QTimeZone: don't use QSet, use sorted QList

QSet, as a node-based container, requires one memory allocation per element
inserted. QList, as a contiguous-memory container (at least in the case of
a QByteArray payload), requires one memory allocation per container.

The higher lookup speed might still speak for using QSet, but there are only
two uses of the sets:

1. Checking for existence (or lack thereof) of timezone names.
   For this, first generating a container full of data just to check for
   existence of one item of data is extremely wasteful. The QTZPrivate
   API should be extended to allow said lookup to be performed on the
   native data store instead.

That leaves

2. Returning a sorted(!) list(!) from the public QTimeZone API.
   There is no reason why, during the construction of those sorted
   lists, the data should be held in a set. Instead, the well-known
   technique of first cramming everything into a result container,
   which is subsequently sorted and has its duplicates removed,
   can be used here.

Saves more than 8K of text size on AMD64 stripped release builds.

Change-Id: I71c2298e94e02d55b0c9fb6f7ebeaed79a1fe2db
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>

7 files changed, 125 insertions, 95 deletions

diff --git a/src/corelib/tools/qtimezone.cpp b/src/corelib/tools/qtimezone.cpp
index b6160a3fd8..c2ee99a2d8 100644
--- a/src/corelib/tools/qtimezone.cpp
+++ b/src/corelib/tools/qtimezone.cpp
@@ -378,12 +378,10 @@ QTimeZone::QTimeZone(int offsetSeconds)
 
 QTimeZone::QTimeZone(const QByteArray &ianaId, int offsetSeconds, const QString &name,
                      const QString &abbreviation, QLocale::Country country, const QString &comment)
+    : d()
 {
-    // ianaId must be a valid ID and must not clash with the standard system names
-    if (QTimeZonePrivate::isValidId(ianaId) && !availableTimeZoneIds().contains(ianaId))
+    if (!isTimeZoneIdAvailable(ianaId))
         d = new QUtcTimeZonePrivate(ianaId, offsetSeconds, name, abbreviation, country, comment);
-    else
-        d = 0;
 }
 
 /*!
@@ -821,7 +819,20 @@ bool QTimeZone::isTimeZoneIdAvailable(const QByteArray &ianaId)
 {
     // isValidId is not strictly required, but faster to weed out invalid
     // IDs as availableTimeZoneIds() may be slow
-    return (QTimeZonePrivate::isValidId(ianaId) && (availableTimeZoneIds().contains(ianaId)));
+    if (!QTimeZonePrivate::isValidId(ianaId))
+        return false;
+    const QList<QByteArray> tzIds = availableTimeZoneIds();
+    return std::binary_search(tzIds.begin(), tzIds.end(), ianaId);
+}
+
+static QList<QByteArray> set_union(const QList<QByteArray> &l1, const QList<QByteArray> &l2)
+{
+    QList<QByteArray> result;
+    result.reserve(l1.size() + l2.size());
+    std::set_union(l1.begin(), l1.end(),
+                   l2.begin(), l2.end(),
+                   std::back_inserter(result));
+    return result;
 }
 
 /*!
@@ -832,11 +843,8 @@ bool QTimeZone::isTimeZoneIdAvailable(const QByteArray &ianaId)
 
 QList<QByteArray> QTimeZone::availableTimeZoneIds()
 {
-    QSet<QByteArray> set = QUtcTimeZonePrivate().availableTimeZoneIds()
-                           + global_tz->backend->availableTimeZoneIds();
-    QList<QByteArray> list = set.toList();
-    std::sort(list.begin(), list.end());
-    return list;
+    return set_union(QUtcTimeZonePrivate().availableTimeZoneIds(),
+                     global_tz->backend->availableTimeZoneIds());
 }
 
 /*!
@@ -852,11 +860,8 @@ QList<QByteArray> QTimeZone::availableTimeZoneIds()
 
