Bump github.com/gookit/config/v2 from 2.1.8 to 2.2.2

Bumps [github.com/gookit/config/v2](https://github.com/gookit/config) from 2.1.8 to 2.2.2.
- [Release notes](https://github.com/gookit/config/releases)
- [Commits](https://github.com/gookit/config/compare/v2.1.8...v2.2.2)

---
updated-dependencies:
- dependency-name: github.com/gookit/config/v2
  dependency-type: direct:production
  update-type: version-update:semver-minor
...

Signed-off-by: dependabot[bot] <support@github.com>

vendor/github.com/goccy/go-yaml/.codecov.yml

@@ -0,0 +1,31 @@
+codecov:
+  require_ci_to_pass: yes
+
+coverage:
+  precision: 2
+  round: down
+  range: "70...100"
+
+  status:
+    project:
+      default:
+        target: 75%
+        threshold: 2%
+    patch: off
+    changes: no
+
+parsers:
+  gcov:
+    branch_detection:
+      conditional: yes
+      loop: yes
+      method: no
+      macro: no
+
+comment:
+  layout: "header,diff"
+  behavior: default
+  require_changes: no
+
+ignore:
+  - ast

vendor/github.com/goccy/go-yaml/CHANGELOG.md

@@ -0,0 +1,163 @@
+# 1.10.1 - 2023-03-28
+
+### Features
+
+- Quote YAML 1.1 bools at encoding time for compatibility with other legacy parsers
+- Add support of 32-bit architecture
+
+### Fix bugs
+
+- Don't trim all space characters in block style sequence
+- Support strings starting with `@`
+
+# 1.10.0 - 2023-03-01
+
+### Fix bugs
+
+Reversible conversion of comments was not working in various cases, which has been corrected.
+**Breaking Change** exists in the comment map interface. However, if you are dealing with CommentMap directly, there is no problem.
+
+
+# 1.9.8 - 2022-12-19
+
+### Fix feature
+
+- Append new line at the end of file ( #329 )
+
+### Fix bugs
+
+- Fix custom marshaler ( #333, #334 )
+- Fix behavior when struct fields conflicted( #335 )
+- Fix position calculation for literal, folded and raw folded strings ( #330 )
+
+# 1.9.7 - 2022-12-03
+
+### Fix bugs
+
+- Fix handling of quoted map key ( #328 )
+- Fix resusing process of scanning context ( #322 )
+
+## v1.9.6 - 2022-10-26
+
+### New Features
+
+- Introduce MapKeyNode interface to limit node types for map key ( #312 )
+
+### Fix bugs
+
+- Quote strings with special characters in flow mode ( #270 )
+- typeError implements PrettyPrinter interface ( #280 )
+- Fix incorrect const type ( #284 )
+- Fix large literals type inference on 32 bits ( #293 )
+- Fix UTF-8 characters ( #294 )
+- Fix decoding of unknown aliases ( #317 )
+- Fix stream encoder for insert a separator between each encoded document ( #318 )
+
+### Update
+
+- Update golang.org/x/sys ( #289 )
+- Update Go version in CI ( #295 )
+- Add test cases for missing keys to struct literals ( #300 )
+
+## v1.9.5 - 2022-01-12
+
+### New Features
+
+* Add UseSingleQuote option ( #265 )
+
+### Fix bugs
+
+* Preserve defaults while decoding nested structs ( #260 )
+* Fix minor typo in decodeInit error ( #264 )
+* Handle empty sequence entries ( #275 )
+* Fix encoding of sequence with multiline string ( #276 )
+* Fix encoding of BytesMarshaler type ( #277 )
+* Fix indentState logic for multi-line value ( #278 )
+
+## v1.9.4 - 2021-10-12
+
+### Fix bugs
+
+* Keep prev/next reference between tokens containing comments when filtering comment tokens ( #257 )
+* Supports escaping reserved keywords in PathBuilder ( #258 )
+
+## v1.9.3 - 2021-09-07
+
+### New Features
+
+* Support encoding and decoding `time.Duration` fields ( #246 )
+* Allow reserved characters for key name in YAMLPath ( #251 )
+* Support getting YAMLPath from ast.Node ( #252 )
+* Support CommentToMap option ( #253 )
+
+### Fix bugs
+
+* Fix encoding nested sequences with `yaml.IndentSequence` ( #241 )
+* Fix error reporting on inline structs in strict mode ( #244, #245 )
+* Fix encoding of large floats ( #247 )
+
+### Improve workflow
+
+* Migrate CI from CircleCI to GitHub Action ( #249 )
+* Add workflow for ycat ( #250 )
+
+## v1.9.2 - 2021-07-26
+
+### Support WithComment option ( #238 )
+
+`yaml.WithComment` is a option for encoding with comment.
+The position where you want to add a comment is represented by YAMLPath, and it is the key of `yaml.CommentMap`.
+Also, you can select `Head` comment or `Line` comment as the comment type.
+
+## v1.9.1 - 2021-07-20
+
+### Fix DecodeFromNode ( #237 )
+
+- Fix YAML handling where anchor exists
+
+## v1.9.0 - 2021-07-19
+
+### New features
+
+- Support encoding of comment node ( #233 )
+- Support `yaml.NodeToValue(ast.Node, interface{}, ...DecodeOption) error` ( #236 )
+  - Can convert a AST node to a value directly
+
+### Fix decoder for comment
+
+- Fix parsing of literal with comment ( #234 )
+
+### Rename API ( #235 )
+
+- Rename `MarshalWithContext` to `MarshalContext`
+- Rename `UnmarshalWithContext` to `UnmarshalContext`
+
+## v1.8.10 - 2021-07-02
+
+### Fixed bugs
+
+- Fix searching anchor by alias name ( #212 )
+- Fixing Issue 186, scanner should account for newline characters when processing multi-line text. Without this source annotations line/column number (for this and all subsequent tokens) is inconsistent with plain text editors. e.g. https://github.com/goccy/go-yaml/issues/186. This addresses the issue specifically for single and double quote text only. ( #210 )
+- Add error for unterminated flow mapping node ( #213 )
+- Handle missing required field validation ( #221 )
+- Nicely format unexpected node type errors ( #229 )
+- Support to encode map which has defined type key ( #231 )
+
+### New features
+
+- Support sequence indentation by EncodeOption ( #232 )
+
+## v1.8.9 - 2021-03-01
+
+### Fixed bugs
+
+- Fix origin buffer for DocumentHeader and DocumentEnd and Directive
+- Fix origin buffer for anchor value
+- Fix syntax error about map value
+- Fix parsing MergeKey ('<<') characters
+- Fix encoding of float value
+- Fix incorrect column annotation when single or double quotes are used
+
+### New features
+
+- Support to encode/decode of ast.Node directly

vendor/github.com/goccy/go-yaml/LICENSE

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2019 Masaaki Goshima
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.

vendor/github.com/goccy/go-yaml/Makefile

@@ -0,0 +1,19 @@
+.PHONY: test
+test:
+	go test -v -race ./...
+
+.PHONY: simple-test
+simple-test:
+	go test -v ./...
+
+.PHONY: cover
+cover:
+	go test -coverprofile=cover.out ./...
+
+.PHONY: cover-html
+cover-html: cover
+	go tool cover -html=cover.out
+
+.PHONY: ycat/build
+ycat/build:
+	go build -o ycat ./cmd/ycat

