Ruby’s question-mark boolean methods are inconsistent

By convention, functions that return true or false have a name which ends with a question mark. The grammar of those function names is odd, though. Compare the following:

• Array#empty?
• Hash#key?(k)
• File#exist?(f)

When read out loud, the full sentence sounds different for each:

• Array#empty? — array is empty
• Hash#key?(k) — hash has key k
• File#exist?(f) — file f does exist

The (unofficial) Ruby style guide suggests exist? over exists? (extra s), but Rails only provides exists? as method names.

An alternative to key? is has_key?, which the (unofficial) Ruby style guide discourages. For consistency with the #exist?, it’d have to be have_key? which is not used anywhere, as far as I can tell, probably because it sounds weird.

For grammatical consistency, the methods would better be named as follows:

• Array#is_empty?
• Hash#has_key?(k)
• File#exists?(f)

But in these new forms, there’s not much of a reason left to include a trailing ? . The meaning of these functions is clear without it. Consider:

• Array#empty — used for emptying an array
• Array#is_empty — used for checking whether an array is empty
• Car#key — used for keying a car (what kind of example is this, Denis?!)
• Car#keyed — used for returning a keyed copy of a car
• Car#is_keyed — used for checking if the car is keyed.

(I don’t own a car, by the way.)

References

• ruby-core:43765 ruby-trunk - Feature #5555Rejected rename #include? to #includes?