terraform/terraform/node_output.go

package terraform

import (
	"fmt"
	"log"

	"github.com/hashicorp/hcl/v2"
	"github.com/hashicorp/terraform/addrs"
	"github.com/hashicorp/terraform/configs"
	"github.com/hashicorp/terraform/dag"
	"github.com/hashicorp/terraform/lang"
	"github.com/hashicorp/terraform/plans"
	"github.com/hashicorp/terraform/states"
	"github.com/hashicorp/terraform/tfdiags"
	"github.com/zclconf/go-cty/cty"
)

// nodeExpandOutput is the placeholder for a non-root module output that has
// not yet had its module path expanded.
type nodeExpandOutput struct {
	Addr    addrs.OutputValue
	Module  addrs.Module
	Config  *configs.Output
	Changes []*plans.OutputChangeSrc
	Destroy bool
}

var (
	_ GraphNodeReferenceable     = (*nodeExpandOutput)(nil)
	_ GraphNodeReferencer        = (*nodeExpandOutput)(nil)
	_ GraphNodeReferenceOutside  = (*nodeExpandOutput)(nil)
	_ GraphNodeDynamicExpandable = (*nodeExpandOutput)(nil)
	_ graphNodeTemporaryValue    = (*nodeExpandOutput)(nil)
	_ graphNodeExpandsInstances  = (*nodeExpandOutput)(nil)
)

func (n *nodeExpandOutput) expandsInstances() {}

func (n *nodeExpandOutput) temporaryValue() bool {
	// non root outputs are temporary
	return !n.Module.IsRoot()
}

func (n *nodeExpandOutput) DynamicExpand(ctx EvalContext) (*Graph, error) {
	if n.Destroy {
		// if we're planning a destroy, we only need to handle the root outputs.
		// The destroy plan doesn't evaluate any other config, so we can skip
		// the rest of the outputs.
		return n.planDestroyRootOutput(ctx)
	}

	expander := ctx.InstanceExpander()

	var g Graph
	for _, module := range expander.ExpandModule(n.Module) {
		absAddr := n.Addr.Absolute(module)

		// Find any recorded change for this output
		var change *plans.OutputChangeSrc
		for _, c := range n.Changes {
			if c.Addr.String() == absAddr.String() {
				change = c
				break
			}
		}

		o := &NodeApplyableOutput{
			Addr:   absAddr,
			Config: n.Config,
			Change: change,
		}
		log.Printf("[TRACE] Expanding output: adding %s as %T", o.Addr.String(), o)
		g.Add(o)
	}
	return &g, nil
}

// if we're planing a destroy operation, add a destroy node for any root output
func (n *nodeExpandOutput) planDestroyRootOutput(ctx EvalContext) (*Graph, error) {
	if !n.Module.IsRoot() {
		return nil, nil
	}
	state := ctx.State()
	if state == nil {
		return nil, nil
	}

	var g Graph
	o := &NodeDestroyableOutput{
		Addr:   n.Addr.Absolute(addrs.RootModuleInstance),
		Config: n.Config,
	}
	log.Printf("[TRACE] Expanding output: adding %s as %T", o.Addr.String(), o)
	g.Add(o)

	return &g, nil
}

func (n *nodeExpandOutput) Name() string {
	path := n.Module.String()
	addr := n.Addr.String() + " (expand)"
	if path != "" {
		return path + "." + addr
	}
	return addr
}

// GraphNodeModulePath
func (n *nodeExpandOutput) ModulePath() addrs.Module {
	return n.Module
}

// GraphNodeReferenceable
func (n *nodeExpandOutput) ReferenceableAddrs() []addrs.Referenceable {
	// An output in the root module can't be referenced at all.
	if n.Module.IsRoot() {
		return nil
	}

	// the output is referenced through the module call, and via the
	// module itself.
	_, call := n.Module.Call()
	callOutput := addrs.ModuleCallOutput{
		Call: call,
		Name: n.Addr.Name,
	}

	// Otherwise, we can reference the output via the
	// module call itself
	return []addrs.Referenceable{call, callOutput}
}

// GraphNodeReferenceOutside implementation
func (n *nodeExpandOutput) ReferenceOutside() (selfPath, referencePath addrs.Module) {
	// Output values have their expressions resolved in the context of the
	// module where they are defined.
	referencePath = n.Module

