Files
732b49d236 Feat: defkit api completeness (#7064)
* docs: map existing codebase

Signed-off-by: Vishal Kumar <vishal210893@gmail.com>

* docs: initialize project

Signed-off-by: Vishal Kumar <vishal210893@gmail.com>

* Fix: add Short() and Ignore() methods to FloatParam for API completeness

FloatParam was the only param type missing Short(string) and Ignore()
fluent methods; the underlying baseParam fields already existed. Adds
matching test cases following the existing BoolParam/IntParam pattern.

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude <noreply@anthropic.com>
Signed-off-by: Vishal Kumar <vishal210893@gmail.com>

* feat(defkit): add StatusDetails() to all 4 definition types via baseDefinition

Adds statusDetails string field to baseDefinition with setStatusDetails()
setter and GetStatusDetails() getter. Exposes StatusDetails(string) fluent
method on ComponentDefinition, TraitDefinition, WorkflowStepDefinition, and
PolicyDefinition. Updates writeStatus in cuegen.go and the inline status
render block in trait.go to render statusDetails as a #"""..."""# CUE block
alongside customStatus and healthPolicy when set.

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude <noreply@anthropic.com>
Signed-off-by: Vishal Kumar <vishal210893@gmail.com>

* feat(01-02): add ForceOptional() to IntParam, FloatParam, EnumParam

- IntParam.ForceOptional() *IntParam
- FloatParam.ForceOptional() *FloatParam
- EnumParam.ForceOptional() *EnumParam
- Test cases for all three types following BoolParam/StringParam patterns

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Signed-off-by: Vishal Kumar <vishal210893@gmail.com>

* feat(02-01): add Labels()/GetLabels() to PolicyDefinition and fix hardcoded labels CUE

- Add labels map[string]string field to PolicyDefinition struct
- Add Labels() fluent setter and GetLabels() getter
- Replace hardcoded labels: {} in PolicyCUEGenerator with nil-conditional sorted output
- Add sort import to policy.go

Signed-off-by: Vishal Kumar <vishal210893@gmail.com>

* test(02-01): add Ginkgo tests for PolicyDefinition.Labels

- Test store/return labels
- Test sorted key output in CUE
- Test labels block omitted when Labels() never called
- Test empty labels block when Labels called with empty map

Signed-off-by: Vishal Kumar <vishal210893@gmail.com>

* docs(02-01): complete Labels on PolicyDefinition plan

- SUMMARY.md: plan 02-01 execution results
- STATE.md: phase 2 in progress, B1 satisfied, key decisions recorded
- ROADMAP.md: phase 2 progress updated (1/3 plans)
- REQUIREMENTS.md: B1 marked complete

Signed-off-by: Vishal Kumar <vishal210893@gmail.com>

* feat(02-02): add annotations field to baseDefinition with getter/setter

Signed-off-by: Vishal Kumar <vishal210893@gmail.com>

* feat(02-03): wire status block rendering into WorkflowStep and Policy CUE generators

- Add status: { customStatus, healthPolicy, statusDetails } rendering to WorkflowStepCUEGenerator.GenerateTemplate
- Add same status block rendering to PolicyCUEGenerator.GenerateTemplate
- Block only emitted when at least one status field is non-empty, matching trait.go pattern

Signed-off-by: Vishal Kumar <vishal210893@gmail.com>

* feat(02-02): add Annotations() fluent method to all 4 definition types

Signed-off-by: Vishal Kumar <vishal210893@gmail.com>

* feat(02-02): conditional sorted annotations CUE block for component, trait, policy

Signed-off-by: Vishal Kumar <vishal210893@gmail.com>

* test(02-03): add Ginkgo status block CUE render tests for WorkflowStep and Policy

- 5 new tests in WorkflowStepDefinition/Status Block CUE Render context
- 5 new tests in PolicyDefinition/Status Block CUE Render context
- Rule 3 fix: restore sort import in trait.go (used by annotations render, spuriously flagged)

Signed-off-by: Vishal Kumar <vishal210893@gmail.com>

* feat(02-02): merge user annotations before category in workflow step CUE block

Signed-off-by: Vishal Kumar <vishal210893@gmail.com>

* feat(02-02): add sorted annotations CUE block to WorkflowStepCUEGenerator

- Wire sorted user annotations rendering into WorkflowStep.GenerateFullDefinition
- Matches pattern already added to Component, Trait, and Policy generators

Signed-off-by: Vishal Kumar <vishal210893@gmail.com>

* feat(02-02): merge user annotations into metadata.annotations in ToYAML for all 4 types

Signed-off-by: Vishal Kumar <vishal210893@gmail.com>

* docs(02-03): complete Wire Status Block Rendering plan

- Add 02-03-SUMMARY.md with full execution record
- Update STATE.md: phase 2 complete, new decision recorded
- Update ROADMAP.md progress for phase 2
- Mark B4 complete in REQUIREMENTS.md

