Thinking about Stream and Enum in Terms of Function Composition

The post reads Enum and Stream as tools for function composition rather than as generalized arrays. Enum is built from a collection’s basic functions, especially reduce, and its helpers are shortcuts over that mechanism.

Reading notes

#

• Enum’s basic protocol methods are count, member?, and reduce, not an each method.
• Enum uses reduce to implement higher-level operations such as all?.
• These operations can stop early when the result is already determined.
• Stream is similar, but it does not execute reduce right away.
• Stream creates an anonymous function that transforms one collection into another.
• A Stream-based collection can be used where a regular collection is used.
• The example with 1..10 mapped by Stream shows lazy evaluation when Enum.take is called.
• A collection defined by Stream does not imply state and remains immutable.
• Each use of a Stream starts from the beginning again.
• The choice to use Stream depends on the tradeoff between storing a collection and creating it at runtime.
• Stream supports reduce, but count and member? from Enumerable fail directly.
• Enum.count can emulate count by using reduce.