Cause

On this page

The Effect<A, E, R> type is polymorphic in values of type E, which means we can work with any error type we want. However, there is additional information related to failures that is not captured by the E value alone.

To preserve and provide comprehensive information about failures, Effect uses the Cause<E> data type. Cause is responsible for storing various details, such as:

  • Unexpected errors or defects
  • Stack and execution traces
  • Causes of fiber interruptions

Effect is designed to be strict in preserving all the information related to a failure. It captures and stores the complete story of failure in the Cause data type. This ensures that no information about the failure is lost, allowing us to precisely determine what happened during the execution of our effects.

Although we don't typically work directly with Cause values, it is an underlying data type that represents errors occurring within an Effect workflow. It provides us with total access to all concurrent and sequential errors within our codebase. This gives us the ability to analyze and handle failures in a comprehensive manner whenever needed.

Creating Causes

We can intentionally create effects with specific causes using the Effect.failCause constructor:

ts
import { Effect, Cause } from "effect"
 
// Create an effect that intentionally fails with an empty cause
const effect = Effect.failCause(Cause.empty)
ts
import { Effect, Cause } from "effect"
 
// Create an effect that intentionally fails with an empty cause
const effect = Effect.failCause(Cause.empty)

To uncover the underlying cause of an effect, we can use the Effect.cause function:

ts
Effect.cause(effect).pipe(
Effect.flatMap((cause) => ...)
)
ts
Effect.cause(effect).pipe(
Effect.flatMap((cause) => ...)
)

Cause Variations

There are several causes for various errors, in this section, we will describe each of these causes.

Empty

The Empty cause represents a lack of errors.

Fail

The Fail cause represents a Cause which failed with an expected error of type E.

Die

The Die cause represents a Cause which failed as a result of a defect, or in other words, an unexpected error.

Interrupt

The Interrupt cause represents failure due to Fiber interruption, which contains the FiberId of the interrupted Fiber.

Sequential

The Sequential cause represents the composition of two causes which occurred sequentially.

For example, if we perform Effect's analog of try-finally (i.e. Effect.ensuring), and both the try and finally blocks fail, we have two errors which occurred sequentially. In these cases, the errors can be represented by the Sequential cause.

Parallel

The Parallel cause represents the composition of two causes which occurred in parallel.

In Effect programs, it is possible that two operations may be performed in parallel. In these cases, the Effect workflow can fail for more than one reason. If both computations fail, then there are actually two errors which occurred in parallel. In these cases, the errors can be represented by the Parallel cause.

Guards

To identify the type of a Cause, you can use specific guards provided by the Cause module:

  • Cause.isEmpty
  • Cause.isFailType
  • Cause.isDie
  • Cause.isInterruptType
  • Cause.isSequentialType
  • Cause.isParallelType

Let's see an example of how you can utilize these guards:

ts
import { Cause } from "effect"
 
const cause = Cause.fail(new Error("my message"))
 
if (Cause.isFailType(cause)) {
console.log(cause.error.message) // Output: my message
}
ts
import { Cause } from "effect"
 
const cause = Cause.fail(new Error("my message"))
 
if (Cause.isFailType(cause)) {
console.log(cause.error.message) // Output: my message
}

By using these guards, you can effectively determine the nature of a Cause, enabling you to handle different error scenarios appropriately in your code. Whether it's an empty cause, failure, defect, interruption, sequential composition, or parallel composition, these guards provide a clear way to identify and manage various types of errors.

Pattern Matching

In addition to using guards, you can handle different cases of a Cause using the Cause.match function. This function allows you to define separate callbacks for each possible case of a Cause.