Signed-off-by: Vishal Kumar <vishal210893@gmail.com>

* test(02-02): add Annotations Ginkgo tests for all 4 definition types

Signed-off-by: Vishal Kumar <vishal210893@gmail.com>

* docs(02-02): complete Annotations plan - SUMMARY, STATE, ROADMAP updated

Signed-off-by: Vishal Kumar <vishal210893@gmail.com>

* docs(03): create phase 3 plans for missing CRD spec fields

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Signed-off-by: Vishal Kumar <vishal210893@gmail.com>

* feat(03-01): add Version() fluent method and GetVersion() to all 4 definition types

- version string field added to baseDefinition struct
- setVersion/GetVersion methods on baseDefinition following existing pattern
- Version(string) fluent setter on TraitDefinition, ComponentDefinition, PolicyDefinition, WorkflowStepDefinition
- TDD RED+GREEN: tests for round-trip pass; CUE/YAML render tests added (fail, to be fixed in next task)

Signed-off-by: Vishal Kumar <vishal210893@gmail.com>

* feat(03-02): add ManageWorkload/ControlPlaneOnly/RevisionEnabled to TraitDefinition

- Add manageWorkload, controlPlaneOnly, revisionEnabled bool fields to TraitDefinition struct
- Add ManageWorkload(), ControlPlaneOnly(), RevisionEnabled() fluent setters
- Add IsManageWorkload(), IsControlPlaneOnly(), IsRevisionEnabled() getters
- Emit conditionally in ToYAML only when true; ToCue() unaffected
- Add 11 round-trip tests covering defaults, setters, and YAML emission

Signed-off-by: Vishal Kumar <vishal210893@gmail.com>

* test(03-03): add Ginkgo tests for ChildResourceKind accumulator on ComponentDefinition

- 7 specs covering nil default, single entry, multi-entry accumulation,
  selector preservation, ToYAML emit/omit, and chaining

Signed-off-by: Vishal Kumar <vishal210893@gmail.com>

* feat(03-02): add ManageHealthCheck to PolicyDefinition

- Add manageHealthCheck bool field to PolicyDefinition struct
- Add ManageHealthCheck() fluent setter and IsManageHealthCheck() getter
- Emit conditionally in ToYAML only when true; ToCue() unaffected
- Add 5 round-trip tests covering default, setter, YAML emission, and chaining

Signed-off-by: Vishal Kumar <vishal210893@gmail.com>

* test(03-03): add Ginkgo tests for PodSpecPath on ComponentDefinition

- 5 specs covering empty default, set/get round-trip, ToYAML emit/omit,
  and chaining; both fields satisfy C5 (childResourceKinds) and C6 (podSpecPath)

Signed-off-by: Vishal Kumar <vishal210893@gmail.com>

* feat(03-01): render version in CUE output and spec.version in ToYAML for all 4 definition types

- cuegen.go: conditional version emit in ComponentDefinition GenerateFullDefinition (after description)
- trait.go: conditional version emit in TraitCUEGenerator.GenerateFullDefinition; spec.version in ToYAML
- policy.go: conditional version emit in PolicyCUEGenerator.GenerateFullDefinition; spec.version in ToYAML
- workflow_step.go: conditional version emit in WorkflowStepCUEGenerator.GenerateFullDefinition; spec.version in ToYAML
- version omitted entirely when not set; TDD GREEN phase complete

Signed-off-by: Vishal Kumar <vishal210893@gmail.com>

* docs(03-03): complete ChildResourceKind+PodSpecPath plan - SUMMARY, STATE, ROADMAP updated

- C5 (childResourceKinds accumulator) and C6 (podSpecPath) requirements satisfied
- 12 Ginkgo specs added covering both fields

Signed-off-by: Vishal Kumar <vishal210893@gmail.com>

* docs(03-02): complete boolean CRD spec fields plan — manageWorkload/controlPlaneOnly/revisionEnabled/manageHealthCheck

Signed-off-by: Vishal Kumar <vishal210893@gmail.com>

* docs(03-01): complete Version() plan — SUMMARY, STATE, ROADMAP updated

- 03-01-SUMMARY.md created documenting Version() on all 4 definition types
- STATE.md: 03-01 session log entry added; phase 3 marked complete; decision recorded
- ROADMAP.md: phase 3 updated to 3/3 plans executed; status Complete
- REQUIREMENTS.md: C1 marked complete

Signed-off-by: Vishal Kumar <vishal210893@gmail.com>

* docs(04): add gap closure plan 04-03 for vela-go-definitions call sites

Closes 17 ArrayOf→Of and 1 FilterPred→Filter call sites broken by the phase 04 renames.