	// ...but they are referenced in the context of their calling module.
	selfPath = referencePath.Parent()

	return // uses named return values
}

// GraphNodeReferencer
func (n *nodeExpandOutput) References() []*addrs.Reference {
	// root outputs might be destroyable, and may not reference anything in
	// that case
	return referencesForOutput(n.Config)
}

// NodeApplyableOutput represents an output that is "applyable":
// it is ready to be applied.
type NodeApplyableOutput struct {
	Addr   addrs.AbsOutputValue
	Config *configs.Output // Config is the output in the config
	// If this is being evaluated during apply, we may have a change recorded already
	Change *plans.OutputChangeSrc
}

var (
	_ GraphNodeModuleInstance   = (*NodeApplyableOutput)(nil)
	_ GraphNodeReferenceable    = (*NodeApplyableOutput)(nil)
	_ GraphNodeReferencer       = (*NodeApplyableOutput)(nil)
	_ GraphNodeReferenceOutside = (*NodeApplyableOutput)(nil)
	_ GraphNodeExecutable       = (*NodeApplyableOutput)(nil)
	_ graphNodeTemporaryValue   = (*NodeApplyableOutput)(nil)
	_ dag.GraphNodeDotter       = (*NodeApplyableOutput)(nil)
)

func (n *NodeApplyableOutput) temporaryValue() bool {
	// this must always be evaluated if it is a root module output
	return !n.Addr.Module.IsRoot()
}

func (n *NodeApplyableOutput) Name() string {
	return n.Addr.String()
}

// GraphNodeModuleInstance
func (n *NodeApplyableOutput) Path() addrs.ModuleInstance {
	return n.Addr.Module
}

// GraphNodeModulePath
func (n *NodeApplyableOutput) ModulePath() addrs.Module {
	return n.Addr.Module.Module()
}

func referenceOutsideForOutput(addr addrs.AbsOutputValue) (selfPath, referencePath addrs.Module) {
	// Output values have their expressions resolved in the context of the
	// module where they are defined.
	referencePath = addr.Module.Module()

	// ...but they are referenced in the context of their calling module.
	selfPath = addr.Module.Parent().Module()

	return // uses named return values
}

// GraphNodeReferenceOutside implementation
func (n *NodeApplyableOutput) ReferenceOutside() (selfPath, referencePath addrs.Module) {
	return referenceOutsideForOutput(n.Addr)
}

func referenceableAddrsForOutput(addr addrs.AbsOutputValue) []addrs.Referenceable {
	// An output in the root module can't be referenced at all.
	if addr.Module.IsRoot() {
		return nil
	}

	// Otherwise, we can be referenced via a reference to our output name
	// on the parent module's call, or via a reference to the entire call.
	// e.g. module.foo.bar or just module.foo .
	// Note that our ReferenceOutside method causes these addresses to be
	// relative to the calling module, not the module where the output
	// was declared.
	_, outp := addr.ModuleCallOutput()
	_, call := addr.Module.CallInstance()

	return []addrs.Referenceable{outp, call}
}

// GraphNodeReferenceable
func (n *NodeApplyableOutput) ReferenceableAddrs() []addrs.Referenceable {
	return referenceableAddrsForOutput(n.Addr)
}

func referencesForOutput(c *configs.Output) []*addrs.Reference {
	impRefs, _ := lang.ReferencesInExpr(c.Expr)
	expRefs, _ := lang.References(c.DependsOn)
	l := len(impRefs) + len(expRefs)
	if l == 0 {
		return nil
	}
	refs := make([]*addrs.Reference, 0, l)
	refs = append(refs, impRefs...)
	refs = append(refs, expRefs...)
	return refs

}

// GraphNodeReferencer
func (n *NodeApplyableOutput) References() []*addrs.Reference {
	return referencesForOutput(n.Config)
}

// GraphNodeExecutable
func (n *NodeApplyableOutput) Execute(ctx EvalContext, op walkOperation) (diags tfdiags.Diagnostics) {
	state := ctx.State()
	if state == nil {
		return
	}

	changes := ctx.Changes() // may be nil, if we're not working on a changeset

	val := cty.UnknownVal(cty.DynamicPseudoType)
	changeRecorded := n.Change != nil
	// we we have a change recorded, we don't need to re-evaluate if the value
	// was known
	if changeRecorded {
		change, err := n.Change.Decode()
		diags = diags.Append(err)
		if err == nil {
			val = change.After
		}
	}