vendor/github.com/goccy/go-yaml/README.md

@@ -0,0 +1,370 @@
+# YAML support for the Go language
+
+[![PkgGoDev](https://pkg.go.dev/badge/github.com/goccy/go-yaml)](https://pkg.go.dev/github.com/goccy/go-yaml)
+![Go](https://github.com/goccy/go-yaml/workflows/Go/badge.svg)
+[![codecov](https://codecov.io/gh/goccy/go-yaml/branch/master/graph/badge.svg)](https://codecov.io/gh/goccy/go-yaml)
+[![Go Report Card](https://goreportcard.com/badge/github.com/goccy/go-yaml)](https://goreportcard.com/report/github.com/goccy/go-yaml)
+
+<img width="300px" src="https://user-images.githubusercontent.com/209884/67159116-64d94b80-f37b-11e9-9b28-f8379636a43c.png"></img>
+
+# Why a new library?
+
+As of this writing, there already exists a de facto standard library for YAML processing for Go: [https://github.com/go-yaml/yaml](https://github.com/go-yaml/yaml). However we feel that some features are lacking, namely:
+
+- Pretty format for error notifications
+- Direct manipulation of YAML abstract syntax tree
+- Support for `Anchor` and `Alias` when marshaling
+- Allow referencing elements declared in another file via anchors
+
+# Features
+
+- Pretty format for error notifications
+- Supports `Scanner` or `Lexer` or `Parser` as public API
+- Supports `Anchor` and `Alias` to Marshaler
+- Allow referencing elements declared in another file via anchors
+- Extract value or AST by YAMLPath ( YAMLPath is like a JSONPath )
+
+# Installation
+
+```sh
+go get -u github.com/goccy/go-yaml
+```
+
+# Synopsis
+
+## 1. Simple Encode/Decode
+
+Has an interface like `go-yaml/yaml` using `reflect`
+
+```go
+var v struct {
+	A int
+	B string
+}
+v.A = 1
+v.B = "hello"
+bytes, err := yaml.Marshal(v)
+if err != nil {
+	//...
+}
+fmt.Println(string(bytes)) // "a: 1\nb: hello\n"
+```
+
+```go
+	yml := `
+%YAML 1.2
+---
+a: 1
+b: c
+`
+var v struct {
+	A int
+	B string
+}
+if err := yaml.Unmarshal([]byte(yml), &v); err != nil {
+	//...
+}
+```
+
+To control marshal/unmarshal behavior, you can use the `yaml` tag.
+
+```go
+	yml := `---
+foo: 1
+bar: c
+`
+var v struct {
+	A int    `yaml:"foo"`
+	B string `yaml:"bar"`
+}
+if err := yaml.Unmarshal([]byte(yml), &v); err != nil {
+	//...
+}
+```
+
+For convenience, we also accept the `json` tag. Note that not all options from
+the `json` tag will have significance when parsing YAML documents. If both
+tags exist, `yaml` tag will take precedence.
+
+```go
+	yml := `---
+foo: 1
+bar: c
+`
+var v struct {
+	A int    `json:"foo"`
+	B string `json:"bar"`
+}
+if err := yaml.Unmarshal([]byte(yml), &v); err != nil {
+	//...
+}
+```
+
+For custom marshal/unmarshaling, implement either `Bytes` or `Interface` variant of marshaler/unmarshaler. The difference is that while `BytesMarshaler`/`BytesUnmarshaler` behaves like [`encoding/json`](https://pkg.go.dev/encoding/json) and `InterfaceMarshaler`/`InterfaceUnmarshaler` behaves like [`gopkg.in/yaml.v2`](https://pkg.go.dev/gopkg.in/yaml.v2).
+
+Semantically both are the same, but they differ in performance. Because indentation matters in YAML, you cannot simply accept a valid YAML fragment from a Marshaler, and expect it to work when it is attached to the parent container's serialized form. Therefore when we receive use the `BytesMarshaler`, which returns `[]byte`, we must decode it once to figure out how to make it work in the given context. If you use the `InterfaceMarshaler`, we can skip the decoding.
+
+If you are repeatedly marshaling complex objects, the latter is always better
+performance wise. But if you are, for example, just providing a choice between
+a config file format that is read only once, the former is probably easier to
+code.
+
+## 2. Reference elements declared in another file
+
+`testdata` directory contains `anchor.yml` file:
+
+```shell
+├── testdata
+   └── anchor.yml
+```
+
+And `anchor.yml` is defined as follows:
+
+```yaml
+a: &a
+  b: 1
+  c: hello
+```
+
+Then, if `yaml.ReferenceDirs("testdata")` option is passed to `yaml.Decoder`, 
+ `Decoder` tries to find the anchor definition from YAML files the under `testdata` directory.
+ 
+```go
+buf := bytes.NewBufferString("a: *a\n")
+dec := yaml.NewDecoder(buf, yaml.ReferenceDirs("testdata"))
+var v struct {
+	A struct {
+		B int
+		C string
+	}
+}
+if err := dec.Decode(&v); err != nil {
+	//...
+}
+fmt.Printf("%+v\n", v) // {A:{B:1 C:hello}}
+```
+
+## 3. Encode with `Anchor` and `Alias`
+
+### 3.1. Explicitly declared `Anchor` name and `Alias` name
+
+If you want to use `anchor` or `alias`, you can define it as a struct tag.
+
+```go
+type T struct {
+  A int
+  B string
+}
+var v struct {
+  C *T `yaml:"c,anchor=x"`
+  D *T `yaml:"d,alias=x"`
+}
+v.C = &T{A: 1, B: "hello"}
+v.D = v.C
+bytes, err := yaml.Marshal(v)
+if err != nil {
+  panic(err)
+}
+fmt.Println(string(bytes))
+/*
+c: &x
+  a: 1
+  b: hello
+d: *x
+*/
+```
+
+### 3.2. Implicitly declared `Anchor` and `Alias` names
+
+If you do not explicitly declare the anchor name, the default behavior is to
+use the equivalent of `strings.ToLower($FieldName)` as the name of the anchor.
+
+If you do not explicitly declare the alias name AND the value is a pointer
+to another element, we look up the anchor name by finding out which anchor
+field the value is assigned to by looking up its pointer address.
+
+```go
+type T struct {
+	I int
+	S string
+}
+var v struct {
+	A *T `yaml:"a,anchor"`
+	B *T `yaml:"b,anchor"`
+	C *T `yaml:"c,alias"`
+	D *T `yaml:"d,alias"`
+}
+v.A = &T{I: 1, S: "hello"}
+v.B = &T{I: 2, S: "world"}
+v.C = v.A // C has same pointer address to A
+v.D = v.B // D has same pointer address to B
+bytes, err := yaml.Marshal(v)
+if err != nil {
+	//...
+}
+fmt.Println(string(bytes)) 
+/*
+a: &a
+  i: 1
+  s: hello
+b: &b
+  i: 2
+  s: world
+c: *a
+d: *b
+*/
+```
+
+### 3.3 MergeKey and Alias
+
+Merge key and alias ( `<<: *alias` ) can be used by embedding a structure with the `inline,alias` tag.
+
+```go
+type Person struct {
+	*Person `yaml:",omitempty,inline,alias"` // embed Person type for default value
+	Name    string `yaml:",omitempty"`
+	Age     int    `yaml:",omitempty"`
+}
+defaultPerson := &Person{
+	Name: "John Smith",
+	Age:  20,
+}
+people := []*Person{
+	{
+		Person: defaultPerson, // assign default value
+		Name:   "Ken",         // override Name property
+		Age:    10,            // override Age property
+	},
+	{
+		Person: defaultPerson, // assign default value only
+	},
+}
+var doc struct {
+	Default *Person   `yaml:"default,anchor"`
+	People  []*Person `yaml:"people"`
+}
+doc.Default = defaultPerson
+doc.People = people
+bytes, err := yaml.Marshal(doc)
+if err != nil {
+	//...
+}
+fmt.Println(string(bytes))
+/*
+default: &default
+  name: John Smith
+  age: 20
+people:
+- <<: *default
+  name: Ken
+  age: 10
+- <<: *default
+*/
+```
+
+## 4. Pretty Formatted Errors
+
+Error values produced during parsing have two extra features over regular
+error values.
+
+First, by default, they contain extra information on the location of the error
+from the source YAML document, to make it easier to find the error location.
+
+Second, the error messages can optionally be colorized.
+
+If you would like to control exactly how the output looks like, consider
+using  `yaml.FormatError`, which accepts two boolean values to
+control turning these features on or off.
+
+<img src="https://user-images.githubusercontent.com/209884/67358124-587f0980-f59a-11e9-96fc-7205aab77695.png"></img>
+
+## 5. Use YAMLPath
+
+```go
+yml := `
+store:
+  book:
+    - author: john
+      price: 10
+    - author: ken
+      price: 12
+  bicycle:
+    color: red
+    price: 19.95
+`
+path, err := yaml.PathString("$.store.book[*].author")
+if err != nil {
+  //...
+}
+var authors []string
+if err := path.Read(strings.NewReader(yml), &authors); err != nil {
+  //...
+}
+fmt.Println(authors)
+// [john ken]
+```
+
+### 5.1 Print customized error with YAML source code
+
+```go
+package main
+
+import (
+  "fmt"
+
+  "github.com/goccy/go-yaml"
+)
+
+func main() {
+  yml := `
+a: 1
+b: "hello"
+`
+  var v struct {
+    A int
+    B string
+  }
+  if err := yaml.Unmarshal([]byte(yml), &v); err != nil {
+    panic(err)
+  }
+  if v.A != 2 {
+    // output error with YAML source
+    path, err := yaml.PathString("$.a")
+    if err != nil {
+      panic(err)
+    }
+    source, err := path.AnnotateSource([]byte(yml), true)
+    if err != nil {
+      panic(err)
+    }
+    fmt.Printf("a value expected 2 but actual %d:\n%s\n", v.A, string(source))
+  }
+}
+```
+
+output result is the following:
+
+<img src="https://user-images.githubusercontent.com/209884/84148813-7aca8680-aa9a-11ea-8fc9-37dece2ebdac.png"></img>
+
+
+# Tools
+
+## ycat
+
+print yaml file with color
+
+<img width="713" alt="ycat" src="https://user-images.githubusercontent.com/209884/66986084-19b00600-f0f9-11e9-9f0e-1f91eb072fe0.png">
+
+### Installation
+
+```sh
+go install github.com/goccy/go-yaml/cmd/ycat@latest
+```
+
+# Looking for Sponsors
+
+I'm looking for sponsors this library. This library is being developed as a personal project in my spare time. If you want a quick response or problem resolution when using this library in your project, please register as a [sponsor](https://github.com/sponsors/goccy). I will cooperate as much as possible. Of course, this library is developed as an MIT license, so you can use it freely for free.
+
+# License
+
+MIT