Co-Authored-By: Claude <noreply@anthropic.com>
Signed-off-by: Vishal Kumar <vishal210893@gmail.com>

* feat(phase-4): low-risk renames — SetField→Set, ArrayOf→Of, FilterPred→Filter

A1: PolicyTemplate.SetField() renamed to Set()
A3: StructField.ArrayOf() renamed to Of()
A5: HelperBuilder.FilterPred(Predicate) renamed to Filter(Predicate);
    HelperBuilder.Filter(Condition) renamed to FilterCond(Condition)

All callers within defkit updated. Verification: 894 Ginkgo specs pass,
go build exits 0, zero occurrences of old names in pkg/definition/defkit/.

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude <noreply@anthropic.com>
Signed-off-by: Vishal Kumar <vishal210893@gmail.com>

* docs(05-high-impact-renames): create phase plan

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Signed-off-by: Vishal Kumar <vishal210893@gmail.com>

* feat(phase-5): high-impact renames — Values→Enum, Fields→WithFields

A4: EnumParam.Values() renamed to Enum() — aligns with StringParam.Enum()
    and StructField.Enum(); 10 call sites updated in defkit tests

A2: StructParam.Fields() and OneOfVariant.Fields() renamed to WithFields()
    — aligns with ArrayParam and MapParam; ~50 call sites updated in defkit tests

Both repos build clean. go test ./pkg/definition/defkit/... passes.
Non-target .Fields() methods (InCondition, StructBuilder, ArrayElement etc.)
correctly preserved.

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude <noreply@anthropic.com>
Signed-off-by: Vishal Kumar <vishal210893@gmail.com>

* chore: exclude .planning/ and .claude/ from version control

Local development artifacts only — not for upstream.

Signed-off-by: Vishal Kumar <vishal210893@gmail.com>

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude <noreply@anthropic.com>
Signed-off-by: Vishal Kumar <vishal210893@gmail.com>

* refactor: adjust formatting for consistency in component and policy definitions

Signed-off-by: Vishal Kumar <vishal210893@gmail.com>

* feat: add FilterCond method for filtering items by Condition expression

Signed-off-by: Vishal Kumar <vishal210893@gmail.com>

* feat: enhance filter condition handling with AND-composition for multiple filters

Signed-off-by: Vishal Kumar <vishal210893@gmail.com>

* feat: rename Enum method to Values for consistency in parameter definitions

Signed-off-by: Jerrin Francis <jfo@>

* feat: simplify labels handling in policy and trait definitions

Signed-off-by: Jerrin Francis <jerrinfrancis7@gmail.com>

---------

Signed-off-by: Vishal Kumar <vishal210893@gmail.com>
Signed-off-by: Jerrin Francis <jfo@>
Signed-off-by: Jerrin Francis <jerrinfrancis7@gmail.com>
Co-authored-by: Claude <noreply@anthropic.com>
2026-03-10 11:28:23 +00:00