 QList<QByteArray> QTimeZone::availableTimeZoneIds(QLocale::Country country)
 {
-    QSet<QByteArray> set = QUtcTimeZonePrivate().availableTimeZoneIds(country)
-                           + global_tz->backend->availableTimeZoneIds(country);
-    QList<QByteArray> list = set.toList();
-    std::sort(list.begin(), list.end());
-    return list;
+    return set_union(QUtcTimeZonePrivate().availableTimeZoneIds(country),
+                     global_tz->backend->availableTimeZoneIds(country));
 }
 
 /*!
@@ -868,11 +873,8 @@ QList<QByteArray> QTimeZone::availableTimeZoneIds(QLocale::Country country)
 
 QList<QByteArray> QTimeZone::availableTimeZoneIds(int offsetSeconds)
 {
-    QSet<QByteArray> set = QUtcTimeZonePrivate().availableTimeZoneIds(offsetSeconds)
-                           + global_tz->backend->availableTimeZoneIds(offsetSeconds);
-    QList<QByteArray> list = set.toList();
-    std::sort(list.begin(), list.end());
-    return list;
+    return set_union(QUtcTimeZonePrivate().availableTimeZoneIds(offsetSeconds),
+                     global_tz->backend->availableTimeZoneIds(offsetSeconds));
 }
 
 /*!
diff --git a/src/corelib/tools/qtimezoneprivate.cpp b/src/corelib/tools/qtimezoneprivate.cpp
index 337f0e6bfb..164db5ed0c 100644
--- a/src/corelib/tools/qtimezoneprivate.cpp
+++ b/src/corelib/tools/qtimezoneprivate.cpp
@@ -39,6 +39,8 @@
 #include <qdatastream.h>
 #include <qdebug.h>
 
+#include <algorithm>
+
 QT_BEGIN_NAMESPACE
 
 enum {
@@ -341,36 +343,38 @@ QByteArray QTimeZonePrivate::systemTimeZoneId() const
     return QByteArray();
 }
 
-QSet<QByteArray> QTimeZonePrivate::availableTimeZoneIds() const
+QList<QByteArray> QTimeZonePrivate::availableTimeZoneIds() const
 {
-    return QSet<QByteArray>();
+    return QList<QByteArray>();
 }
 
-QSet<QByteArray> QTimeZonePrivate::availableTimeZoneIds(QLocale::Country country) const
+QList<QByteArray> QTimeZonePrivate::availableTimeZoneIds(QLocale::Country country) const
 {
     // Default fall-back mode, use the zoneTable to find Region of know Zones
-    QSet<QByteArray> regionSet;
+    QList<QByteArray> regions;
 
     // First get all Zones in the Zones table belonging to the Region
     for (int i = 0; i < zoneDataTableSize; ++i) {
         if (zoneData(i)->country == country)
-            regionSet += ianaId(zoneData(i)).split(' ').toSet();
+            regions += ianaId(zoneData(i)).split(' ');
     }
 
-    // Then select just those that are available
-    QSet<QByteArray> set;
-    foreach (const QByteArray &ianaId, availableTimeZoneIds()) {
-        if (regionSet.contains(ianaId))
-            set << ianaId;
-    }
+    std::sort(regions.begin(), regions.end());
+    regions.erase(std::unique(regions.begin(), regions.end()), regions.end());
 
-    return set;
+    // Then select just those that are available
+    const QList<QByteArray> all = availableTimeZoneIds();
+    QList<QByteArray> result;
+    result.reserve(qMin(all.size(), regions.size()));
+    std::set_intersection(all.begin(), all.end(), regions.cbegin(), regions.cend(),
+                          std::back_inserter(result));
+    return result;
 }
 
-QSet<QByteArray> QTimeZonePrivate::availableTimeZoneIds(int offsetFromUtc) const
+QList<QByteArray> QTimeZonePrivate::availableTimeZoneIds(int offsetFromUtc) const
 {
     // Default fall-back mode, use the zoneTable to find Offset of know Zones
-    QSet<QByteArray> offsetSet;
+    QList<QByteArray> offsets;
     // First get all Zones in the table using the Offset
     for (int i = 0; i < windowsDataTableSize; ++i) {
         const QWindowsData *winData = windowsData(i);
@@ -378,19 +382,21 @@ QSet<QByteArray> QTimeZonePrivate::availableTimeZoneIds(int offsetFromUtc) const
             for (int j = 0; j < zoneDataTableSize; ++j) {
                 const QZoneData *data = zoneData(j);
                 if (data->windowsIdKey == winData->windowsIdKey)
-                    offsetSet += ianaId(data).split(' ').toSet();
+                    offsets += ianaId(data).split(' ');
             }
         }
     }
 