	// If there was no change recorded, or the recorded change was not wholly
	// known, then we need to re-evaluate the output
	if !changeRecorded || !val.IsWhollyKnown() {
		// This has to run before we have a state lock, since evaluation also
		// reads the state
		val, diags = ctx.EvaluateExpr(n.Config.Expr, cty.DynamicPseudoType, nil)
		// We'll handle errors below, after we have loaded the module.
		// Outputs don't have a separate mode for validation, so validate
		// depends_on expressions here too
		diags = diags.Append(validateDependsOn(ctx, n.Config.DependsOn))

		// For root module outputs in particular, an output value must be
		// statically declared as sensitive in order to dynamically return
		// a sensitive result, to help avoid accidental exposure in the state
		// of a sensitive value that the user doesn't want to include there.
		_, marks := val.UnmarkDeep()
		_, hasSensitive := marks["sensitive"]
		if n.Addr.Module.IsRoot() {
			if !n.Config.Sensitive && hasSensitive {
				diags = diags.Append(&hcl.Diagnostic{
					Severity: hcl.DiagError,
					Summary:  "Output refers to sensitive values",
					Detail: `To reduce the risk of accidentally exporting sensitive data that was intended to be only internal, Terraform requires that any root module output containing sensitive data be explicitly marked as sensitive, to confirm your intent.

If you do intend to export this data, annotate the output value as sensitive by adding the following argument:
    sensitive = true`,
					Subject: n.Config.DeclRange.Ptr(),
				})
			}
		}
	}

	// handling the interpolation error
	if diags.HasErrors() {
		if flagWarnOutputErrors {
			log.Printf("[ERROR] Output interpolation %q failed: %s", n.Addr, diags.Err())
			// if we're continuing, make sure the output is included, and
			// marked as unknown. If the evaluator was able to find a type
			// for the value in spite of the error then we'll use it.
			n.setValue(state, changes, cty.UnknownVal(val.Type()))

			// Keep existing warnings, while converting errors to warnings.
			// This is not meant to be the normal path, so there no need to
			// make the errors pretty.
			var warnings tfdiags.Diagnostics
			for _, d := range diags {
				switch d.Severity() {
				case tfdiags.Warning:
					warnings = warnings.Append(d)
				case tfdiags.Error:
					desc := d.Description()
					warnings = warnings.Append(tfdiags.SimpleWarning(fmt.Sprintf("%s:%s", desc.Summary, desc.Detail)))
				}
			}

			return warnings
		}
		return diags
	}
	n.setValue(state, changes, val)

	// If we were able to evaluate a new value, we can update that in the
	// refreshed state as well.
	if state = ctx.RefreshState(); state != nil && val.IsWhollyKnown() {
		n.setValue(state, changes, val)
	}

	return diags
}

// dag.GraphNodeDotter impl.
func (n *NodeApplyableOutput) DotNode(name string, opts *dag.DotOpts) *dag.DotNode {
	return &dag.DotNode{
		Name: name,
		Attrs: map[string]string{
			"label": n.Name(),
			"shape": "note",
		},
	}
}

// NodeDestroyableOutput represents an output that is "destroyable":
// its application will remove the output from the state.
type NodeDestroyableOutput struct {
	Addr   addrs.AbsOutputValue
	Config *configs.Output // Config is the output in the config
}

var (
	_ GraphNodeExecutable = (*NodeDestroyableOutput)(nil)
	_ dag.GraphNodeDotter = (*NodeDestroyableOutput)(nil)
)

func (n *NodeDestroyableOutput) Name() string {
	return fmt.Sprintf("%s (destroy)", n.Addr.String())
}

// GraphNodeModulePath
func (n *NodeDestroyableOutput) ModulePath() addrs.Module {
	return n.Addr.Module.Module()
}

func (n *NodeDestroyableOutput) temporaryValue() bool {
	// this must always be evaluated if it is a root module output
	return !n.Addr.Module.IsRoot()
}

// GraphNodeExecutable
func (n *NodeDestroyableOutput) Execute(ctx EvalContext, op walkOperation) tfdiags.Diagnostics {
	state := ctx.State()
	if state == nil {
		return nil
	}