891 lines
26 KiB
Go

/*
Copyright 2025 The KubeVela Authors.
Licensed under the Apache License, Version 2.0 (the "License");
you may not use this file except in compliance with the License.
You may obtain a copy of the License at
http://www.apache.org/licenses/LICENSE-2.0
Unless required by applicable law or agreed to in writing, software
distributed under the License is distributed on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
See the License for the specific language governing permissions and
limitations under the License.
*/
package defkit
import (
"fmt"
"iter"
"maps"
"slices"
"strings"
)
// CollectionOp represents an operation on a collection (array/list).
type CollectionOp struct {
source Value
ops []collectionOperation
guard Condition // optional guard: wraps entire comprehension in if cond
}
type collectionOperation interface {
apply(items []any) []any
}
// Each creates a collection operation pipeline from a list parameter.
// Usage: ports.Each().Filter(...).Map(...).Pick(...)
func Each(source Value) *CollectionOp {
return &CollectionOp{
source: source,
ops: make([]collectionOperation, 0),
}
}
// From creates a collection operation pipeline from a source value.
// This is an alias for Each, providing a more readable API for transformations.
// Usage: From(volumes).Map(FieldMap{"mountPath": F("path"), "name": F("name")})
func From(source Value) *CollectionOp {
return Each(source)
}
// F creates a field reference for use in Map operations.
// This is a convenience function that creates a FieldRef.
// Usage: From(volumes).Map(FieldMap{"mountPath": F("path")})
func F(name string) FieldRef {
return FieldRef(name)
}
func (c *CollectionOp) expr() {}
func (c *CollectionOp) value() {}
// Guard adds a guard condition that wraps the entire comprehension.
// When the guard is set, the generated CUE is:
//
// [if guard for v in source if filter {v}]
//
// This is used when the source may not exist and needs a
// protective condition (e.g., if parameter.privileges != _|_).
func (c *CollectionOp) Guard(cond Condition) *CollectionOp {
c.guard = cond
return c
}
// GetGuard returns the guard condition, if any.
func (c *CollectionOp) GetGuard() Condition { return c.guard }
// Filter keeps only items matching the predicate.
// Usage: .Filter(Field("expose").Eq(true))
func (c *CollectionOp) Filter(pred Predicate) *CollectionOp {
c.ops = append(c.ops, &filterOp{pred: pred})
return c
}
// FilterCond keeps only items matching a Condition expression.
// The condition is used for CUE generation (generates an `if condition` guard);
// runtime apply is a passthrough since Condition is a CUE-level concept.
func (c *CollectionOp) FilterCond(cond Condition) *CollectionOp {
c.ops = append(c.ops, &filterCondCollectionOp{cond: cond})
return c
}
// Map transforms each item using the given field mappings.
// Usage: .Map(Fields{"containerPort": Field("port"), "name": Field("name")})
func (c *CollectionOp) Map(mappings FieldMap) *CollectionOp {
c.ops = append(c.ops, &mapOp{mappings: mappings})
return c
}
// MapVariant adds a conditional field mapping that only applies when the iteration
// variable's discriminator field equals the given variant name.
// Generates: if v.discriminator == "variantName" { ...mappings... }
// Usage: .MapVariant("type", "pvc", FieldMap{"persistentVolumeClaim.claimName": FieldRef("claimName")})
func (c *CollectionOp) MapVariant(discriminator, variantName string, mappings FieldMap) *CollectionOp {
c.ops = append(c.ops, &mapVariantOp{
discriminator: discriminator,
variantName: variantName,
mappings: mappings,
})
return c
}
// Pick selects only the specified fields from each item.
// Usage: .Pick("name", "mountPath")
func (c *CollectionOp) Pick(fields ...string) *CollectionOp {
c.ops = append(c.ops, &pickOp{fields: fields})
return c
}
// Rename renames a field in each item.
// Usage: .Rename("port", "containerPort")
func (c *CollectionOp) Rename(from, to string) *CollectionOp {
c.ops = append(c.ops, &renameOp{from: from, to: to})
return c
}
// Wrap wraps each item value under a new key.
// Usage: .Wrap("name") transforms "secret1" to {name: "secret1"}
func (c *CollectionOp) Wrap(key string) *CollectionOp {
c.ops = append(c.ops, &wrapOp{key: key})
return c
}
// DefaultField sets a default value for a field if not present.
// Usage: .DefaultField("name", Format("port-%d", Field("port")))
func (c *CollectionOp) DefaultField(field string, defaultVal FieldValue) *CollectionOp {
c.ops = append(c.ops, &defaultFieldOp{field: field, defaultVal: defaultVal})
return c
}
// Flatten flattens nested collections from multiple sources.
// Used for volumeMounts which has pvc, configMap, secret, etc.
func (c *CollectionOp) Flatten() *CollectionOp {
c.ops = append(c.ops, &flattenOp{})
return c
}
// Dedupe removes duplicate items by a key field, keeping the first occurrence.
func (c *CollectionOp) Dedupe(keyField string) *CollectionOp {
c.ops = append(c.ops, &dedupeOp{keyField: keyField})
return c
}
// Source returns the source value.