vendor/github.com/goccy/go-yaml/encode.go

@@ -0,0 +1,871 @@
+package yaml
+
+import (
+	"context"
+	"encoding"
+	"fmt"
+	"io"
+	"math"
+	"reflect"
+	"sort"
+	"strconv"
+	"strings"
+	"time"
+
+	"github.com/goccy/go-yaml/ast"
+	"github.com/goccy/go-yaml/internal/errors"
+	"github.com/goccy/go-yaml/parser"
+	"github.com/goccy/go-yaml/printer"
+	"github.com/goccy/go-yaml/token"
+	"golang.org/x/xerrors"
+)
+
+const (
+	// DefaultIndentSpaces default number of space for indent
+	DefaultIndentSpaces = 2
+)
+
+// Encoder writes YAML values to an output stream.
+type Encoder struct {
+	writer                     io.Writer
+	opts                       []EncodeOption
+	indent                     int
+	indentSequence             bool
+	singleQuote                bool
+	isFlowStyle                bool
+	isJSONStyle                bool
+	useJSONMarshaler           bool
+	anchorCallback             func(*ast.AnchorNode, interface{}) error
+	anchorPtrToNameMap         map[uintptr]string
+	customMarshalerMap         map[reflect.Type]func(interface{}) ([]byte, error)
+	useLiteralStyleIfMultiline bool
+	commentMap                 map[*Path][]*Comment
+	written                    bool
+
+	line        int
+	column      int
+	offset      int
+	indentNum   int
+	indentLevel int
+}
+
+// NewEncoder returns a new encoder that writes to w.
+// The Encoder should be closed after use to flush all data to w.
+func NewEncoder(w io.Writer, opts ...EncodeOption) *Encoder {
+	return &Encoder{
+		writer:             w,
+		opts:               opts,
+		indent:             DefaultIndentSpaces,
+		anchorPtrToNameMap: map[uintptr]string{},
+		customMarshalerMap: map[reflect.Type]func(interface{}) ([]byte, error){},
+		line:               1,
+		column:             1,
+		offset:             0,
+	}
+}
+
+// Close closes the encoder by writing any remaining data.
+// It does not write a stream terminating string "...".
+func (e *Encoder) Close() error {
+	return nil
+}
+
+// Encode writes the YAML encoding of v to the stream.
+// If multiple items are encoded to the stream,
+// the second and subsequent document will be preceded with a "---" document separator,
+// but the first will not.
+//
+// See the documentation for Marshal for details about the conversion of Go values to YAML.
+func (e *Encoder) Encode(v interface{}) error {
+	return e.EncodeContext(context.Background(), v)
+}
+
+// EncodeContext writes the YAML encoding of v to the stream with context.Context.
+func (e *Encoder) EncodeContext(ctx context.Context, v interface{}) error {
+	node, err := e.EncodeToNodeContext(ctx, v)
+	if err != nil {
+		return errors.Wrapf(err, "failed to encode to node")
+	}
+	if err := e.setCommentByCommentMap(node); err != nil {
+		return errors.Wrapf(err, "failed to set comment by comment map")
+	}
+	if !e.written {
+		e.written = true
+	} else {
+		// write document separator
+		e.writer.Write([]byte("---\n"))
+	}
+	var p printer.Printer
+	e.writer.Write(p.PrintNode(node))
+	return nil
+}
+
+// EncodeToNode convert v to ast.Node.
+func (e *Encoder) EncodeToNode(v interface{}) (ast.Node, error) {
+	return e.EncodeToNodeContext(context.Background(), v)
+}
+
+// EncodeToNodeContext convert v to ast.Node with context.Context.
+func (e *Encoder) EncodeToNodeContext(ctx context.Context, v interface{}) (ast.Node, error) {
+	for _, opt := range e.opts {
+		if err := opt(e); err != nil {
+			return nil, errors.Wrapf(err, "failed to run option for encoder")
+		}
+	}
+	node, err := e.encodeValue(ctx, reflect.ValueOf(v), 1)
+	if err != nil {
+		return nil, errors.Wrapf(err, "failed to encode value")
+	}
+	return node, nil
+}
+
+func (e *Encoder) setCommentByCommentMap(node ast.Node) error {
+	if e.commentMap == nil {
+		return nil
+	}
+	for path, comments := range e.commentMap {
+		n, err := path.FilterNode(node)
+		if err != nil {
+			return errors.Wrapf(err, "failed to filter node")
+		}
+		if n == nil {
+			continue
+		}
+		for _, comment := range comments {
+			commentTokens := []*token.Token{}
+			for _, text := range comment.Texts {
+				commentTokens = append(commentTokens, token.New(text, text, nil))
+			}
+			commentGroup := ast.CommentGroup(commentTokens)
+			switch comment.Position {
+			case CommentHeadPosition:
+				if err := e.setHeadComment(node, n, commentGroup); err != nil {
+					return errors.Wrapf(err, "failed to set head comment")
+				}
+			case CommentLinePosition:
+				if err := e.setLineComment(node, n, commentGroup); err != nil {
+					return errors.Wrapf(err, "failed to set line comment")
+				}
+			case CommentFootPosition:
+				if err := e.setFootComment(node, n, commentGroup); err != nil {
+					return errors.Wrapf(err, "failed to set foot comment")
+				}
+			default:
+				return ErrUnknownCommentPositionType
+			}
+		}
+	}
+	return nil
+}
+
+func (e *Encoder) setHeadComment(node ast.Node, filtered ast.Node, comment *ast.CommentGroupNode) error {
+	parent := ast.Parent(node, filtered)
+	if parent == nil {
+		return ErrUnsupportedHeadPositionType(node)
+	}
+	switch p := parent.(type) {
+	case *ast.MappingValueNode:
+		if err := p.SetComment(comment); err != nil {
+			return errors.Wrapf(err, "failed to set comment")
+		}
+	case *ast.MappingNode:
+		if err := p.SetComment(comment); err != nil {
+			return errors.Wrapf(err, "failed to set comment")
+		}
+	case *ast.SequenceNode:
+		if len(p.ValueHeadComments) == 0 {
+			p.ValueHeadComments = make([]*ast.CommentGroupNode, len(p.Values))
+		}
+		var foundIdx int
+		for idx, v := range p.Values {
+			if v == filtered {
+				foundIdx = idx
+				break
+			}
+		}
+		p.ValueHeadComments[foundIdx] = comment
+	default:
+		return ErrUnsupportedHeadPositionType(node)
+	}
+	return nil
+}
+
+func (e *Encoder) setLineComment(node ast.Node, filtered ast.Node, comment *ast.CommentGroupNode) error {
+	switch filtered.(type) {
+	case *ast.MappingValueNode, *ast.SequenceNode:
+		// Line comment cannot be set for mapping value node.
+		// It should probably be set for the parent map node
+		if err := e.setLineCommentToParentMapNode(node, filtered, comment); err != nil {
+			return errors.Wrapf(err, "failed to set line comment to parent node")
+		}
+	default:
+		if err := filtered.SetComment(comment); err != nil {
+			return errors.Wrapf(err, "failed to set comment")
+		}
+	}
+	return nil
+}
+
+func (e *Encoder) setLineCommentToParentMapNode(node ast.Node, filtered ast.Node, comment *ast.CommentGroupNode) error {
+	parent := ast.Parent(node, filtered)
+	if parent == nil {
+		return ErrUnsupportedLinePositionType(node)
+	}
+	switch p := parent.(type) {
+	case *ast.MappingValueNode:
+		if err := p.Key.SetComment(comment); err != nil {
+			return errors.Wrapf(err, "failed to set comment")
+		}
+	case *ast.MappingNode:
+		if err := p.SetComment(comment); err != nil {
+			return errors.Wrapf(err, "failed to set comment")
+		}
+	default:
+		return ErrUnsupportedLinePositionType(parent)
+	}
+	return nil
+}
+
+func (e *Encoder) setFootComment(node ast.Node, filtered ast.Node, comment *ast.CommentGroupNode) error {
+	parent := ast.Parent(node, filtered)
+	if parent == nil {
+		return ErrUnsupportedFootPositionType(node)
+	}
+	switch n := parent.(type) {
+	case *ast.MappingValueNode:
+		n.FootComment = comment
+	case *ast.MappingNode:
+		n.FootComment = comment
+	case *ast.SequenceNode:
+		n.FootComment = comment
+	default:
+		return ErrUnsupportedFootPositionType(n)
+	}
+	return nil
+}
+
+func (e *Encoder) encodeDocument(doc []byte) (ast.Node, error) {
+	f, err := parser.ParseBytes(doc, 0)
+	if err != nil {
+		return nil, errors.Wrapf(err, "failed to parse yaml")
+	}
+	for _, docNode := range f.Docs {
+		if docNode.Body != nil {
+			return docNode.Body, nil
+		}
+	}
+	return nil, nil
+}
+
+func (e *Encoder) isInvalidValue(v reflect.Value) bool {
+	if !v.IsValid() {
+		return true
+	}
+	kind := v.Type().Kind()
+	if kind == reflect.Ptr && v.IsNil() {
+		return true
+	}
+	if kind == reflect.Interface && v.IsNil() {
+		return true
+	}
+	return false
+}
+
+type jsonMarshaler interface {
+	MarshalJSON() ([]byte, error)
+}
+
+func (e *Encoder) existsTypeInCustomMarshalerMap(t reflect.Type) bool {
+	if _, exists := e.customMarshalerMap[t]; exists {
+		return true
+	}
+
+	globalCustomMarshalerMu.Lock()
+	defer globalCustomMarshalerMu.Unlock()
+	if _, exists := globalCustomMarshalerMap[t]; exists {
+		return true
+	}
+	return false
+}
+
+func (e *Encoder) marshalerFromCustomMarshalerMap(t reflect.Type) (func(interface{}) ([]byte, error), bool) {
+	if marshaler, exists := e.customMarshalerMap[t]; exists {
+		return marshaler, exists
+	}
+
+	globalCustomMarshalerMu.Lock()
+	defer globalCustomMarshalerMu.Unlock()
+	if marshaler, exists := globalCustomMarshalerMap[t]; exists {
+		return marshaler, exists
+	}
+	return nil, false
+}
+
+func (e *Encoder) canEncodeByMarshaler(v reflect.Value) bool {
+	if !v.CanInterface() {
+		return false
+	}
+	if e.existsTypeInCustomMarshalerMap(v.Type()) {
+		return true
+	}
+	iface := v.Interface()
+	switch iface.(type) {
+	case BytesMarshalerContext:
+		return true
+	case BytesMarshaler:
+		return true
+	case InterfaceMarshalerContext:
+		return true
+	case InterfaceMarshaler:
+		return true
+	case time.Time:
+		return true
+	case time.Duration:
+		return true
+	case encoding.TextMarshaler:
+		return true
+	case jsonMarshaler:
+		return e.useJSONMarshaler
+	}
+	return false
+}
+
+func (e *Encoder) encodeByMarshaler(ctx context.Context, v reflect.Value, column int) (ast.Node, error) {
+	iface := v.Interface()
+
+	if marshaler, exists := e.marshalerFromCustomMarshalerMap(v.Type()); exists {
+		doc, err := marshaler(iface)
+		if err != nil {
+			return nil, errors.Wrapf(err, "failed to MarshalYAML")
+		}
+		node, err := e.encodeDocument(doc)
+		if err != nil {
+			return nil, errors.Wrapf(err, "failed to encode document")
+		}
+		return node, nil
+	}
+
+	if marshaler, ok := iface.(BytesMarshalerContext); ok {
+		doc, err := marshaler.MarshalYAML(ctx)
+		if err != nil {
+			return nil, errors.Wrapf(err, "failed to MarshalYAML")
+		}
+		node, err := e.encodeDocument(doc)
+		if err != nil {
+			return nil, errors.Wrapf(err, "failed to encode document")
+		}
+		return node, nil
+	}
+
+	if marshaler, ok := iface.(BytesMarshaler); ok {
+		doc, err := marshaler.MarshalYAML()
+		if err != nil {
+			return nil, errors.Wrapf(err, "failed to MarshalYAML")
+		}
+		node, err := e.encodeDocument(doc)
+		if err != nil {
+			return nil, errors.Wrapf(err, "failed to encode document")
+		}
+		return node, nil
+	}
+
+	if marshaler, ok := iface.(InterfaceMarshalerContext); ok {
+		marshalV, err := marshaler.MarshalYAML(ctx)
+		if err != nil {
+			return nil, errors.Wrapf(err, "failed to MarshalYAML")
+		}
+		return e.encodeValue(ctx, reflect.ValueOf(marshalV), column)
+	}
+
+	if marshaler, ok := iface.(InterfaceMarshaler); ok {
+		marshalV, err := marshaler.MarshalYAML()
+		if err != nil {
+			return nil, errors.Wrapf(err, "failed to MarshalYAML")
+		}
+		return e.encodeValue(ctx, reflect.ValueOf(marshalV), column)
+	}
+
+	if t, ok := iface.(time.Time); ok {
+		return e.encodeTime(t, column), nil
+	}
+
+	if t, ok := iface.(time.Duration); ok {
+		return e.encodeDuration(t, column), nil
+	}
+
+	if marshaler, ok := iface.(encoding.TextMarshaler); ok {
+		doc, err := marshaler.MarshalText()
+		if err != nil {
+			return nil, errors.Wrapf(err, "failed to MarshalText")
+		}
+		node, err := e.encodeDocument(doc)
+		if err != nil {
+			return nil, errors.Wrapf(err, "failed to encode document")
+		}
+		return node, nil
+	}
+
+	if e.useJSONMarshaler {
+		if marshaler, ok := iface.(jsonMarshaler); ok {
+			jsonBytes, err := marshaler.MarshalJSON()
+			if err != nil {
+				return nil, errors.Wrapf(err, "failed to MarshalJSON")
+			}
+			doc, err := JSONToYAML(jsonBytes)
+			if err != nil {
+				return nil, errors.Wrapf(err, "failed to convert json to yaml")
+			}
+			node, err := e.encodeDocument(doc)
+			if err != nil {
+				return nil, errors.Wrapf(err, "failed to encode document")
+			}
+			return node, nil
+		}
+	}
+
+	return nil, xerrors.Errorf("does not implemented Marshaler")
+}
+
+func (e *Encoder) encodeValue(ctx context.Context, v reflect.Value, column int) (ast.Node, error) {
+	if e.isInvalidValue(v) {
+		return e.encodeNil(), nil
+	}
+	if e.canEncodeByMarshaler(v) {
+		node, err := e.encodeByMarshaler(ctx, v, column)
+		if err != nil {
+			return nil, errors.Wrapf(err, "failed to encode by marshaler")
+		}
+		return node, nil
+	}
+	switch v.Type().Kind() {
+	case reflect.Int, reflect.Int8, reflect.Int16, reflect.Int32, reflect.Int64:
+		return e.encodeInt(v.Int()), nil
+	case reflect.Uint, reflect.Uint8, reflect.Uint16, reflect.Uint32, reflect.Uint64:
+		return e.encodeUint(v.Uint()), nil
+	case reflect.Float32:
+		return e.encodeFloat(v.Float(), 32), nil
+	case reflect.Float64:
+		return e.encodeFloat(v.Float(), 64), nil
+	case reflect.Ptr:
+		anchorName := e.anchorPtrToNameMap[v.Pointer()]
+		if anchorName != "" {