-    // Then select just those that are available
-    QSet<QByteArray> set;
-    foreach (const QByteArray &ianaId, availableTimeZoneIds()) {
-        if (offsetSet.contains(ianaId))
-            set << ianaId;
-    }
+    std::sort(offsets.begin(), offsets.end());
+    offsets.erase(std::unique(offsets.begin(), offsets.end()), offsets.end());
 
-    return set;
+    // Then select just those that are available
+    const QList<QByteArray> all = availableTimeZoneIds();
+    QList<QByteArray> result;
+    result.reserve(qMin(all.size(), offsets.size()));
+    std::set_intersection(all.begin(), all.end(), offsets.cbegin(), offsets.cend(),
+                          std::back_inserter(result));
+    return result;
 }
 
 #ifndef QT_NO_DATASTREAM
@@ -682,31 +688,35 @@ QByteArray QUtcTimeZonePrivate::systemTimeZoneId() const
     return utcQByteArray();
 }
 
-QSet<QByteArray> QUtcTimeZonePrivate::availableTimeZoneIds() const
+QList<QByteArray> QUtcTimeZonePrivate::availableTimeZoneIds() const
 {
-    QSet<QByteArray> set;
+    QList<QByteArray> result;
     for (int i = 0; i < utcDataTableSize; ++i)
-        set << utcId(utcData(i));
-    return set;
+        result << utcId(utcData(i));
+    std::sort(result.begin(), result.end()); // ### or already sorted??
+    // ### assuming no duplicates
+    return result;
 }
 
-QSet<QByteArray> QUtcTimeZonePrivate::availableTimeZoneIds(QLocale::Country country) const
+QList<QByteArray> QUtcTimeZonePrivate::availableTimeZoneIds(QLocale::Country country) const
 {
     // If AnyCountry then is request for all non-region offset codes
     if (country == QLocale::AnyCountry)
         return availableTimeZoneIds();
-    return QSet<QByteArray>();
+    return QList<QByteArray>();
 }
 
-QSet<QByteArray> QUtcTimeZonePrivate::availableTimeZoneIds(qint32 offsetSeconds) const
+QList<QByteArray> QUtcTimeZonePrivate::availableTimeZoneIds(qint32 offsetSeconds) const
 {
-    QSet<QByteArray> set;
+    QList<QByteArray> result;
     for (int i = 0; i < utcDataTableSize; ++i) {
         const QUtcData *data = utcData(i);
         if (data->offsetFromUtc == offsetSeconds)
-            set << utcId(data);
+            result << utcId(data);
     }
-    return set;
+    std::sort(result.begin(), result.end()); // ### or already sorted??
+    // ### assuming no duplicates
+    return result;
 }
 
 #ifndef QT_NO_DATASTREAM
diff --git a/src/corelib/tools/qtimezoneprivate_icu.cpp b/src/corelib/tools/qtimezoneprivate_icu.cpp
index 3ab42661c0..dd57464a72 100644
--- a/src/corelib/tools/qtimezoneprivate_icu.cpp
+++ b/src/corelib/tools/qtimezoneprivate_icu.cpp
@@ -37,6 +37,9 @@
 #include <unicode/ucal.h>
 
 #include <qdebug.h>
+#include <qlist.h>
+
+#include <algorithm>
 
 QT_BEGIN_NAMESPACE
 
@@ -234,19 +237,21 @@ static QTimeZonePrivate::Data ucalTimeZoneTransition(UCalendar *m_ucal,
 #endif // U_ICU_VERSION_SHORT
 
