Details
-
Task
-
Resolution: Unresolved
-
P3: Somewhat important
-
None
-
None
Description
Acceptance criterion:
- all existing tests still pass
- the following new tests don't fail:
{
constexpr auto table = qOffsetStringArray("Foo", "oo", "");
QVERIFY(table.viewAt(0) == "Foo");
QVERIFY(table.viewAt(1) == "oo");
QVERIFY(table.viewAt(2) == "");
QCOMPARE(table.m_string.size(), 4);
}
{
constexpr auto table = qOffsetStringArray("oo", "", "Foo");
QVERIFY(table.viewAt(0) == "oo");
QVERIFY(table.viewAt(1) == "");
QVERIFY(table.viewAt(2) == "Foo");
QCOMPARE(table.m_string.size(), 4);
}
{
constexpr auto table = qOffsetStringArray("Foo", "Zoo", "oo", "");
QVERIFY(table.viewAt(0) == "Foo");
QVERIFY(table.viewAt(1) == "Zoo");
QVERIFY(table.viewAt(2) == "oo");
QVERIFY(table.viewAt(3) == "");
QCOMPARE(table.m_string.size(), 8);
}
{
constexpr auto table = qOffsetStringArray("oo", "", "Zoo", "Foo");
QVERIFY(table.viewAt(0) == "oo");
QVERIFY(table.viewAt(1) == "");
QVERIFY(table.viewAt(2) == "Zoo");
QVERIFY(table.viewAt(3) == "Foo");
QCOMPARE(table.m_string.size(), 8);
}
To maintain prefetcher-friendliness, it would be desirable if the string data would not have to be re-arranged, except when a later longer string replaces an existing earlier suffix of it. That is, qOffsetStringArray("a", "ba", "dba", "aba").m_strings == "dba\0aba", not == "aba\0dba", reeking of a sorting artifact.