu-small/source/i18n/measunit.cpp
+++ b/deps/icu-small/source/i18n/measunit.cpp
@@ -33,7 +33,8 @@ UOBJECT_DEFINE_RTTI_IMPLEMENTATION(MeasureUnit)
 // update this code, refer to:
 // http://site.icu-project.org/design/formatting/measureformat/updating-measure-unit
 //
-// Start generated code for measunit.cpp
+// Start generated code
+// TODO(ICU-21076): improve how this generated code is produced.
 
 // Maps from Type ID to offset in gSubTypes.
 static const int32_t gOffsets[] = {
@@ -54,15 +55,44 @@ static const int32_t gOffsets[] = {
     404,
     408,
     423,
-    424,
-    430,
-    440,
-    444,
-    448,
+    426,
+    432,
+    442,
+    446,
     450,
-    484
+    452,
+    486
 };
 
+// TODO: FIX CODE GENERATION - leaving this here but commented-out to make it
+// clear that we no longer want this array. We needed it for only one thing: efficient checking of "currency".
+//
+// static const int32_t gIndexes[] = {
+//     0,
+//     2,
+//     7,
+//     17,
+//     25,
+//     29,
+//     29,
+//     40,
+//     56,
+//     60,
+//     69,
+//     71,
+//     75,
+//     83,
+//     105,
+//     109,
+//     124,
+//     127,
+//     133,
+//     143,
+//     147,
+//     151,
+//     153,
+//     187
+// };
 static const int32_t kCurrencyOffset = 5;
 
 // Must be sorted alphabetically.
@@ -517,7 +547,9 @@ static const char * const gSubTypes[] = {
     "solar-mass",
     "stone",
     "ton",
-    "",
+    "", // TODO(ICU-21076): manual edit of what should have been generated by Java.
+    "percent", // TODO(ICU-21076): regenerate, deal with duplication.
+    "permille", // TODO(ICU-21076): regenerate, deal with duplication.
     "gigawatt",
     "horsepower",
     "kilowatt",
@@ -580,6 +612,8 @@ static const char * const gSubTypes[] = {
     "teaspoon"
 };
 
+// unitPerUnitToSingleUnit no longer in use! TODO: remove from code-generation code.
+
 // Shortcuts to the base unit in order to make the default constructor fast
 static const int32_t kBaseTypeIdx = 16;
 static const int32_t kBaseSubTypeIdx = 0;
@@ -2056,7 +2090,7 @@ MeasureUnit MeasureUnit::getTeaspoon() {
     return MeasureUnit(22, 33);
 }
 
-// End generated code for measunit.cpp
+// End generated code
 
 static int32_t binarySearch(
         const char * const * array, int32_t start, int32_t end, StringPiece key) {
@@ -2105,7 +2139,9 @@ MeasureUnit &MeasureUnit::operator=(const MeasureUnit &other) {
     if (this == &other) {
         return *this;
     }
-    delete fImpl;
+    if (fImpl != nullptr) {
+        delete fImpl;
+    }
     if (other.fImpl) {
         ErrorCode localStatus;
         fImpl = new MeasureUnitImpl(other.fImpl->copy(localStatus));
@@ -2126,7 +2162,9 @@ MeasureUnit &MeasureUnit::operator=(MeasureUnit &&other) noexcept {
     if (this == &other) {
         return *this;
     }
-    delete fImpl;
+    if (fImpl != nullptr) {
+        delete fImpl;
+    }
     fImpl = other.fImpl;
     other.fImpl = nullptr;
     fTypeId = other.fTypeId;
@@ -2139,8 +2177,10 @@ MeasureUnit *MeasureUnit::clone() const {
 }
 
 MeasureUnit::~MeasureUnit() {
-    delete fImpl;
-    fImpl = nullptr;
+    if (fImpl != nullptr) {
+        delete fImpl;
+        fImpl = nullptr;
+    }
 }
 
 const char *MeasureUnit::getType() const {
@@ -2237,11 +2277,9 @@ StringEnumeration* MeasureUnit::getAvailableTypes(UErrorCode &errorCode) {
 }
 