+			aliasName := anchorName
+			alias := ast.Alias(token.New("*", "*", e.pos(column)))
+			alias.Value = ast.String(token.New(aliasName, aliasName, e.pos(column)))
+			return alias, nil
+		}
+		return e.encodeValue(ctx, v.Elem(), column)
+	case reflect.Interface:
+		return e.encodeValue(ctx, v.Elem(), column)
+	case reflect.String:
+		return e.encodeString(v.String(), column), nil
+	case reflect.Bool:
+		return e.encodeBool(v.Bool()), nil
+	case reflect.Slice:
+		if mapSlice, ok := v.Interface().(MapSlice); ok {
+			return e.encodeMapSlice(ctx, mapSlice, column)
+		}
+		return e.encodeSlice(ctx, v)
+	case reflect.Array:
+		return e.encodeArray(ctx, v)
+	case reflect.Struct:
+		if v.CanInterface() {
+			if mapItem, ok := v.Interface().(MapItem); ok {
+				return e.encodeMapItem(ctx, mapItem, column)
+			}
+			if t, ok := v.Interface().(time.Time); ok {
+				return e.encodeTime(t, column), nil
+			}
+		}
+		return e.encodeStruct(ctx, v, column)
+	case reflect.Map:
+		return e.encodeMap(ctx, v, column), nil
+	default:
+		return nil, xerrors.Errorf("unknown value type %s", v.Type().String())
+	}
+}
+
+func (e *Encoder) pos(column int) *token.Position {
+	return &token.Position{
+		Line:        e.line,
+		Column:      column,
+		Offset:      e.offset,
+		IndentNum:   e.indentNum,
+		IndentLevel: e.indentLevel,
+	}
+}
+
+func (e *Encoder) encodeNil() *ast.NullNode {
+	value := "null"
+	return ast.Null(token.New(value, value, e.pos(e.column)))
+}
+
+func (e *Encoder) encodeInt(v int64) *ast.IntegerNode {
+	value := fmt.Sprint(v)
+	return ast.Integer(token.New(value, value, e.pos(e.column)))
+}
+
+func (e *Encoder) encodeUint(v uint64) *ast.IntegerNode {
+	value := fmt.Sprint(v)
+	return ast.Integer(token.New(value, value, e.pos(e.column)))
+}
+
+func (e *Encoder) encodeFloat(v float64, bitSize int) ast.Node {
+	if v == math.Inf(0) {
+		value := ".inf"
+		return ast.Infinity(token.New(value, value, e.pos(e.column)))
+	} else if v == math.Inf(-1) {
+		value := "-.inf"
+		return ast.Infinity(token.New(value, value, e.pos(e.column)))
+	} else if math.IsNaN(v) {
+		value := ".nan"
+		return ast.Nan(token.New(value, value, e.pos(e.column)))
+	}
+	value := strconv.FormatFloat(v, 'g', -1, bitSize)
+	if !strings.Contains(value, ".") && !strings.Contains(value, "e") {
+		// append x.0 suffix to keep float value context
+		value = fmt.Sprintf("%s.0", value)
+	}
+	return ast.Float(token.New(value, value, e.pos(e.column)))
+}
+
+func (e *Encoder) isNeedQuoted(v string) bool {
+	if e.isJSONStyle {
+		return true
+	}
+	if e.useLiteralStyleIfMultiline && strings.ContainsAny(v, "\n\r") {
+		return false
+	}
+	if e.isFlowStyle && strings.ContainsAny(v, `]},'"`) {
+		return true
+	}
+	if token.IsNeedQuoted(v) {
+		return true
+	}
+	return false
+}
+
+func (e *Encoder) encodeString(v string, column int) *ast.StringNode {
+	if e.isNeedQuoted(v) {
+		if e.singleQuote {
+			v = quoteWith(v, '\'')
+		} else {
+			v = strconv.Quote(v)
+		}
+	}
+	return ast.String(token.New(v, v, e.pos(column)))
+}
+
+func (e *Encoder) encodeBool(v bool) *ast.BoolNode {
+	value := fmt.Sprint(v)
+	return ast.Bool(token.New(value, value, e.pos(e.column)))
+}
+
+func (e *Encoder) encodeSlice(ctx context.Context, value reflect.Value) (*ast.SequenceNode, error) {
+	if e.indentSequence {
+		e.column += e.indent
+	}
+	column := e.column
+	sequence := ast.Sequence(token.New("-", "-", e.pos(column)), e.isFlowStyle)
+	for i := 0; i < value.Len(); i++ {
+		node, err := e.encodeValue(ctx, value.Index(i), column)
+		if err != nil {
+			return nil, errors.Wrapf(err, "failed to encode value for slice")
+		}
+		sequence.Values = append(sequence.Values, node)
+	}
+	if e.indentSequence {
+		e.column -= e.indent
+	}
+	return sequence, nil
+}
+
+func (e *Encoder) encodeArray(ctx context.Context, value reflect.Value) (*ast.SequenceNode, error) {
+	if e.indentSequence {
+		e.column += e.indent
+	}
+	column := e.column
+	sequence := ast.Sequence(token.New("-", "-", e.pos(column)), e.isFlowStyle)
+	for i := 0; i < value.Len(); i++ {
+		node, err := e.encodeValue(ctx, value.Index(i), column)
+		if err != nil {
+			return nil, errors.Wrapf(err, "failed to encode value for array")
+		}
+		sequence.Values = append(sequence.Values, node)
+	}
+	if e.indentSequence {
+		e.column -= e.indent
+	}
+	return sequence, nil
+}
+
+func (e *Encoder) encodeMapItem(ctx context.Context, item MapItem, column int) (*ast.MappingValueNode, error) {
+	k := reflect.ValueOf(item.Key)
+	v := reflect.ValueOf(item.Value)
+	value, err := e.encodeValue(ctx, v, column)
+	if err != nil {
+		return nil, errors.Wrapf(err, "failed to encode MapItem")
+	}
+	if e.isMapNode(value) {
+		value.AddColumn(e.indent)
+	}
+	return ast.MappingValue(
+		token.New("", "", e.pos(column)),
+		e.encodeString(k.Interface().(string), column),
+		value,
+	), nil
+}
+
+func (e *Encoder) encodeMapSlice(ctx context.Context, value MapSlice, column int) (*ast.MappingNode, error) {
+	node := ast.Mapping(token.New("", "", e.pos(column)), e.isFlowStyle)
+	for _, item := range value {
+		value, err := e.encodeMapItem(ctx, item, column)
+		if err != nil {
+			return nil, errors.Wrapf(err, "failed to encode MapItem for MapSlice")
+		}
+		node.Values = append(node.Values, value)
+	}
+	return node, nil
+}
+
+func (e *Encoder) isMapNode(node ast.Node) bool {
+	_, ok := node.(ast.MapNode)
+	return ok
+}
+
+func (e *Encoder) encodeMap(ctx context.Context, value reflect.Value, column int) ast.Node {
+	node := ast.Mapping(token.New("", "", e.pos(column)), e.isFlowStyle)
+	keys := make([]interface{}, len(value.MapKeys()))
+	for i, k := range value.MapKeys() {
+		keys[i] = k.Interface()
+	}
+	sort.Slice(keys, func(i, j int) bool {
+		return fmt.Sprint(keys[i]) < fmt.Sprint(keys[j])
+	})
+	for _, key := range keys {
+		k := reflect.ValueOf(key)
+		v := value.MapIndex(k)
+		value, err := e.encodeValue(ctx, v, column)
+		if err != nil {
+			return nil
+		}
+		if e.isMapNode(value) {
+			value.AddColumn(e.indent)
+		}
+		node.Values = append(node.Values, ast.MappingValue(
+			nil,
+			e.encodeString(fmt.Sprint(key), column),
+			value,
+		))
+	}
+	return node
+}
+
+// IsZeroer is used to check whether an object is zero to determine
+// whether it should be omitted when marshaling with the omitempty flag.
+// One notable implementation is time.Time.
+type IsZeroer interface {
+	IsZero() bool
+}
+
+func (e *Encoder) isZeroValue(v reflect.Value) bool {
+	kind := v.Kind()
+	if z, ok := v.Interface().(IsZeroer); ok {
+		if (kind == reflect.Ptr || kind == reflect.Interface) && v.IsNil() {
+			return true
+		}
+		return z.IsZero()
+	}
+	switch kind {
+	case reflect.String:
+		return len(v.String()) == 0
+	case reflect.Interface, reflect.Ptr:
+		return v.IsNil()
+	case reflect.Slice:
+		return v.Len() == 0
+	case reflect.Map:
+		return v.Len() == 0
+	case reflect.Int, reflect.Int8, reflect.Int16, reflect.Int32, reflect.Int64:
+		return v.Int() == 0
+	case reflect.Float32, reflect.Float64:
+		return v.Float() == 0
+	case reflect.Uint, reflect.Uint8, reflect.Uint16, reflect.Uint32, reflect.Uint64, reflect.Uintptr:
+		return v.Uint() == 0
+	case reflect.Bool:
+		return !v.Bool()
+	case reflect.Struct:
+		vt := v.Type()
+		for i := v.NumField() - 1; i >= 0; i-- {
+			if vt.Field(i).PkgPath != "" {
+				continue // private field
+			}
+			if !e.isZeroValue(v.Field(i)) {
+				return false
+			}
+		}
+		return true
+	}
+	return false
+}
+
+func (e *Encoder) encodeTime(v time.Time, column int) *ast.StringNode {
+	value := v.Format(time.RFC3339Nano)
+	if e.isJSONStyle {
+		value = strconv.Quote(value)
+	}
+	return ast.String(token.New(value, value, e.pos(column)))
+}
+
+func (e *Encoder) encodeDuration(v time.Duration, column int) *ast.StringNode {
+	value := v.String()
+	if e.isJSONStyle {
+		value = strconv.Quote(value)
+	}
+	return ast.String(token.New(value, value, e.pos(column)))
+}
+
+func (e *Encoder) encodeAnchor(anchorName string, value ast.Node, fieldValue reflect.Value, column int) (*ast.AnchorNode, error) {
+	anchorNode := ast.Anchor(token.New("&", "&", e.pos(column)))
+	anchorNode.Name = ast.String(token.New(anchorName, anchorName, e.pos(column)))
+	anchorNode.Value = value
+	if e.anchorCallback != nil {
+		if err := e.anchorCallback(anchorNode, fieldValue.Interface()); err != nil {
+			return nil, errors.Wrapf(err, "failed to marshal anchor")
+		}
+		if snode, ok := anchorNode.Name.(*ast.StringNode); ok {
+			anchorName = snode.Value
+		}
+	}
+	if fieldValue.Kind() == reflect.Ptr {
+		e.anchorPtrToNameMap[fieldValue.Pointer()] = anchorName
+	}
+	return anchorNode, nil
+}
+
+func (e *Encoder) encodeStruct(ctx context.Context, value reflect.Value, column int) (ast.Node, error) {
+	node := ast.Mapping(token.New("", "", e.pos(column)), e.isFlowStyle)
+	structType := value.Type()
+	structFieldMap, err := structFieldMap(structType)
+	if err != nil {
+		return nil, errors.Wrapf(err, "failed to get struct field map")
+	}
+	hasInlineAnchorField := false
+	var inlineAnchorValue reflect.Value
+	for i := 0; i < value.NumField(); i++ {
+		field := structType.Field(i)
+		if isIgnoredStructField(field) {
+			continue
+		}
+		fieldValue := value.FieldByName(field.Name)
+		structField := structFieldMap[field.Name]
+		if structField.IsOmitEmpty && e.isZeroValue(fieldValue) {
+			// omit encoding
+			continue
+		}
+		ve := e
+		if !e.isFlowStyle && structField.IsFlow {
+			ve = &Encoder{}
+			*ve = *e
+			ve.isFlowStyle = true
+		}
+		value, err := ve.encodeValue(ctx, fieldValue, column)
+		if err != nil {
+			return nil, errors.Wrapf(err, "failed to encode value")
+		}
+		if e.isMapNode(value) {
+			value.AddColumn(e.indent)
+		}
+		var key ast.MapKeyNode = e.encodeString(structField.RenderName, column)
+		switch {
+		case structField.AnchorName != "":
+			anchorNode, err := e.encodeAnchor(structField.AnchorName, value, fieldValue, column)
+			if err != nil {
+				return nil, errors.Wrapf(err, "failed to encode anchor")
+			}
+			value = anchorNode
+		case structField.IsAutoAlias:
+			if fieldValue.Kind() != reflect.Ptr {
+				return nil, xerrors.Errorf(
+					"%s in struct is not pointer type. but required automatically alias detection",
+					structField.FieldName,
+				)
+			}
+			anchorName := e.anchorPtrToNameMap[fieldValue.Pointer()]
+			if anchorName == "" {
+				return nil, xerrors.Errorf(
+					"cannot find anchor name from pointer address for automatically alias detection",
+				)
+			}
+			aliasName := anchorName
+			alias := ast.Alias(token.New("*", "*", e.pos(column)))
+			alias.Value = ast.String(token.New(aliasName, aliasName, e.pos(column)))
+			value = alias
+			if structField.IsInline {
+				// if both used alias and inline, output `<<: *alias`
+				key = ast.MergeKey(token.New("<<", "<<", e.pos(column)))
+			}
+		case structField.AliasName != "":
+			aliasName := structField.AliasName
+			alias := ast.Alias(token.New("*", "*", e.pos(column)))
+			alias.Value = ast.String(token.New(aliasName, aliasName, e.pos(column)))
+			value = alias
+			if structField.IsInline {
+				// if both used alias and inline, output `<<: *alias`
+				key = ast.MergeKey(token.New("<<", "<<", e.pos(column)))
+			}
+		case structField.IsInline:
+			isAutoAnchor := structField.IsAutoAnchor
+			if !hasInlineAnchorField {
+				hasInlineAnchorField = isAutoAnchor
+			}
+			if isAutoAnchor {
+				inlineAnchorValue = fieldValue
+			}
+			mapNode, ok := value.(ast.MapNode)
+			if !ok {
+				return nil, xerrors.Errorf("inline value is must be map or struct type")
+			}
+			mapIter := mapNode.MapRange()
+			for mapIter.Next() {
+				key := mapIter.Key()
+				value := mapIter.Value()
+				keyName := key.GetToken().Value
+				if structFieldMap.isIncludedRenderName(keyName) {
+					// if declared same key name, skip encoding this field
+					continue
+				}
+				key.AddColumn(-e.indent)
+				value.AddColumn(-e.indent)
+				node.Values = append(node.Values, ast.MappingValue(nil, key, value))
+			}
+			continue
+		case structField.IsAutoAnchor:
+			anchorNode, err := e.encodeAnchor(structField.RenderName, value, fieldValue, column)
+			if err != nil {
+				return nil, errors.Wrapf(err, "failed to encode anchor")
+			}
+			value = anchorNode
+		}
+		node.Values = append(node.Values, ast.MappingValue(nil, key, value))
+	}
+	if hasInlineAnchorField {
+		node.AddColumn(e.indent)
+		anchorName := "anchor"
+		anchorNode := ast.Anchor(token.New("&", "&", e.pos(column)))
+		anchorNode.Name = ast.String(token.New(anchorName, anchorName, e.pos(column)))
+		anchorNode.Value = node
+		if e.anchorCallback != nil {
+			if err := e.anchorCallback(anchorNode, value.Addr().Interface()); err != nil {
+				return nil, errors.Wrapf(err, "failed to marshal anchor")
+			}
+			if snode, ok := anchorNode.Name.(*ast.StringNode); ok {
+				anchorName = snode.Value
+			}
+		}
+		if inlineAnchorValue.Kind() == reflect.Ptr {
+			e.anchorPtrToNameMap[inlineAnchorValue.Pointer()] = anchorName
+		}
+		return anchorNode, nil
+	}
+	return node, nil
+}