func (c *CollectionOp) Source() Value { return c.source }
// Operations returns the operations to apply.
func (c *CollectionOp) Operations() []collectionOperation { return c.ops }
// --- Go 1.23 Iterator Methods (iter.Seq) ---
// All returns an iterator over all items after applying operations.
// This is a Go 1.23 range-over-function iterator that can be used in for-range loops.
// Example: for item := range col.All() { ... }
func (c *CollectionOp) All(items []any) iter.Seq[map[string]any] {
return func(yield func(map[string]any) bool) {
// Apply all operations in sequence
result := items
for _, op := range c.ops {
result = op.apply(result)
}
// Yield each item
for _, item := range result {
if m, ok := item.(map[string]any); ok {
if !yield(m) {
return
}
}
}
}
}
// AllPairs returns a key-value iterator with index and item.
// This is a Go 1.23 iter.Seq2 that provides (index, item) pairs.
// Example: for i, item := range col.AllPairs() { ... }
func (c *CollectionOp) AllPairs(items []any) iter.Seq2[int, map[string]any] {
return func(yield func(int, map[string]any) bool) {
// Apply all operations in sequence
result := items
for _, op := range c.ops {
result = op.apply(result)
}
// Yield each item with index
for i, item := range result {
if m, ok := item.(map[string]any); ok {
if !yield(i, m) {
return
}
}
}
}
}
// Collect materializes the iterator into a slice.
// Uses Go 1.23 slices.Collect for efficient collection.
func (c *CollectionOp) Collect(items []any) []map[string]any {
return slices.Collect(c.All(items))
}
// Count returns the number of items after applying operations.
func (c *CollectionOp) Count(items []any) int {
count := 0
for range c.All(items) {
count++
}
return count
}
// First returns the first item after applying operations, or nil if empty.
func (c *CollectionOp) First(items []any) map[string]any {
for item := range c.All(items) {
return item
}
return nil
}
// FieldMap defines field mappings for Map operation.
type FieldMap map[string]FieldValue
// --- Go 1.23 Iterator Methods for FieldMap (maps package) ---
// All returns an iterator over all key-value pairs using Go 1.23 maps.All.
// Example: for key, val := range fm.All() { ... }
func (fm FieldMap) All() iter.Seq2[string, FieldValue] {
return maps.All(fm)
}
// Keys returns an iterator over all keys using Go 1.23 maps.Keys.
// Example: for key := range fm.Keys() { ... }
func (fm FieldMap) Keys() iter.Seq[string] {
return maps.Keys(fm)
}
// Values returns an iterator over all values using Go 1.23 maps.Values.
// Example: for val := range fm.Values() { ... }
func (fm FieldMap) Values() iter.Seq[FieldValue] {
return maps.Values(fm)
}
// FieldValue represents a value to use in field mapping.
type FieldValue interface {
resolve(item map[string]any) any
}
// FieldRef references a field from the current item in collection operations.
type FieldRef string
func (f FieldRef) resolve(item map[string]any) any {
return item[string(f)]
}
// Or provides a fallback if the field is nil or empty.
// Generates CUE like: *v.field | fallback
func (f FieldRef) Or(fallback FieldValue) *OrFieldRef {
return &OrFieldRef{primary: f, fallback: fallback}
}
// OrConditional provides a fallback using if/else blocks instead of default syntax.
// Generates CUE like:
//
// if v.field != _|_ { name: v.field }
// if v.field == _|_ { name: fallbackExpr }
func (f FieldRef) OrConditional(fallback FieldValue) *ConditionalOrFieldRef {
return &ConditionalOrFieldRef{primary: f, fallback: fallback}
}
// ConditionalOrFieldRef represents a field reference with a conditional fallback.
// Instead of generating CUE default syntax (*v.field | fallback), it generates
// two if/else blocks for the field.
type ConditionalOrFieldRef struct {
primary FieldRef
fallback FieldValue
}
func (c *ConditionalOrFieldRef) resolve(item map[string]any) any {
val := item[string(c.primary)]
if val == nil || val == "" {
return c.fallback.resolve(item)
}
return val
}
// OrFieldRef represents a field reference with a fallback value.
type OrFieldRef struct {
primary FieldRef
fallback FieldValue
}
func (o *OrFieldRef) resolve(item map[string]any) any {
val := item[string(o.primary)]
if val == nil || val == "" {
return o.fallback.resolve(item)
}
return val
}
// LitVal is a literal value for field mapping.
type LitVal struct {
val any
}
func (l LitVal) resolve(_ map[string]any) any {
return l.val
}
// LitField creates a literal field value.
func LitField(val any) LitVal {
return LitVal{val: val}
}
// FormatField formats a string using item fields.
// Usage: FormatField("port-%v", Field("port"))
type FormatField struct {
format string
args []FieldValue
}
func (f *FormatField) resolve(item map[string]any) any {
args := make([]any, len(f.args))
for i, arg := range f.args {
args[i] = arg.resolve(item)
}
return fmt.Sprintf(f.format, args...)
}
// Format creates a formatted string field value.
func Format(format string, args ...FieldValue) *FormatField {
return &FormatField{format: format, args: args}
}
// RequiredImports returns the CUE imports required by FormatField.
