You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Previously, it turns out, we didn't actually test the reverse sorted keys.
Instead we just renamed the `Reversed = Sorted` and then tested the `Sorted`
case.
But it's even worse than that because in order to "force" the test to pass we
had effectively disable the check so make it count element whether they were
there or not (Count + 1) in this snippet:
```diff
FoldFun = fun(Element, {[HAcc | TAcc], Count}) ->
case Element == HAcc of
true -> {ok, {TAcc, Count + 1}};
- _ -> {ok, {TAcc, Count + 1}}
+ _ -> {ok, {TAcc, Count}}
end
end,
```
Luckily, it seems we survived with the code being correct as far as fixing the
test didn't reveal any actual btree logic errors.
In addition, we were missing tests trying out various chunk sizes, so added two
extra cases: one with much smaller chunks, and another with much larger chunks.
0 commit comments