vendor/github.com/goccy/go-yaml/error.go

@@ -0,0 +1,62 @@
+package yaml
+
+import (
+	"github.com/goccy/go-yaml/ast"
+	"golang.org/x/xerrors"
+)
+
+var (
+	ErrInvalidQuery               = xerrors.New("invalid query")
+	ErrInvalidPath                = xerrors.New("invalid path instance")
+	ErrInvalidPathString          = xerrors.New("invalid path string")
+	ErrNotFoundNode               = xerrors.New("node not found")
+	ErrUnknownCommentPositionType = xerrors.New("unknown comment position type")
+	ErrInvalidCommentMapValue     = xerrors.New("invalid comment map value. it must be not nil value")
+)
+
+func ErrUnsupportedHeadPositionType(node ast.Node) error {
+	return xerrors.Errorf("unsupported comment head position for %s", node.Type())
+}
+
+func ErrUnsupportedLinePositionType(node ast.Node) error {
+	return xerrors.Errorf("unsupported comment line position for %s", node.Type())
+}
+
+func ErrUnsupportedFootPositionType(node ast.Node) error {
+	return xerrors.Errorf("unsupported comment foot position for %s", node.Type())
+}
+
+// IsInvalidQueryError whether err is ErrInvalidQuery or not.
+func IsInvalidQueryError(err error) bool {
+	return xerrors.Is(err, ErrInvalidQuery)
+}
+
+// IsInvalidPathError whether err is ErrInvalidPath or not.
+func IsInvalidPathError(err error) bool {
+	return xerrors.Is(err, ErrInvalidPath)
+}
+
+// IsInvalidPathStringError whether err is ErrInvalidPathString or not.
+func IsInvalidPathStringError(err error) bool {
+	return xerrors.Is(err, ErrInvalidPathString)
+}
+
+// IsNotFoundNodeError whether err is ErrNotFoundNode or not.
+func IsNotFoundNodeError(err error) bool {
+	return xerrors.Is(err, ErrNotFoundNode)
+}
+
+// IsInvalidTokenTypeError whether err is ast.ErrInvalidTokenType or not.
+func IsInvalidTokenTypeError(err error) bool {
+	return xerrors.Is(err, ast.ErrInvalidTokenType)
+}
+
+// IsInvalidAnchorNameError whether err is ast.ErrInvalidAnchorName or not.
+func IsInvalidAnchorNameError(err error) bool {
+	return xerrors.Is(err, ast.ErrInvalidAnchorName)
+}
+
+// IsInvalidAliasNameError whether err is ast.ErrInvalidAliasName or not.
+func IsInvalidAliasNameError(err error) bool {
+	return xerrors.Is(err, ast.ErrInvalidAliasName)
+}