// FormatField typically uses strconv.FormatInt for numeric formatting.
func (f *FormatField) RequiredImports() []string {
// Check if any args are numeric field references that would need strconv
// The format "port-%v" with numeric args generates strconv.FormatInt
for _, arg := range f.args {
switch arg.(type) {
case FieldRef, *OrFieldRef:
// Field references with numeric formatting use strconv
if strings.Contains(f.format, "%v") || strings.Contains(f.format, "%d") {
return []string{"strconv"}
}
}
}
return nil
}
// Predicate represents a filter condition.
type Predicate interface {
matches(item map[string]any) bool
}
// FieldEq checks if a field equals a value.
type FieldEq struct {
field string
value any
}
func (f FieldEq) matches(item map[string]any) bool {
return item[f.field] == f.value
}
// FieldEquals creates a field equality predicate.
func FieldEquals(field string, value any) FieldEq {
return FieldEq{field: field, value: value}
}
// FieldIsSet checks if a field is set (not nil).
type FieldIsSet struct {
field string
}
func (f FieldIsSet) matches(item map[string]any) bool {
val, exists := item[f.field]
return exists && val != nil
}
// FieldExists creates a field-is-set predicate.
func FieldExists(field string) FieldIsSet {
return FieldIsSet{field: field}
}
// --- Internal operation implementations ---
type filterOp struct {
pred Predicate
}
func (f *filterOp) apply(items []any) []any {
result := make([]any, 0, len(items))
for _, item := range items {
if m, ok := item.(map[string]any); ok {
if f.pred.matches(m) {
result = append(result, item)
}
}
}
return result
}
// filterCondCollectionOp filters items by a Condition (CUE-level expression).
// apply is a runtime passthrough: Condition is a CUE AST concept with no Go runtime
// evaluation semantics. Filtering only takes effect in generated CUE output.
// Callers must not rely on Collect/Count/First to reflect condition-based filters.
type filterCondCollectionOp struct {
cond Condition
}
func (f *filterCondCollectionOp) apply(items []any) []any { return items }
// Cond returns the condition used for CUE generation.
func (f *filterCondCollectionOp) Cond() Condition { return f.cond }
type mapOp struct {
mappings FieldMap
}
func (m *mapOp) apply(items []any) []any {
result := make([]any, 0, len(items))
for _, item := range items {
if itemMap, ok := item.(map[string]any); ok {
newItem := make(map[string]any)
for newKey, fieldVal := range m.mappings {
resolved := fieldVal.resolve(itemMap)
// Skip nil values from optional field types — they should be
// omitted from the output entirely, not included as nil.
if resolved == nil {
if _, isOpt := fieldVal.(*OptionalField); isOpt {
continue
}
if _, isCompOpt := fieldVal.(*CompoundOptionalField); isCompOpt {
continue
}
}
newItem[newKey] = resolved
}
result = append(result, newItem)
}
}
return result
}
// mapVariantOp represents a conditional map operation based on a discriminator field value.
// When the iteration variable's discriminator field equals the variant name,
// the variant's field mappings are included.
// Generates: if v.discriminator == "variantName" { ...mappings... }
type mapVariantOp struct {
discriminator string
variantName string
mappings FieldMap
}
func (m *mapVariantOp) apply(items []any) []any {
// Runtime apply: merge variant mappings into matching items, pass others through.
// This mirrors CUE semantics where each variant condition is evaluated for every
// item in the loop body — non-matching items are kept so later MapVariant ops
// can process them.
result := make([]any, 0, len(items))
for _, item := range items {
itemMap, ok := item.(map[string]any)
if !ok {
result = append(result, item)
continue
}
disc, exists := itemMap[m.discriminator]
if !exists || fmt.Sprintf("%v", disc) != m.variantName {
// Non-matching: pass through unchanged
result = append(result, item)
continue
}
// Matching: copy existing fields and merge variant mappings
newItem := make(map[string]any, len(itemMap)+len(m.mappings))
for k, v := range itemMap {
newItem[k] = v
}
for newKey, fieldVal := range m.mappings {
newItem[newKey] = fieldVal.resolve(itemMap)
}
result = append(result, newItem)
}
return result
}
type pickOp struct {
fields []string
}
func (p *pickOp) apply(items []any) []any {
result := make([]any, 0, len(items))
for _, item := range items {
if m, ok := item.(map[string]any); ok {
newItem := make(map[string]any)
for _, field := range p.fields {
if val, exists := m[field]; exists {
newItem[field] = val
}
}
result = append(result, newItem)
}
}
return result
}
type renameOp struct {
from, to string
}
func (r *renameOp) apply(items []any) []any {
result := make([]any, 0, len(items))
for _, item := range items {
if m, ok := item.(map[string]any); ok {
newItem := make(map[string]any)
for k, v := range m {
if k == r.from {
newItem[r.to] = v
} else {
newItem[k] = v
}
}
result = append(result, newItem)
}
}
return result
}
type wrapOp struct {
key string
}
func (w *wrapOp) apply(items []any) []any {