 bool MeasureUnit::findBySubType(StringPiece subType, MeasureUnit* output) {
-    // Sanity checking kCurrencyOffset and final entry in gOffsets
-    U_ASSERT(uprv_strcmp(gTypes[kCurrencyOffset], "currency") == 0);
-    U_ASSERT(gOffsets[UPRV_LENGTHOF(gOffsets) - 1] == UPRV_LENGTHOF(gSubTypes));
-
     for (int32_t t = 0; t < UPRV_LENGTHOF(gOffsets) - 1; t++) {
+        // Ensure kCurrencyOffset is set correctly
+        U_ASSERT(uprv_strcmp(gTypes[kCurrencyOffset], "currency") == 0);
         // Skip currency units
         if (t == kCurrencyOffset) {
             continue;
@@ -2298,8 +2336,10 @@ void MeasureUnit::initCurrency(StringPiece isoCurrency) {
 void MeasureUnit::setTo(int32_t typeId, int32_t subTypeId) {
     fTypeId = typeId;
     fSubTypeId = subTypeId;
-    delete fImpl;
-    fImpl = nullptr;
+    if (fImpl != nullptr) {
+        delete fImpl;
+        fImpl = nullptr;
+    }
 }
 
 int32_t MeasureUnit::getOffset() const {
diff --git a/deps/icu-small/source/i18n/number_skeletons.cpp b/deps/icu-small/source/i18n/number_skeletons.cpp
index e6d94d27b2ba5b..028525a589db91 100644
--- a/deps/icu-small/source/i18n/number_skeletons.cpp
+++ b/deps/icu-small/source/i18n/number_skeletons.cpp
@@ -732,6 +732,7 @@ skeleton::parseStem(const StringSegment& segment, const UCharsTrie& stemTrie, Se
 
         case STEM_CURRENCY:
             CHECK_NULL(seen, unit, status);
+            CHECK_NULL(seen, perUnit, status);
             return STATE_CURRENCY_UNIT;
 
         case STEM_INTEGER_WIDTH:
@@ -1500,32 +1501,33 @@ bool GeneratorHelpers::notation(const MacroProps& macros, UnicodeString& sb, UEr
 }
 
 bool GeneratorHelpers::unit(const MacroProps& macros, UnicodeString& sb, UErrorCode& status) {
-    if (utils::unitIsCurrency(macros.unit)) {
+    MeasureUnit unit = macros.unit;
+    if (!utils::unitIsBaseUnit(macros.perUnit)) {
+        if (utils::unitIsCurrency(macros.unit) || utils::unitIsCurrency(macros.perUnit)) {
+            status = U_UNSUPPORTED_ERROR;
+            return false;
+        }
+        unit = unit.product(macros.perUnit.reciprocal(status), status);
+    }
+
+    if (utils::unitIsCurrency(unit)) {
         sb.append(u"currency/", -1);
-        CurrencyUnit currency(macros.unit, status);
+        CurrencyUnit currency(unit, status);
         if (U_FAILURE(status)) {
             return false;
         }
         blueprint_helpers::generateCurrencyOption(currency, sb, status);
         return true;
-    } else if (utils::unitIsBaseUnit(macros.unit)) {
+    } else if (utils::unitIsBaseUnit(unit)) {
         // Default value is not shown in normalized form
         return false;
-    } else if (utils::unitIsPercent(macros.unit)) {
+    } else if (utils::unitIsPercent(unit)) {
         sb.append(u"percent", -1);
         return true;
-    } else if (utils::unitIsPermille(macros.unit)) {
+    } else if (utils::unitIsPermille(unit)) {
         sb.append(u"permille", -1);
         return true;
     } else {
-        MeasureUnit unit = macros.unit;
-        if (utils::unitIsCurrency(macros.perUnit)) {
-            status = U_UNSUPPORTED_ERROR;
-            return false;
-        }
-        if (!utils::unitIsBaseUnit(macros.perUnit)) {
-            unit = unit.product(macros.perUnit.reciprocal(status), status);
-        }
         sb.append(u"unit/", -1);
         sb.append(unit.getIdentifier());
         return true;
diff --git a/deps/icu-small/source/i18n/unicode/dtitvfmt.h b/deps/icu-small/source/i18n/unicode/dtitvfmt.h
index 3d20d8e9c44990..4a1ab801a04c9d 100644
--- a/deps/icu-small/source/i18n/unicode/dtitvfmt.h
+++ b/deps/icu-small/source/i18n/unicode/dtitvfmt.h
@@ -1037,6 +1037,17 @@ class U_I18N_API DateIntervalFormat : public Format {
                             UBool suppressDayPeriodField,
                             UnicodeString& adjustedIntervalPattern);
 
