I often come across code like this which is bad, from a performance POV:

```
QVector<Type> l1 = ...;
QVector<Type> l2 = ...;
foreach(const auto& item : l1 + l2) { ... }
```

Instead, it would be faster to use a lambda or a nested loop to get rid of the temporary allocation. E.g.:

```
QVector<Type> l1 = ...;
QVector<Type> l2 = ...;
for (const auto& list : {l1, l2}) {
    for (const auto& item : list) { ... }
}
```