result := make([]any, 0, len(items))
for _, item := range items {
result = append(result, map[string]any{w.key: item})
}
return result
}
type defaultFieldOp struct {
field string
defaultVal FieldValue
}
func (d *defaultFieldOp) apply(items []any) []any {
result := make([]any, 0, len(items))
for _, item := range items {
if m, ok := item.(map[string]any); ok {
newItem := make(map[string]any)
for k, v := range m {
newItem[k] = v
}
if newItem[d.field] == nil || newItem[d.field] == "" {
newItem[d.field] = d.defaultVal.resolve(m)
}
result = append(result, newItem)
}
}
return result
}
type flattenOp struct{}
func (f *flattenOp) apply(items []any) []any {
result := make([]any, 0)
for _, item := range items {
if arr, ok := item.([]any); ok {
result = append(result, arr...)
} else {
result = append(result, item)
}
}
return result
}
// --- Multi-source collection for volumeMounts pattern ---
// MultiSource combines items from multiple named sources.
type MultiSource struct {
source Value
sources []string
ops []collectionOperation
mapBySource map[string]FieldMap // Per-source-type mappings
}
func (m *MultiSource) expr() {}
func (m *MultiSource) value() {}
// FromFields creates a multi-source collection from named fields of an object.
// Usage: FromFields(volumeMounts, "pvc", "configMap", "secret", "emptyDir", "hostPath")
func FromFields(source Value, fields ...string) *MultiSource {
return &MultiSource{
source: source,
sources: fields,
ops: make([]collectionOperation, 0),
mapBySource: make(map[string]FieldMap),
}
}
// Pick selects only specified fields from each item.
func (m *MultiSource) Pick(fields ...string) *MultiSource {
m.ops = append(m.ops, &pickOp{fields: fields})
return m
}
// Dedupe removes duplicate items by a key field.
func (m *MultiSource) Dedupe(keyField string) *MultiSource {
m.ops = append(m.ops, &dedupeOp{keyField: keyField})
return m
}
// Filter keeps only items matching the predicate.
func (m *MultiSource) Filter(pred Predicate) *MultiSource {
m.ops = append(m.ops, &filterOp{pred: pred})
return m
}
// FilterCond keeps only items matching a Condition expression.
// The condition is used for CUE generation; runtime apply is a passthrough.
func (m *MultiSource) FilterCond(cond Condition) *MultiSource {
m.ops = append(m.ops, &filterCondCollectionOp{cond: cond})
return m
}
// MapBySource applies different field mappings based on the source field name.
// This is useful for volumeMounts where pvc, configMap, secret each have different output formats.
//
// Usage:
//
// FromFields(volumeMounts, "pvc", "configMap", "secret", "emptyDir", "hostPath").
// MapBySource(map[string]FieldMap{
// "pvc": {"name": FieldRef("name"), "persistentVolumeClaim": Nested(FieldMap{"claimName": FieldRef("claimName")})},
// "configMap": {"name": FieldRef("name"), "configMap": Nested(FieldMap{"name": FieldRef("cmName"), "defaultMode": FieldRef("defaultMode")})},
// ...
// }).
// Dedupe("name")
func (m *MultiSource) MapBySource(mappings map[string]FieldMap) *MultiSource {
m.mapBySource = mappings
return m
}
// Source returns the source value.
func (m *MultiSource) Source() Value { return m.source }
// Sources returns the source field names.
func (m *MultiSource) Sources() []string { return m.sources }
// Operations returns the operations.
func (m *MultiSource) Operations() []collectionOperation { return m.ops }
// MapBySourceMappings returns the per-source-type field mappings.
func (m *MultiSource) MapBySourceMappings() map[string]FieldMap { return m.mapBySource }
// --- Go 1.23 Iterator Methods for MultiSource (iter.Seq) ---
// All returns an iterator over all items from all sources after applying operations.
// Items are collected from each source field and transformed according to MapBySource mappings.
// Example: for item := range ms.All(sourceData) { ... }
func (m *MultiSource) All(sourceData map[string]any) iter.Seq[map[string]any] {
return func(yield func(map[string]any) bool) {
// Collect items from all sources
var allItems []any
for _, sourceName := range m.sources {
if sourceItems, ok := sourceData[sourceName]; ok {
if arr, ok := sourceItems.([]map[string]any); ok {
// Apply per-source mapping if defined
if mapping, hasMapping := m.mapBySource[sourceName]; hasMapping {
for _, item := range arr {
transformed := applyFieldMap(item, mapping)
allItems = append(allItems, transformed)
}
} else {
for _, item := range arr {
allItems = append(allItems, item)
}
}
} else if arr, ok := sourceItems.([]any); ok {
// Apply per-source mapping if defined
if mapping, hasMapping := m.mapBySource[sourceName]; hasMapping {
for _, item := range arr {
if itemMap, ok := item.(map[string]any); ok {
transformed := applyFieldMap(itemMap, mapping)
allItems = append(allItems, transformed)
}
}
} else {
allItems = append(allItems, arr...)
}
}
}
}
// Apply operations
result := allItems
for _, op := range m.ops {
result = op.apply(result)
}
// Yield each item
for _, item := range result {
if itemMap, ok := item.(map[string]any); ok {
if !yield(itemMap) {
return
}
}
}
}
}