+    /**
+     * Does the same thing as UnicodeString::findAndReplace(), except that it won't perform
+     * the substitution inside quoted literal text.
+     * @param targetString The string to perform the find-replace operation on.
+     * @param strToReplace The string to search for and replace in the target string.
+     * @param strToReplaceWith The string to substitute in wherever `stringToReplace` was found.
+     */
+    static void U_EXPORT2 findReplaceInPattern(UnicodeString& targetString,
+                                               const UnicodeString& strToReplace,
+                                               const UnicodeString& strToReplaceWith);
+
     /**
      * Concat a single date pattern with a time interval pattern,
      * set it into the intervalPatterns, while field is time field.
diff --git a/deps/icu-small/source/i18n/unicode/listformatter.h b/deps/icu-small/source/i18n/unicode/listformatter.h
index a969a8744dcf58..eddb5dab6701b3 100644
--- a/deps/icu-small/source/i18n/unicode/listformatter.h
+++ b/deps/icu-small/source/i18n/unicode/listformatter.h
@@ -23,6 +23,8 @@
 
 #if U_SHOW_CPLUSPLUS_API
 
+#if !UCONFIG_NO_FORMATTING
+
 #include "unicode/unistr.h"
 #include "unicode/locid.h"
 #include "unicode/formattedvalue.h"
@@ -65,7 +67,6 @@ struct ListFormatData : public UMemory {
  */
 
 
-#if !UCONFIG_NO_FORMATTING
 /**
  * An immutable class containing the result of a list formatting operation.
  *
@@ -135,7 +136,6 @@ class U_I18N_API FormattedList : public UMemory, public FormattedValue {
         : fData(nullptr), fErrorCode(errorCode) {}
     friend class ListFormatter;
 };
-#endif // !UCONFIG_NO_FORMATTING
 
 
 /**
@@ -185,8 +185,6 @@ class U_I18N_API ListFormatter : public UObject{
      */
     static ListFormatter* createInstance(const Locale& locale, UErrorCode& errorCode);
 
-#ifndef U_HIDE_DRAFT_API
-#if !UCONFIG_NO_FORMATTING
     /**
      * Creates a ListFormatter for the given locale, list type, and style.
      *
@@ -195,12 +193,10 @@ class U_I18N_API ListFormatter : public UObject{
      * @param width The width of formatting to use.
      * @param errorCode ICU error code, set if no data available for the given locale.
      * @return A ListFormatter object created from internal data derived from CLDR data.
-     * @draft ICU 67
+     * @stable ICU 67
      */
     static ListFormatter* createInstance(
       const Locale& locale, UListFormatterType type, UListFormatterWidth width, UErrorCode& errorCode);
-#endif  /* !UCONFIG_NO_FORMATTING */
-#endif  /* U_HIDE_DRAFT_API */
 
 #ifndef U_HIDE_INTERNAL_API
     /**
@@ -239,7 +235,6 @@ class U_I18N_API ListFormatter : public UObject{
     UnicodeString& format(const UnicodeString items[], int32_t n_items,
         UnicodeString& appendTo, UErrorCode& errorCode) const;
 
-#if !UCONFIG_NO_FORMATTING
     /**
      * Formats a list of strings to a FormattedList, which exposes field
      * position information. The FormattedList contains more information than
@@ -255,7 +250,6 @@ class U_I18N_API ListFormatter : public UObject{
         const UnicodeString items[],
         int32_t n_items,
         UErrorCode& errorCode) const;
-#endif // !UCONFIG_NO_FORMATTING
 
 #ifndef U_HIDE_INTERNAL_API
     /**
@@ -296,6 +290,8 @@ class U_I18N_API ListFormatter : public UObject{
 
 U_NAMESPACE_END
 
+#endif /* #if !UCONFIG_NO_FORMATTING */
+
 #endif /* U_SHOW_CPLUSPLUS_API */
 
 #endif // __LISTFORMATTER_H__
diff --git a/deps/icu-small/source/i18n/unicode/measunit.h b/deps/icu-small/source/i18n/unicode/measunit.h
index b9f732ae99011d..ed8773c7710f3e 100644
--- a/deps/icu-small/source/i18n/unicode/measunit.h
+++ b/deps/icu-small/source/i18n/unicode/measunit.h
@@ -3519,6 +3519,7 @@ class U_I18N_API MeasureUnit: public UObject {
      */
     static MeasureUnit getTeaspoon();
 