 // Convert a uenum to a QList<QByteArray>
-static QSet<QByteArray> uenumToIdSet(UEnumeration *uenum)
+static QList<QByteArray> uenumToIdList(UEnumeration *uenum)
 {
-    QSet<QByteArray> set;
+    QList<QByteArray> list;
     int32_t size = 0;
     UErrorCode status = U_ZERO_ERROR;
     // TODO Perhaps use uenum_unext instead?
     QByteArray result = uenum_next(uenum, &size, &status);
     while (U_SUCCESS(status) && !result.isEmpty()) {
-        set << result;
+        list << result;
         status = U_ZERO_ERROR;
         result = uenum_next(uenum, &size, &status);
     }
-    return set;
+    std::sort(list.begin(), list.end());
+    list.erase(std::unique(list.begin(), list.end()), list.end());
+    return list;
 }
 
 // Qt wrapper around ucal_getDSTSavings()
@@ -453,41 +458,41 @@ QByteArray QIcuTimeZonePrivate::systemTimeZoneId() const
     return ucalDefaultTimeZoneId();
 }
 
-QSet<QByteArray> QIcuTimeZonePrivate::availableTimeZoneIds() const
+QList<QByteArray> QIcuTimeZonePrivate::availableTimeZoneIds() const
 {
     UErrorCode status = U_ZERO_ERROR;
     UEnumeration *uenum = ucal_openTimeZones(&status);
-    QSet<QByteArray> set;
+    QList<QByteArray> result;
     if (U_SUCCESS(status))
-        set = uenumToIdSet(uenum);
+        result = uenumToIdList(uenum);
     uenum_close(uenum);
-    return set;
+    return result;
 }
 
-QSet<QByteArray> QIcuTimeZonePrivate::availableTimeZoneIds(QLocale::Country country) const
+QList<QByteArray> QIcuTimeZonePrivate::availableTimeZoneIds(QLocale::Country country) const
 {
     QByteArray regionCode = QLocalePrivate::countryToCode(country).toUtf8();
     UErrorCode status = U_ZERO_ERROR;
     UEnumeration *uenum = ucal_openCountryTimeZones(regionCode, &status);
-    QSet<QByteArray> set;
+    QList<QByteArray> result;
     if (U_SUCCESS(status))
-        set = uenumToIdSet(uenum);
+        result = uenumToIdList(uenum);
     uenum_close(uenum);
-    return set;
+    return result;
 }
 
-QSet<QByteArray> QIcuTimeZonePrivate::availableTimeZoneIds(int offsetFromUtc) const
+QList<QByteArray> QIcuTimeZonePrivate::availableTimeZoneIds(int offsetFromUtc) const
 {
 // TODO Available directly in C++ api but not C api, from 4.8 onwards new filter method works
 #if U_ICU_VERSION_MAJOR_NUM >= 49 || (U_ICU_VERSION_MAJOR_NUM == 4 && U_ICU_VERSION_MINOR_NUM == 8)
     UErrorCode status = U_ZERO_ERROR;
     UEnumeration *uenum = ucal_openTimeZoneIDEnumeration(UCAL_ZONE_TYPE_ANY, 0,
                                                          &offsetFromUtc, &status);
-    QSet<QByteArray> set;
+    QList<QByteArray> result;
     if (U_SUCCESS(status))
-        set = uenumToIdSet(uenum);
+        result = uenumToIdList(uenum);
     uenum_close(uenum);
-    return set;
+    return result;
 #else
     return QTimeZonePrivate::availableTimeZoneIds(offsetFromUtc);
 #endif
diff --git a/src/corelib/tools/qtimezoneprivate_mac.mm b/src/corelib/tools/qtimezoneprivate_mac.mm
index b0c3c9e64f..f316fe0a6d 100644
--- a/src/corelib/tools/qtimezoneprivate_mac.mm
+++ b/src/corelib/tools/qtimezoneprivate_mac.mm
@@ -41,6 +41,8 @@
 
 #include <qdebug.h>
 
+#include <algorithm>
+
 QT_BEGIN_NAMESPACE
 
 /*
@@ -247,18 +249,21 @@ QByteArray QMacTimeZonePrivate::systemTimeZoneId() const
     return QCFString::toQString([[NSTimeZone systemTimeZone] name]).toUtf8();
 }
 