vendor/github.com/goccy/go-yaml/internal/errors/error.go

@@ -0,0 +1,260 @@
+package errors
+
+import (
+	"bytes"
+	"fmt"
+	"reflect"
+
+	"github.com/goccy/go-yaml/printer"
+	"github.com/goccy/go-yaml/token"
+	"golang.org/x/xerrors"
+)
+
+const (
+	defaultColorize      = false
+	defaultIncludeSource = true
+)
+
+var (
+	ErrDecodeRequiredPointerType = xerrors.New("required pointer type value")
+)
+
+// Wrapf wrap error for stack trace
+func Wrapf(err error, msg string, args ...interface{}) error {
+	return &wrapError{
+		baseError: &baseError{},
+		err:       xerrors.Errorf(msg, args...),
+		nextErr:   err,
+		frame:     xerrors.Caller(1),
+	}
+}
+
+// ErrSyntax create syntax error instance with message and token
+func ErrSyntax(msg string, tk *token.Token) *syntaxError {
+	return &syntaxError{
+		baseError: &baseError{},
+		msg:       msg,
+		token:     tk,
+		frame:     xerrors.Caller(1),
+	}
+}
+
+type baseError struct {
+	state fmt.State
+	verb  rune
+}
+
+func (e *baseError) Error() string {
+	return ""
+}
+
+func (e *baseError) chainStateAndVerb(err error) {
+	wrapErr, ok := err.(*wrapError)
+	if ok {
+		wrapErr.state = e.state
+		wrapErr.verb = e.verb
+	}
+	syntaxErr, ok := err.(*syntaxError)
+	if ok {
+		syntaxErr.state = e.state
+		syntaxErr.verb = e.verb
+	}
+}
+
+type wrapError struct {
+	*baseError
+	err     error
+	nextErr error
+	frame   xerrors.Frame
+}
+
+type FormatErrorPrinter struct {
+	xerrors.Printer
+	Colored    bool
+	InclSource bool
+}
+
+func (e *wrapError) As(target interface{}) bool {
+	err := e.nextErr
+	for {
+		if wrapErr, ok := err.(*wrapError); ok {
+			err = wrapErr.nextErr
+			continue
+		}
+		break
+	}
+	return xerrors.As(err, target)
+}
+
+func (e *wrapError) Unwrap() error {
+	return e.nextErr
+}
+
+func (e *wrapError) PrettyPrint(p xerrors.Printer, colored, inclSource bool) error {
+	return e.FormatError(&FormatErrorPrinter{Printer: p, Colored: colored, InclSource: inclSource})
+}
+
+func (e *wrapError) FormatError(p xerrors.Printer) error {
+	if _, ok := p.(*FormatErrorPrinter); !ok {
+		p = &FormatErrorPrinter{
+			Printer:    p,
+			Colored:    defaultColorize,
+			InclSource: defaultIncludeSource,
+		}
+	}
+	if e.verb == 'v' && e.state.Flag('+') {
+		// print stack trace for debugging
+		p.Print(e.err, "\n")
+		e.frame.Format(p)
+		e.chainStateAndVerb(e.nextErr)
+		return e.nextErr
+	}
+	err := e.nextErr
+	for {
+		if wrapErr, ok := err.(*wrapError); ok {
+			err = wrapErr.nextErr
+			continue
+		}
+		break
+	}
+	e.chainStateAndVerb(err)
+	if fmtErr, ok := err.(xerrors.Formatter); ok {
+		fmtErr.FormatError(p)
+	} else {
+		p.Print(err)
+	}
+	return nil
+}
+
+type wrapState struct {
+	org fmt.State
+}
+
+func (s *wrapState) Write(b []byte) (n int, err error) {
+	return s.org.Write(b)
+}
+
+func (s *wrapState) Width() (wid int, ok bool) {
+	return s.org.Width()
+}
+
+func (s *wrapState) Precision() (prec int, ok bool) {
+	return s.org.Precision()
+}
+
+func (s *wrapState) Flag(c int) bool {
+	// set true to 'printDetail' forced because when p.Detail() is false, xerrors.Printer no output any text
+	if c == '#' {
+		// ignore '#' keyword because xerrors.FormatError doesn't set true to printDetail.
+		// ( see https://github.com/golang/xerrors/blob/master/adaptor.go#L39-L43 )
+		return false
+	}
+	return true
+}
+
+func (e *wrapError) Format(state fmt.State, verb rune) {
+	e.state = state
+	e.verb = verb
+	xerrors.FormatError(e, &wrapState{org: state}, verb)
+}
+
+func (e *wrapError) Error() string {
+	var buf bytes.Buffer
+	e.PrettyPrint(&Sink{&buf}, defaultColorize, defaultIncludeSource)
+	return buf.String()
+}
+
+type syntaxError struct {
+	*baseError
+	msg   string
+	token *token.Token
+	frame xerrors.Frame
+}
+
+func (e *syntaxError) PrettyPrint(p xerrors.Printer, colored, inclSource bool) error {
+	return e.FormatError(&FormatErrorPrinter{Printer: p, Colored: colored, InclSource: inclSource})
+}
+
+func (e *syntaxError) FormatError(p xerrors.Printer) error {
+	var pp printer.Printer
+
+	var colored, inclSource bool
+	if fep, ok := p.(*FormatErrorPrinter); ok {
+		colored = fep.Colored
+		inclSource = fep.InclSource
+	}
+
+	pos := fmt.Sprintf("[%d:%d] ", e.token.Position.Line, e.token.Position.Column)
+	msg := pp.PrintErrorMessage(fmt.Sprintf("%s%s", pos, e.msg), colored)
+	if inclSource {
+		msg += "\n" + pp.PrintErrorToken(e.token, colored)
+	}
+	p.Print(msg)
+
+	if e.verb == 'v' && e.state.Flag('+') {
+		// %+v
+		// print stack trace for debugging
+		e.frame.Format(p)
+	}
+	return nil
+}
+
+type PrettyPrinter interface {
+	PrettyPrint(xerrors.Printer, bool, bool) error
+}
+type Sink struct{ *bytes.Buffer }
+
+func (es *Sink) Print(args ...interface{}) {
+	fmt.Fprint(es.Buffer, args...)
+}
+
+func (es *Sink) Printf(f string, args ...interface{}) {
+	fmt.Fprintf(es.Buffer, f, args...)
+}
+
+func (es *Sink) Detail() bool {
+	return false
+}
+
+func (e *syntaxError) Error() string {
+	var buf bytes.Buffer
+	e.PrettyPrint(&Sink{&buf}, defaultColorize, defaultIncludeSource)
+	return buf.String()
+}
+
+type TypeError struct {
+	DstType         reflect.Type
+	SrcType         reflect.Type
+	StructFieldName *string
+	Token           *token.Token
+}
+
+func (e *TypeError) Error() string {
+	if e.StructFieldName != nil {
+		return fmt.Sprintf("cannot unmarshal %s into Go struct field %s of type %s", e.SrcType, *e.StructFieldName, e.DstType)
+	}
+	return fmt.Sprintf("cannot unmarshal %s into Go value of type %s", e.SrcType, e.DstType)
+}
+
+func (e *TypeError) PrettyPrint(p xerrors.Printer, colored, inclSource bool) error {
+	return e.FormatError(&FormatErrorPrinter{Printer: p, Colored: colored, InclSource: inclSource})
+}
+
+func (e *TypeError) FormatError(p xerrors.Printer) error {
+	var pp printer.Printer
+
+	var colored, inclSource bool
+	if fep, ok := p.(*FormatErrorPrinter); ok {
+		colored = fep.Colored
+		inclSource = fep.InclSource
+	}
+
+	pos := fmt.Sprintf("[%d:%d] ", e.Token.Position.Line, e.Token.Position.Column)
+	msg := pp.PrintErrorMessage(fmt.Sprintf("%s%s", pos, e.Error()), colored)
+	if inclSource {
+		msg += "\n" + pp.PrintErrorToken(e.Token, colored)
+	}
+	p.Print(msg)
+
+	return nil
+}

vendor/github.com/goccy/go-yaml/lexer/lexer.go

@@ -0,0 +1,23 @@
+package lexer
+
+import (
+	"io"
+
+	"github.com/goccy/go-yaml/scanner"
+	"github.com/goccy/go-yaml/token"
+)
+
+// Tokenize split to token instances from string
+func Tokenize(src string) token.Tokens {
+	var s scanner.Scanner
+	s.Init(src)
+	var tokens token.Tokens
+	for {
+		subTokens, err := s.Scan()
+		if err == io.EOF {
+			break
+		}
+		tokens.Add(subTokens...)
+	}
+	return tokens
+}