	// if this is a root module, try to get a before value from the state for
	// the diff
	sensitiveBefore := false
	before := cty.NullVal(cty.DynamicPseudoType)
	mod := state.Module(n.Addr.Module)
	if n.Addr.Module.IsRoot() && mod != nil {
		for name, o := range mod.OutputValues {
			if name == n.Addr.OutputValue.Name {
				sensitiveBefore = o.Sensitive
				before = o.Value
				break
			}
		}
	}

	changes := ctx.Changes()
	if changes != nil {
		change := &plans.OutputChange{
			Addr:      n.Addr,
			Sensitive: sensitiveBefore,
			Change: plans.Change{
				Action: plans.Delete,
				Before: before,
				After:  cty.NullVal(cty.DynamicPseudoType),
			},
		}

		cs, err := change.Encode()
		if err != nil {
			// Should never happen, since we just constructed this right above
			panic(fmt.Sprintf("planned change for %s could not be encoded: %s", n.Addr, err))
		}
		log.Printf("[TRACE] NodeDestroyableOutput: Saving %s change for %s in changeset", change.Action, n.Addr)
		changes.RemoveOutputChange(n.Addr) // remove any existing planned change, if present
		changes.AppendOutputChange(cs)     // add the new planned change
	}

	state.RemoveOutputValue(n.Addr)
	return nil
}

// dag.GraphNodeDotter impl.
func (n *NodeDestroyableOutput) DotNode(name string, opts *dag.DotOpts) *dag.DotNode {
	return &dag.DotNode{
		Name: name,
		Attrs: map[string]string{
			"label": n.Name(),
			"shape": "note",
		},
	}
}

func (n *NodeApplyableOutput) setValue(state *states.SyncState, changes *plans.ChangesSync, val cty.Value) {
	// If we have an active changeset then we'll first replicate the value in
	// there and lookup the prior value in the state. This is used in
	// preference to the state where present, since it *is* able to represent
	// unknowns, while the state cannot.
	if changes != nil {
		// if this is a root module, try to get a before value from the state for
		// the diff
		sensitiveBefore := false
		before := cty.NullVal(cty.DynamicPseudoType)

		// is this output new to our state?
		newOutput := true

		mod := state.Module(n.Addr.Module)
		if n.Addr.Module.IsRoot() && mod != nil {
			for name, o := range mod.OutputValues {
				if name == n.Addr.OutputValue.Name {
					before = o.Value
					sensitiveBefore = o.Sensitive
					newOutput = false
					break
				}
			}
		}

		// We will not show the value is either the before or after are marked
		// as sensitivity. We can show the value again once sensitivity is
		// removed from both the config and the state.
		sensitiveChange := sensitiveBefore || n.Config.Sensitive

		// strip any marks here just to be sure we don't panic on the True comparison
		unmarkedVal, _ := val.UnmarkDeep()

		action := plans.Update
		switch {
		case val.IsNull() && before.IsNull():
			// This is separate from the NoOp case below, since we can ignore
			// sensitivity here when there are only null values.
			action = plans.NoOp

		case newOutput:
			// This output was just added to the configuration
			action = plans.Create

		case val.IsWhollyKnown() &&
			unmarkedVal.Equals(before).True() &&
			n.Config.Sensitive == sensitiveBefore:
			// Sensitivity must also match to be a NoOp.
			// Theoretically marks may not match here, but sensitivity is the
			// only one we can act on, and the state will have been loaded
			// without any marks to consider.
			action = plans.NoOp
		}

		change := &plans.OutputChange{
			Addr:      n.Addr,
			Sensitive: sensitiveChange,
			Change: plans.Change{
				Action: action,
				Before: before,
				After:  val,
			},
		}

		cs, err := change.Encode()
		if err != nil {
			// Should never happen, since we just constructed this right above
			panic(fmt.Sprintf("planned change for %s could not be encoded: %s", n.Addr, err))
		}
		log.Printf("[TRACE] setValue: Saving %s change for %s in changeset", change.Action, n.Addr)
		changes.RemoveOutputChange(n.Addr) // remove any existing planned change, if present
		changes.AppendOutputChange(cs)     // add the new planned change
	}

	if val.IsKnown() && !val.IsNull() {
		// The state itself doesn't represent unknown values, so we null them
		// out here and then we'll save the real unknown value in the planned
		// changeset below, if we have one on this graph walk.
		log.Printf("[TRACE] setValue: Saving value for %s in state", n.Addr)
		unmarkedVal, _ := val.UnmarkDeep()
		stateVal := cty.UnknownAsNull(unmarkedVal)
		state.SetOutputValue(n.Addr, stateVal, n.Config.Sensitive)
	} else {
		log.Printf("[TRACE] setValue: Removing %s from state (it is now null)", n.Addr)
		state.RemoveOutputValue(n.Addr)
	}

}