-QSet<QByteArray> QMacTimeZonePrivate::availableTimeZoneIds() const
+QList<QByteArray> QMacTimeZonePrivate::availableTimeZoneIds() const
 {
     NSEnumerator *enumerator = [[NSTimeZone knownTimeZoneNames] objectEnumerator];
     QByteArray tzid = QCFString::toQString([enumerator nextObject]).toUtf8();
 
-    QSet<QByteArray> set;
+    QList<QByteArray> list;
     while (!tzid.isEmpty()) {
-        set << tzid;
+        list << tzid;
         tzid = QCFString::toQString([enumerator nextObject]).toUtf8();
     }
 
-    return set;
+    std::sort(list.begin(), list.end());
+    list.erase(std::unique(list.begin(), list.end()), list.end());
+
+    return list;
 }
 
 QT_END_NAMESPACE
diff --git a/src/corelib/tools/qtimezoneprivate_p.h b/src/corelib/tools/qtimezoneprivate_p.h
index 609ed50c56..64827950e2 100644
--- a/src/corelib/tools/qtimezoneprivate_p.h
+++ b/src/corelib/tools/qtimezoneprivate_p.h
@@ -126,9 +126,9 @@ public:
 
     virtual QByteArray systemTimeZoneId() const;
 
-    virtual QSet<QByteArray> availableTimeZoneIds() const;
-    virtual QSet<QByteArray> availableTimeZoneIds(QLocale::Country country) const;
-    virtual QSet<QByteArray> availableTimeZoneIds(int utcOffset) const;
+    virtual QList<QByteArray> availableTimeZoneIds() const;
+    virtual QList<QByteArray> availableTimeZoneIds(QLocale::Country country) const;
+    virtual QList<QByteArray> availableTimeZoneIds(int utcOffset) const;
 
     virtual void serialize(QDataStream &ds) const;
 
@@ -199,9 +199,9 @@ public:
 
     QByteArray systemTimeZoneId() const Q_DECL_OVERRIDE;
 
-    QSet<QByteArray> availableTimeZoneIds() const Q_DECL_OVERRIDE;
-    QSet<QByteArray> availableTimeZoneIds(QLocale::Country country) const Q_DECL_OVERRIDE;
-    QSet<QByteArray> availableTimeZoneIds(int utcOffset) const Q_DECL_OVERRIDE;
+    QList<QByteArray> availableTimeZoneIds() const Q_DECL_OVERRIDE;
+    QList<QByteArray> availableTimeZoneIds(QLocale::Country country) const Q_DECL_OVERRIDE;
+    QList<QByteArray> availableTimeZoneIds(int utcOffset) const Q_DECL_OVERRIDE;
 
     void serialize(QDataStream &ds) const Q_DECL_OVERRIDE;
 
@@ -250,9 +250,9 @@ public:
 
     QByteArray systemTimeZoneId() const Q_DECL_OVERRIDE;
 
-    QSet<QByteArray> availableTimeZoneIds() const Q_DECL_OVERRIDE;
-    QSet<QByteArray> availableTimeZoneIds(QLocale::Country country) const Q_DECL_OVERRIDE;
-    QSet<QByteArray> availableTimeZoneIds(int offsetFromUtc) const Q_DECL_OVERRIDE;
+    QList<QByteArray> availableTimeZoneIds() const Q_DECL_OVERRIDE;
+    QList<QByteArray> availableTimeZoneIds(QLocale::Country country) const Q_DECL_OVERRIDE;
+    QList<QByteArray> availableTimeZoneIds(int offsetFromUtc) const Q_DECL_OVERRIDE;
 
 private:
     void init(const QByteArray &ianaId);
@@ -318,8 +318,8 @@ public:
 
     QByteArray systemTimeZoneId() const Q_DECL_OVERRIDE;
 
