This Commented CUE demonstrates how to use the built-in functions encoding/yaml.Validate and encoding/yaml.ValidatePartial as field validators.

They assert that properly-formed YAML, encoded in a string, adheres to specific constraints by checking that the data and schema unify successfully. encoding.yaml/Validate also requires that all non-optional fields are present.

file.cue
package example

import "encoding/yaml"

data: """
	a: 1
	b: "two"
	"""

// Validate requires that all non-optional fields in a schema are present in
// data, and that data adheres to schema constraints.
data: yaml.Validate(_outOfBoundsSchema)
data: yaml.Validate(_missingFieldSchema)

// ValidatePartial requires only that data adheres to schema constraints.
data: yaml.ValidatePartial(_outOfBoundsSchema)
data: yaml.ValidatePartial(_missingFieldSchema)

_outOfBoundsSchema: {
	a!: >99 // validation failure for both functions
	b!: string
}
_missingFieldSchema: {
	a!: int
	b!: string
	c!: bool // validation failure for yaml.Validate only
}
TERMINAL
$ cue vet
data: invalid value "a: 1\nb: \"two\"" (does not satisfy encoding/yaml.Validate): error in call to encoding/yaml.Validate: invalid value 1 (out of bound >99):
    ./file.cue:12:7
    ./file.cue:5:7
    ./file.cue:13:7
    ./file.cue:16:7
    ./file.cue:17:7
    ./file.cue:20:6
    yaml.Validate:1:4
data: invalid value "a: 1\nb: \"two\"" (does not satisfy encoding/yaml.Validate): error in call to encoding/yaml.Validate: field is required but not present:
    ./file.cue:13:7
    ./file.cue:5:7
    ./file.cue:12:7
    ./file.cue:16:7
    ./file.cue:17:7
    ./file.cue:26:2
data: invalid value "a: 1\nb: \"two\"" (does not satisfy encoding/yaml.ValidatePartial): error in call to encoding/yaml.ValidatePartial: invalid value 1 (out of bound >99):
    ./file.cue:16:7
    ./file.cue:5:7
    ./file.cue:12:7
    ./file.cue:13:7
    ./file.cue:17:7
    ./file.cue:20:6
    yaml.ValidatePartial:1:4