vendor/github.com/goccy/go-yaml/option.go

@@ -0,0 +1,278 @@
+package yaml
+
+import (
+	"io"
+	"reflect"
+
+	"github.com/goccy/go-yaml/ast"
+)
+
+// DecodeOption functional option type for Decoder
+type DecodeOption func(d *Decoder) error
+
+// ReferenceReaders pass to Decoder that reference to anchor defined by passed readers
+func ReferenceReaders(readers ...io.Reader) DecodeOption {
+	return func(d *Decoder) error {
+		d.referenceReaders = append(d.referenceReaders, readers...)
+		return nil
+	}
+}
+
+// ReferenceFiles pass to Decoder that reference to anchor defined by passed files
+func ReferenceFiles(files ...string) DecodeOption {
+	return func(d *Decoder) error {
+		d.referenceFiles = files
+		return nil
+	}
+}
+
+// ReferenceDirs pass to Decoder that reference to anchor defined by files under the passed dirs
+func ReferenceDirs(dirs ...string) DecodeOption {
+	return func(d *Decoder) error {
+		d.referenceDirs = dirs
+		return nil
+	}
+}
+
+// RecursiveDir search yaml file recursively from passed dirs by ReferenceDirs option
+func RecursiveDir(isRecursive bool) DecodeOption {
+	return func(d *Decoder) error {
+		d.isRecursiveDir = isRecursive
+		return nil
+	}
+}
+
+// Validator set StructValidator instance to Decoder
+func Validator(v StructValidator) DecodeOption {
+	return func(d *Decoder) error {
+		d.validator = v
+		return nil
+	}
+}
+
+// Strict enable DisallowUnknownField and DisallowDuplicateKey
+func Strict() DecodeOption {
+	return func(d *Decoder) error {
+		d.disallowUnknownField = true
+		d.disallowDuplicateKey = true
+		return nil
+	}
+}
+
+// DisallowUnknownField causes the Decoder to return an error when the destination
+// is a struct and the input contains object keys which do not match any
+// non-ignored, exported fields in the destination.
+func DisallowUnknownField() DecodeOption {
+	return func(d *Decoder) error {
+		d.disallowUnknownField = true
+		return nil
+	}
+}
+
+// DisallowDuplicateKey causes an error when mapping keys that are duplicates
+func DisallowDuplicateKey() DecodeOption {
+	return func(d *Decoder) error {
+		d.disallowDuplicateKey = true
+		return nil
+	}
+}
+
+// UseOrderedMap can be interpreted as a map,
+// and uses MapSlice ( ordered map ) aggressively if there is no type specification
+func UseOrderedMap() DecodeOption {
+	return func(d *Decoder) error {
+		d.useOrderedMap = true
+		return nil
+	}
+}
+
+// UseJSONUnmarshaler if neither `BytesUnmarshaler` nor `InterfaceUnmarshaler` is implemented
+// and `UnmashalJSON([]byte)error` is implemented, convert the argument from `YAML` to `JSON` and then call it.
+func UseJSONUnmarshaler() DecodeOption {
+	return func(d *Decoder) error {
+		d.useJSONUnmarshaler = true
+		return nil
+	}
+}
+
+// CustomUnmarshaler overrides any decoding process for the type specified in generics.
+//
+// NOTE: If RegisterCustomUnmarshaler and CustomUnmarshaler of DecodeOption are specified for the same type,
+// the CustomUnmarshaler specified in DecodeOption takes precedence.
+func CustomUnmarshaler[T any](unmarshaler func(*T, []byte) error) DecodeOption {
+	return func(d *Decoder) error {
+		var typ *T
+		d.customUnmarshalerMap[reflect.TypeOf(typ)] = func(v interface{}, b []byte) error {
+			return unmarshaler(v.(*T), b)
+		}
+		return nil
+	}
+}
+
+// EncodeOption functional option type for Encoder
+type EncodeOption func(e *Encoder) error
+
+// Indent change indent number
+func Indent(spaces int) EncodeOption {
+	return func(e *Encoder) error {
+		e.indent = spaces
+		return nil
+	}
+}
+
+// IndentSequence causes sequence values to be indented the same value as Indent
+func IndentSequence(indent bool) EncodeOption {
+	return func(e *Encoder) error {
+		e.indentSequence = indent
+		return nil
+	}
+}
+
+// UseSingleQuote determines if single or double quotes should be preferred for strings.
+func UseSingleQuote(sq bool) EncodeOption {
+	return func(e *Encoder) error {
+		e.singleQuote = sq
+		return nil
+	}
+}
+
+// Flow encoding by flow style
+func Flow(isFlowStyle bool) EncodeOption {
+	return func(e *Encoder) error {
+		e.isFlowStyle = isFlowStyle
+		return nil
+	}
+}
+
+// UseLiteralStyleIfMultiline causes encoding multiline strings with a literal syntax,
+// no matter what characters they include
+func UseLiteralStyleIfMultiline(useLiteralStyleIfMultiline bool) EncodeOption {
+	return func(e *Encoder) error {
+		e.useLiteralStyleIfMultiline = useLiteralStyleIfMultiline
+		return nil
+	}
+}
+
+// JSON encode in JSON format
+func JSON() EncodeOption {
+	return func(e *Encoder) error {
+		e.isJSONStyle = true
+		e.isFlowStyle = true
+		return nil
+	}
+}
+
+// MarshalAnchor call back if encoder find an anchor during encoding
+func MarshalAnchor(callback func(*ast.AnchorNode, interface{}) error) EncodeOption {
+	return func(e *Encoder) error {
+		e.anchorCallback = callback
+		return nil
+	}
+}
+
+// UseJSONMarshaler if neither `BytesMarshaler` nor `InterfaceMarshaler`
+// nor `encoding.TextMarshaler` is implemented and `MarshalJSON()([]byte, error)` is implemented,
+// call `MarshalJSON` to convert the returned `JSON` to `YAML` for processing.
+func UseJSONMarshaler() EncodeOption {
+	return func(e *Encoder) error {
+		e.useJSONMarshaler = true
+		return nil
+	}
+}
+
+// CustomMarshaler overrides any encoding process for the type specified in generics.
+//
+// NOTE: If type T implements MarshalYAML for pointer receiver, the type specified in CustomMarshaler must be *T.
+// If RegisterCustomMarshaler and CustomMarshaler of EncodeOption are specified for the same type,
+// the CustomMarshaler specified in EncodeOption takes precedence.
+func CustomMarshaler[T any](marshaler func(T) ([]byte, error)) EncodeOption {
+	return func(e *Encoder) error {
+		var typ T
+		e.customMarshalerMap[reflect.TypeOf(typ)] = func(v interface{}) ([]byte, error) {
+			return marshaler(v.(T))
+		}
+		return nil
+	}
+}
+
+// CommentPosition type of the position for comment.
+type CommentPosition int
+
+const (
+	CommentHeadPosition CommentPosition = CommentPosition(iota)
+	CommentLinePosition
+	CommentFootPosition
+)
+
+func (p CommentPosition) String() string {
+	switch p {
+	case CommentHeadPosition:
+		return "Head"
+	case CommentLinePosition:
+		return "Line"
+	case CommentFootPosition:
+		return "Foot"
+	default:
+		return ""
+	}
+}
+
+// LineComment create a one-line comment for CommentMap.
+func LineComment(text string) *Comment {
+	return &Comment{
+		Texts:    []string{text},
+		Position: CommentLinePosition,
+	}
+}
+
+// HeadComment create a multiline comment for CommentMap.
+func HeadComment(texts ...string) *Comment {
+	return &Comment{
+		Texts:    texts,
+		Position: CommentHeadPosition,
+	}
+}
+
+// FootComment create a multiline comment for CommentMap.
+func FootComment(texts ...string) *Comment {
+	return &Comment{
+		Texts:    texts,
+		Position: CommentFootPosition,
+	}
+}
+
+// Comment raw data for comment.
+type Comment struct {
+	Texts    []string
+	Position CommentPosition
+}
+
+// CommentMap map of the position of the comment and the comment information.
+type CommentMap map[string][]*Comment
+
+// WithComment add a comment using the location and text information given in the CommentMap.
+func WithComment(cm CommentMap) EncodeOption {
+	return func(e *Encoder) error {
+		commentMap := map[*Path][]*Comment{}
+		for k, v := range cm {
+			path, err := PathString(k)
+			if err != nil {
+				return err
+			}
+			commentMap[path] = v
+		}
+		e.commentMap = commentMap
+		return nil
+	}
+}
+
+// CommentToMap apply the position and content of comments in a YAML document to a CommentMap.
+func CommentToMap(cm CommentMap) DecodeOption {
+	return func(d *Decoder) error {
+		if cm == nil {
+			return ErrInvalidCommentMapValue
+		}
+		d.toCommentMap = cm
+		return nil
+	}
+}

vendor/github.com/goccy/go-yaml/parser/context.go

@@ -0,0 +1,199 @@
+package parser
+
+import (
+	"fmt"
+	"strings"
+
+	"github.com/goccy/go-yaml/token"
+)
+
+// context context at parsing
+type context struct {
+	parent *context
+	idx    int
+	size   int
+	tokens token.Tokens
+	mode   Mode
+	path   string
+}
+
+var pathSpecialChars = []string{
+	"$", "*", ".", "[", "]",
+}
+
+func containsPathSpecialChar(path string) bool {
+	for _, char := range pathSpecialChars {
+		if strings.Contains(path, char) {
+			return true
+		}
+	}
+	return false
+}
+
+func normalizePath(path string) string {
+	if containsPathSpecialChar(path) {
+		return fmt.Sprintf("'%s'", path)
+	}
+	return path
+}
+
+func (c *context) withChild(path string) *context {
+	ctx := c.copy()
+	path = normalizePath(path)
+	ctx.path += fmt.Sprintf(".%s", path)
+	return ctx
+}
+
+func (c *context) withIndex(idx uint) *context {
+	ctx := c.copy()
+	ctx.path += fmt.Sprintf("[%d]", idx)
+	return ctx
+}
+
+func (c *context) copy() *context {
+	return &context{
+		parent: c,
+		idx:    c.idx,
+		size:   c.size,
+		tokens: append(token.Tokens{}, c.tokens...),
+		mode:   c.mode,
+		path:   c.path,
+	}
+}
+
+func (c *context) next() bool {
+	return c.idx < c.size
+}
+
+func (c *context) previousToken() *token.Token {
+	if c.idx > 0 {
+		return c.tokens[c.idx-1]
+	}
+	return nil
+}
+
+func (c *context) insertToken(idx int, tk *token.Token) {
+	if c.parent != nil {
+		c.parent.insertToken(idx, tk)
+	}
+	if c.size < idx {
+		return
+	}
+	if c.size == idx {
+		curToken := c.tokens[c.size-1]
+		tk.Next = curToken
+		curToken.Prev = tk
+
+		c.tokens = append(c.tokens, tk)
+		c.size = len(c.tokens)
+		return
+	}
+
+	curToken := c.tokens[idx]
+	tk.Next = curToken
+	curToken.Prev = tk
+
+	c.tokens = append(c.tokens[:idx+1], c.tokens[idx:]...)
+	c.tokens[idx] = tk
+	c.size = len(c.tokens)
+}
+
+func (c *context) currentToken() *token.Token {
+	if c.idx >= c.size {
+		return nil
+	}
+	return c.tokens[c.idx]
+}
+
+func (c *context) nextToken() *token.Token {
+	if c.idx+1 >= c.size {
+		return nil
+	}
+	return c.tokens[c.idx+1]
+}
+
+func (c *context) afterNextToken() *token.Token {
+	if c.idx+2 >= c.size {
+		return nil
+	}
+	return c.tokens[c.idx+2]
+}
+
+func (c *context) nextNotCommentToken() *token.Token {
+	for i := c.idx + 1; i < c.size; i++ {
+		tk := c.tokens[i]
+		if tk.Type == token.CommentType {
+			continue
+		}
+		return tk
+	}
+	return nil
+}
+
+func (c *context) afterNextNotCommentToken() *token.Token {
+	notCommentTokenCount := 0
+	for i := c.idx + 1; i < c.size; i++ {
+		tk := c.tokens[i]
+		if tk.Type == token.CommentType {
+			continue
+		}
+		notCommentTokenCount++
+		if notCommentTokenCount == 2 {
+			return tk
+		}
+	}
+	return nil
+}
+
+func (c *context) enabledComment() bool {
+	return c.mode&ParseComments != 0
+}
+
+func (c *context) isCurrentCommentToken() bool {
+	tk := c.currentToken()
+	if tk == nil {
+		return false
+	}
+	return tk.Type == token.CommentType
+}
+
+func (c *context) progressIgnoreComment(num int) {
+	if c.parent != nil {
+		c.parent.progressIgnoreComment(num)
+	}
+	if c.size <= c.idx+num {
+		c.idx = c.size
+	} else {
+		c.idx += num
+	}
+}
+
+func (c *context) progress(num int) {
+	if c.isCurrentCommentToken() {
+		return
+	}
+	c.progressIgnoreComment(num)
+}
+
+func newContext(tokens token.Tokens, mode Mode) *context {
+	filteredTokens := []*token.Token{}
+	if mode&ParseComments != 0 {
+		filteredTokens = tokens
+	} else {
+		for _, tk := range tokens {
+			if tk.Type == token.CommentType {
+				continue
+			}
+			// keep prev/next reference between tokens containing comments
+			// https://github.com/goccy/go-yaml/issues/254
+			filteredTokens = append(filteredTokens, tk)
+		}
+	}
+	return &context{
+		idx:    0,
+		size:   len(filteredTokens),
+		tokens: token.Tokens(filteredTokens),
+		mode:   mode,
+		path:   "$",
+	}
+}