-    QSet<QByteArray> availableTimeZoneIds() const Q_DECL_OVERRIDE;
-    QSet<QByteArray> availableTimeZoneIds(QLocale::Country country) const Q_DECL_OVERRIDE;
+    QList<QByteArray> availableTimeZoneIds() const Q_DECL_OVERRIDE;
+    QList<QByteArray> availableTimeZoneIds(QLocale::Country country) const Q_DECL_OVERRIDE;
 
 private:
     void init(const QByteArray &ianaId);
@@ -369,7 +369,7 @@ public:
 
     QByteArray systemTimeZoneId() const Q_DECL_OVERRIDE;
 
-    QSet<QByteArray> availableTimeZoneIds() const Q_DECL_OVERRIDE;
+    QList<QByteArray> availableTimeZoneIds() const Q_DECL_OVERRIDE;
 
 private:
     void init(const QByteArray &zoneId);
@@ -420,7 +420,7 @@ public:
 
     QByteArray systemTimeZoneId() const Q_DECL_OVERRIDE;
 
-    QSet<QByteArray> availableTimeZoneIds() const Q_DECL_OVERRIDE;
+    QList<QByteArray> availableTimeZoneIds() const Q_DECL_OVERRIDE;
 
 private:
     void init(const QByteArray &ianaId);
diff --git a/src/corelib/tools/qtimezoneprivate_tz.cpp b/src/corelib/tools/qtimezoneprivate_tz.cpp
index 0d8ae3b47f..29f0e17012 100644
--- a/src/corelib/tools/qtimezoneprivate_tz.cpp
+++ b/src/corelib/tools/qtimezoneprivate_tz.cpp
@@ -41,6 +41,7 @@
 
 #include <qdebug.h>
 
+#include <algorithm>
 
 QT_BEGIN_NAMESPACE
 
@@ -956,20 +957,23 @@ QByteArray QTzTimeZonePrivate::systemTimeZoneId() const
     return ianaId;
 }
 
-QSet<QByteArray> QTzTimeZonePrivate::availableTimeZoneIds() const
+QList<QByteArray> QTzTimeZonePrivate::availableTimeZoneIds() const
 {
-    return tzZones->keys().toSet();
+    QList<QByteArray> result = tzZones->keys();
+    std::sort(result.begin(), result.end());
+    return result;
 }
 
-QSet<QByteArray> QTzTimeZonePrivate::availableTimeZoneIds(QLocale::Country country) const
+QList<QByteArray> QTzTimeZonePrivate::availableTimeZoneIds(QLocale::Country country) const
 {
     // TODO AnyCountry
-    QSet<QByteArray> set;
+    QList<QByteArray> result;
     foreach (const QByteArray &key, tzZones->keys()) {
         if (tzZones->value(key).country == country)
-            set << key;
+            result << key;
     }
-    return set;
+    std::sort(result.begin(), result.end());
+    return result;
 }
 
 QT_END_NAMESPACE
diff --git a/src/corelib/tools/qtimezoneprivate_win.cpp b/src/corelib/tools/qtimezoneprivate_win.cpp
index 2d0caf07a8..a9bb3aa3b5 100644
--- a/src/corelib/tools/qtimezoneprivate_win.cpp
+++ b/src/corelib/tools/qtimezoneprivate_win.cpp
@@ -38,6 +38,8 @@
 
 #include "qdebug.h"
 
+#include <algorithm>
+
 QT_BEGIN_NAMESPACE
 
 /*
@@ -632,14 +634,16 @@ QByteArray QWinTimeZonePrivate::systemTimeZoneId() const
     return ianaId;
 }
 
-QSet<QByteArray> QWinTimeZonePrivate::availableTimeZoneIds() const
+QList<QByteArray> QWinTimeZonePrivate::availableTimeZoneIds() const
 {
-    QSet<QByteArray> set;
+    QList<QByteArray> result;
     foreach (const QByteArray &winId, availableWindowsIds()) {
         foreach (const QByteArray &ianaId, windowsIdToIanaIds(winId))
-            set << ianaId;
+            result << ianaId;
     }
-    return set;
+    std::sort(result.begin(), result.end());
+    result.erase(std::unique(result.begin(), result.end()), result.end());
+    return result;
 }
 
 QWinTimeZonePrivate::QWinTransitionRule QWinTimeZonePrivate::ruleForYear(int year) const