// AllPairs returns a key-value iterator with index and item from all sources.
// Example: for i, item := range ms.AllPairs(sourceData) { ... }
func (m *MultiSource) AllPairs(sourceData map[string]any) iter.Seq2[int, map[string]any] {
return func(yield func(int, map[string]any) bool) {
i := 0
for item := range m.All(sourceData) {
if !yield(i, item) {
return
}
i++
}
}
}
// Collect materializes the iterator into a slice.
// Uses Go 1.23 slices.Collect for efficient collection.
func (m *MultiSource) Collect(sourceData map[string]any) []map[string]any {
return slices.Collect(m.All(sourceData))
}
// Count returns the number of items from all sources after applying operations.
func (m *MultiSource) Count(sourceData map[string]any) int {
count := 0
for range m.All(sourceData) {
count++
}
return count
}
type dedupeOp struct {
keyField string
}
func (d *dedupeOp) apply(items []any) []any {
seen := make(map[any]bool)
result := make([]any, 0, len(items))
for _, item := range items {
if m, ok := item.(map[string]any); ok {
key := m[d.keyField]
if !seen[key] {
seen[key] = true
result = append(result, item)
}
}
}
return result
}
//lint:ignore U1000 planned for future use
type transformByTypeOp struct {
transforms map[string]FieldMap
}
//lint:ignore U1000 planned for future use
func (t *transformByTypeOp) apply(items []any) []any {
// This is applied during resolution, not here
return items
}
// --- Nested Object Field Mapping ---
// NestedField creates a nested object with the given field mappings.
// Usage: Map(FieldMap{"persistentVolumeClaim": Nested(FieldMap{"claimName": FieldRef("claimName")})})
type NestedField struct {
mapping FieldMap
}
func (n *NestedField) resolve(item map[string]any) any {
result := make(map[string]any)
for k, v := range n.mapping {
resolved := v.resolve(item)
if resolved != nil {
result[k] = resolved
}
}
return result
}
// Nested creates a nested object from field mappings.
// Usage: "persistentVolumeClaim": Nested(FieldMap{"claimName": FieldRef("claimName")})
func Nested(mapping FieldMap) *NestedField {
return &NestedField{mapping: mapping}
}
// --- Optional Field Value ---
// OptionalField includes a field value only if it exists and is not nil.
type OptionalField struct {
field string
}
func (o *OptionalField) resolve(item map[string]any) any {
val, exists := item[o.field]
if !exists || val == nil {
return nil
}
return val
}
// Optional creates a field reference that returns nil if the field doesn't exist.
// This is useful for optional fields that should only appear in output if present.
func Optional(field string) *OptionalField {
return &OptionalField{field: field}
}
// OptionalFieldRef is an alias for Optional, providing a clearer name for the pattern.
// Usage: defkit.OptionalFieldRef("subPath")
func OptionalFieldRef(field string) *OptionalField {
return &OptionalField{field: field}
}
// CompoundOptionalField includes a field value only if the field exists AND an additional condition is met.
// Generates CUE: if v.field != _|_ if additionalCond { fieldName: v.field }
type CompoundOptionalField struct {
field string
additionalCond Condition
}
func (o *CompoundOptionalField) resolve(item map[string]any) any {
val, exists := item[o.field]
if !exists || val == nil {
return nil
}
return val
}
// OptionalFieldWithCond creates a field reference that includes the field only when
// both the field exists and the additional condition is satisfied.
// Generates CUE: if v.field != _|_ if cond { fieldName: v.field }
func OptionalFieldWithCond(field string, cond Condition) *CompoundOptionalField {
return &CompoundOptionalField{field: field, additionalCond: cond}
}
// NestedFieldMap creates a nested object from field mappings.
// This is an alias for Nested, providing a clearer name for struct array helpers.
// Usage: defkit.NestedFieldMap(defkit.FieldMap{"claimName": defkit.FieldRef("claimName")})
func NestedFieldMap(mapping FieldMap) *NestedField {
return &NestedField{mapping: mapping}
}
// ConcatExprValue represents a concatenation expression for array helpers.
// This is used for generating: mountsArray.pvc + mountsArray.configMap + ...
type ConcatExprValue struct {
source *StructArrayHelper
fields []string
}
func (c *ConcatExprValue) value() {}
func (c *ConcatExprValue) expr() {}
// Source returns the source struct array helper.
func (c *ConcatExprValue) Source() *StructArrayHelper { return c.source }
// Fields returns the field names to concatenate.
func (c *ConcatExprValue) Fields() []string { return c.fields }
// ConcatExpr creates a concatenation expression from a struct array helper.
// This generates CUE like: mountsArray.pvc + mountsArray.configMap + mountsArray.secret + ...
// Used when setting volumeMounts on containers.
//
// Usage:
//
// SetIf(volumeMounts.IsSet(), "spec.containers[0].volumeMounts",
// defkit.ConcatExpr(mountsArray, "pvc", "configMap", "secret", "emptyDir", "hostPath"))
func ConcatExpr(source *StructArrayHelper, fields ...string) *ConcatExprValue {
return &ConcatExprValue{
source: source,
fields: fields,
}
}