Let's explore how you can use Cause.match:

ts
import { Cause } from "effect"
 
const cause = Cause.parallel(
Cause.fail(new Error("my fail message")),
Cause.die("my die message")
)
 
console.log(
Cause.match(cause, {
onEmpty: "(empty)",
onFail: (error) => `(error: ${error.message})`,
onDie: (defect) => `(defect: ${defect})`,
onInterrupt: (fiberId) => `(fiberId: ${fiberId})`,
onSequential: (left, right) =>
`(onSequential (left: ${left}) (right: ${right}))`,
onParallel: (left, right) =>
`(onParallel (left: ${left}) (right: ${right})`
})
)
/*
Output:
(onParallel (left: (error: my fail message)) (right: (defect: my die message))
*/
ts
import { Cause } from "effect"
 
const cause = Cause.parallel(
Cause.fail(new Error("my fail message")),
Cause.die("my die message")
)
 
console.log(
Cause.match(cause, {
onEmpty: "(empty)",
onFail: (error) => `(error: ${error.message})`,
onDie: (defect) => `(defect: ${defect})`,
onInterrupt: (fiberId) => `(fiberId: ${fiberId})`,
onSequential: (left, right) =>
`(onSequential (left: ${left}) (right: ${right}))`,
onParallel: (left, right) =>
`(onParallel (left: ${left}) (right: ${right})`
})
)
/*
Output:
(onParallel (left: (error: my fail message)) (right: (defect: my die message))
*/

Pretty Printing

When working with errors in your code, it's crucial to have clear and readable error messages for effective debugging. The Cause.pretty function provides a convenient way to achieve this by generating nicely formatted error messages as strings.

Let's take a look at how you can use Cause.pretty:

ts
import { Cause, FiberId } from "effect"
 
console.log(Cause.pretty(Cause.empty)) // All fibers interrupted without errors.
console.log(Cause.pretty(Cause.fail(new Error("my fail message")))) // Error: my fail message
console.log(Cause.pretty(Cause.die("my die message"))) // Error: my die message
console.log(Cause.pretty(Cause.interrupt(FiberId.make(1, 0)))) // All fibers interrupted without errors.
console.log(
Cause.pretty(Cause.sequential(Cause.fail("fail1"), Cause.fail("fail2")))
)
/*
Output:
Error: fail1
Error: fail2
*/
ts
import { Cause, FiberId } from "effect"
 
console.log(Cause.pretty(Cause.empty)) // All fibers interrupted without errors.
console.log(Cause.pretty(Cause.fail(new Error("my fail message")))) // Error: my fail message
console.log(Cause.pretty(Cause.die("my die message"))) // Error: my die message
console.log(Cause.pretty(Cause.interrupt(FiberId.make(1, 0)))) // All fibers interrupted without errors.
console.log(
Cause.pretty(Cause.sequential(Cause.fail("fail1"), Cause.fail("fail2")))
)
/*
Output:
Error: fail1
Error: fail2
*/

Retrieval of Failures and Defects

If you're specifically interested in obtaining a collection of failures or defects from a Cause, you can use the Cause.failures and Cause.defects functions respectively.

Let's see how you can utilize these functions:

ts
import { Cause } from "effect"
 
const cause = Cause.parallel(
Cause.fail(new Error("my fail message 1")),
Cause.sequential(
Cause.die("my die message"),
Cause.fail(new Error("my fail message 2"))
)
)
 
console.log(Cause.failures(cause))
/*
Output:
{
_id: 'Chunk',
values: [
Error: my fail message 1 ...,
Error: my fail message 2 ...
]
}
*/
 
console.log(Cause.defects(cause))
/*
Output:
{ _id: 'Chunk', values: [ 'my die message' ] }
*/
ts
import { Cause } from "effect"
 
const cause = Cause.parallel(
Cause.fail(new Error("my fail message 1")),
Cause.sequential(
Cause.die("my die message"),
Cause.fail(new Error("my fail message 2"))
)
)
 
console.log(Cause.failures(cause))
/*
Output:
{
_id: 'Chunk',
values: [
Error: my fail message 1 ...,
Error: my fail message 2 ...
]
}
*/
 
console.log(Cause.defects(cause))
/*
Output:
{ _id: 'Chunk', values: [ 'my die message' ] }
*/