An algebraic data type (ADT) is an immutable composite data structure with a well-known interface. It combines values into a meaningful shape and lets operations compose around that shape. Here, “type” describes an object's structure and behavior, not necessarily a statically checked type. section 5.5 section 1.3
Two common forms express different relationships:
- A record, or product type, holds values that belong together: an immutable Pair might hold a username AND password. Unlike a plain array, it communicates their relationship. section 5.5.1
- A choice, or sum type, represents one of several alternatives: success OR failure, never both. It gives validation and error handling explicit states instead of relying on a Boolean. section 5.5.2
The practical payoff is composition when an operation might fail. Repeated validation checks or try/catch blocks interrupt an ordinary function pipeline. A Validation choice can carry a successful value through mapped operations; on failure, later mapped operations are skipped while the error remains available. section 5.0 section 1.3 section 8.2.2
A composable validation chain
Validation.of(block).flatMap(checkLength(64)).flatMap(checkTampering).flatMap(checkDifficulty).flatMap(checkLinkage(previousHash)).flatMap(checkTimestamps(previousTimestamp)); Each check becomes a step in one chain rather than another layer of conditional control flow. section 5.0
Promises offer a related way to think about this pattern: after pending, a Promise settles as fulfilled or rejected, much as Validation represents success or failure. Promise.then composes work on the fulfilled path, while Promise.catch handles rejection. An algebraic data type is not the same thing as an abstract data type such as a Stack or Queue. section 8.2.2 section 5.5