+
 // End generated createXXX methods
 
  protected:
diff --git a/deps/icu-small/source/i18n/unicode/ulistformatter.h b/deps/icu-small/source/i18n/unicode/ulistformatter.h
index 3dfa9f2d56171f..28a1e580370006 100644
--- a/deps/icu-small/source/i18n/unicode/ulistformatter.h
+++ b/deps/icu-small/source/i18n/unicode/ulistformatter.h
@@ -62,17 +62,16 @@ typedef enum UListFormatterField {
     ULISTFMT_ELEMENT_FIELD
 } UListFormatterField;
 
-#ifndef U_HIDE_DRAFT_API
 /**
  * Type of meaning expressed by the list.
  *
- * @draft ICU 67
+ * @stable ICU 67
  */
 typedef enum UListFormatterType {
     /**
      * Conjunction formatting, e.g. "Alice, Bob, Charlie, and Delta".
      *
-     * @draft ICU 67
+     * @stable ICU 67
      */
     ULISTFMT_TYPE_AND,
 
@@ -80,14 +79,14 @@ typedef enum UListFormatterType {
      * Disjunction (or alternative, or simply one of) formatting, e.g.
      * "Alice, Bob, Charlie, or Delta".
      *
-     * @draft ICU 67
+     * @stable ICU 67
      */
     ULISTFMT_TYPE_OR,
 
     /**
      * Formatting of a list of values with units, e.g. "5 pounds, 12 ounces".
      *
-     * @draft ICU 67
+     * @stable ICU 67
      */
     ULISTFMT_TYPE_UNITS
 } UListFormatterType;
@@ -95,29 +94,28 @@ typedef enum UListFormatterType {
 /**
  * Verbosity level of the list patterns.
  *
- * @draft ICU 67
+ * @stable ICU 67
  */
 typedef enum UListFormatterWidth {
     /**
      * Use list formatting with full words (no abbreviations) when possible.
      *
-     * @draft ICU 67
+     * @stable ICU 67
      */
     ULISTFMT_WIDTH_WIDE,
 
     /**
      * Use list formatting of typical length.
-     * @draft ICU 67
+     * @stable ICU 67
      */
     ULISTFMT_WIDTH_SHORT,
 
     /**
      * Use list formatting of the shortest possible length.
-     * @draft ICU 67
+     * @stable ICU 67
      */
     ULISTFMT_WIDTH_NARROW,
 } UListFormatterWidth;
-#endif /* U_HIDE_DRAFT_API */
 
 /**
  * Open a new UListFormatter object using the rules for a given locale.
@@ -141,7 +139,6 @@ U_CAPI UListFormatter* U_EXPORT2
 ulistfmt_open(const char*  locale,
               UErrorCode*  status);
 
-#ifndef U_HIDE_DRAFT_API
 /**
  * Open a new UListFormatter object appropriate for the given locale, list type,
  * and style.
@@ -162,12 +159,11 @@ ulistfmt_open(const char*  locale,
  * @return
  *            A pointer to a UListFormatter object for the specified locale,
  *            or NULL if an error occurred.
- * @draft ICU 67
+ * @stable ICU 67
  */
 U_CAPI UListFormatter* U_EXPORT2
 ulistfmt_openForType(const char*  locale, UListFormatterType type,
                      UListFormatterWidth width, UErrorCode*  status);
-#endif /* U_HIDE_DRAFT_API */
 
 /**
  * Close a UListFormatter object. Once closed it may no longer be used.
diff --git a/tools/icu/current_ver.dep b/tools/icu/current_ver.dep
index b4caab129071e1..1c7f45879621c0 100644
--- a/tools/icu/current_ver.dep
+++ b/tools/icu/current_ver.dep
@@ -1,6 +1,6 @@
 [
   {
-    "url": "https://github.com/unicode-org/icu/releases/download/release-68-1/icu4c-68_1-src.tgz",
-    "md5": "6a99b541ea01f271257b121a4433c7c0"
+    "url": "https://github.com/unicode-org/icu/releases/download/release-68-2/icu4c-68_2-src.tgz",
+    "md5": "c21cbdfe31a1e325afe765a16f907d20"
   }
 ]