vendor/github.com/goccy/go-yaml/parser/parser.go

@@ -0,0 +1,714 @@
+package parser
+
+import (
+	"fmt"
+	"io/ioutil"
+	"strings"
+
+	"github.com/goccy/go-yaml/ast"
+	"github.com/goccy/go-yaml/internal/errors"
+	"github.com/goccy/go-yaml/lexer"
+	"github.com/goccy/go-yaml/token"
+	"golang.org/x/xerrors"
+)
+
+type parser struct{}
+
+func (p *parser) parseMapping(ctx *context) (*ast.MappingNode, error) {
+	mapTk := ctx.currentToken()
+	node := ast.Mapping(mapTk, true)
+	node.SetPath(ctx.path)
+	ctx.progress(1) // skip MappingStart token
+	for ctx.next() {
+		tk := ctx.currentToken()
+		if tk.Type == token.MappingEndType {
+			node.End = tk
+			return node, nil
+		} else if tk.Type == token.CollectEntryType {
+			ctx.progress(1)
+			continue
+		}
+
+		value, err := p.parseMappingValue(ctx)
+		if err != nil {
+			return nil, errors.Wrapf(err, "failed to parse mapping value in mapping node")
+		}
+		mvnode, ok := value.(*ast.MappingValueNode)
+		if !ok {
+			return nil, errors.ErrSyntax("failed to parse flow mapping node", value.GetToken())
+		}
+		node.Values = append(node.Values, mvnode)
+		ctx.progress(1)
+	}
+	return nil, errors.ErrSyntax("unterminated flow mapping", node.GetToken())
+}
+
+func (p *parser) parseSequence(ctx *context) (*ast.SequenceNode, error) {
+	node := ast.Sequence(ctx.currentToken(), true)
+	node.SetPath(ctx.path)
+	ctx.progress(1) // skip SequenceStart token
+	for ctx.next() {
+		tk := ctx.currentToken()
+		if tk.Type == token.SequenceEndType {
+			node.End = tk
+			break
+		} else if tk.Type == token.CollectEntryType {
+			ctx.progress(1)
+			continue
+		}
+
+		value, err := p.parseToken(ctx.withIndex(uint(len(node.Values))), tk)
+		if err != nil {
+			return nil, errors.Wrapf(err, "failed to parse sequence value in flow sequence node")
+		}
+		node.Values = append(node.Values, value)
+		ctx.progress(1)
+	}
+	return node, nil
+}
+
+func (p *parser) parseTag(ctx *context) (*ast.TagNode, error) {
+	tagToken := ctx.currentToken()
+	node := ast.Tag(tagToken)
+	node.SetPath(ctx.path)
+	ctx.progress(1) // skip tag token
+	var (
+		value ast.Node
+		err   error
+	)
+	switch token.ReservedTagKeyword(tagToken.Value) {
+	case token.MappingTag,
+		token.OrderedMapTag:
+		value, err = p.parseMapping(ctx)
+	case token.IntegerTag,
+		token.FloatTag,
+		token.StringTag,
+		token.BinaryTag,
+		token.TimestampTag,
+		token.NullTag:
+		typ := ctx.currentToken().Type
+		if typ == token.LiteralType || typ == token.FoldedType {
+			value, err = p.parseLiteral(ctx)
+		} else {
+			value = p.parseScalarValue(ctx.currentToken())
+		}
+	case token.SequenceTag,
+		token.SetTag:
+		err = errors.ErrSyntax(fmt.Sprintf("sorry, currently not supported %s tag", tagToken.Value), tagToken)
+	default:
+		// custom tag
+		value, err = p.parseToken(ctx, ctx.currentToken())
+	}
+	if err != nil {
+		return nil, errors.Wrapf(err, "failed to parse tag value")
+	}
+	node.Value = value
+	return node, nil
+}
+
+func (p *parser) removeLeftSideNewLineCharacter(src string) string {
+	// CR or LF or CRLF
+	return strings.TrimLeft(strings.TrimLeft(strings.TrimLeft(src, "\r"), "\n"), "\r\n")
+}
+
+func (p *parser) existsNewLineCharacter(src string) bool {
+	if strings.Index(src, "\n") > 0 {
+		return true
+	}
+	if strings.Index(src, "\r") > 0 {
+		return true
+	}
+	return false
+}
+
+func (p *parser) validateMapKey(tk *token.Token) error {
+	if tk.Type != token.StringType {
+		return nil
+	}
+	origin := p.removeLeftSideNewLineCharacter(tk.Origin)
+	if p.existsNewLineCharacter(origin) {
+		return errors.ErrSyntax("unexpected key name", tk)
+	}
+	return nil
+}
+
+func (p *parser) createNullToken(base *token.Token) *token.Token {
+	pos := *(base.Position)
+	pos.Column++
+	return token.New("null", "null", &pos)
+}
+
+func (p *parser) parseMapValue(ctx *context, key ast.MapKeyNode, colonToken *token.Token) (ast.Node, error) {
+	node, err := p.createMapValueNode(ctx, key, colonToken)
+	if err != nil {
+		return nil, errors.Wrapf(err, "failed to create map value node")
+	}
+	if node != nil && node.GetPath() == "" {
+		node.SetPath(ctx.path)
+	}
+	return node, nil
+}
+
+func (p *parser) createMapValueNode(ctx *context, key ast.MapKeyNode, colonToken *token.Token) (ast.Node, error) {
+	tk := ctx.currentToken()
+	if tk == nil {
+		nullToken := p.createNullToken(colonToken)
+		ctx.insertToken(ctx.idx, nullToken)
+		return ast.Null(nullToken), nil
+	}
+
+	if tk.Position.Column == key.GetToken().Position.Column && tk.Type == token.StringType {
+		// in this case,
+		// ----
+		// key: <value does not defined>
+		// next
+		nullToken := p.createNullToken(colonToken)
+		ctx.insertToken(ctx.idx, nullToken)
+		return ast.Null(nullToken), nil
+	}
+
+	if tk.Position.Column < key.GetToken().Position.Column {
+		// in this case,
+		// ----
+		//   key: <value does not defined>
+		// next
+		nullToken := p.createNullToken(colonToken)
+		ctx.insertToken(ctx.idx, nullToken)
+		return ast.Null(nullToken), nil
+	}
+
+	value, err := p.parseToken(ctx, ctx.currentToken())
+	if err != nil {
+		return nil, errors.Wrapf(err, "failed to parse mapping 'value' node")
+	}
+	return value, nil
+}
+
+func (p *parser) validateMapValue(ctx *context, key, value ast.Node) error {
+	keyColumn := key.GetToken().Position.Column
+	valueColumn := value.GetToken().Position.Column
+	if keyColumn != valueColumn {
+		return nil
+	}
+	if value.Type() != ast.StringType {
+		return nil
+	}
+	ntk := ctx.nextToken()
+	if ntk == nil || (ntk.Type != token.MappingValueType && ntk.Type != token.SequenceEntryType) {
+		return errors.ErrSyntax("could not found expected ':' token", value.GetToken())
+	}
+	return nil
+}
+
+func (p *parser) parseMappingValue(ctx *context) (ast.Node, error) {
+	key, err := p.parseMapKey(ctx)
+	if err != nil {
+		return nil, errors.Wrapf(err, "failed to parse map key")
+	}
+	keyText := key.GetToken().Value
+	key.SetPath(ctx.withChild(keyText).path)
+	if err := p.validateMapKey(key.GetToken()); err != nil {
+		return nil, errors.Wrapf(err, "validate mapping key error")
+	}
+	ctx.progress(1)          // progress to mapping value token
+	tk := ctx.currentToken() // get mapping value token
+	if tk == nil {
+		return nil, errors.ErrSyntax("unexpected map", key.GetToken())
+	}
+	ctx.progress(1) // progress to value token
+	if err := p.setSameLineCommentIfExists(ctx.withChild(keyText), key); err != nil {
+		return nil, errors.Wrapf(err, "failed to set same line comment to node")
+	}
+	if key.GetComment() != nil {
+		// if current token is comment, GetComment() is not nil.
+		// then progress to value token
+		ctx.progressIgnoreComment(1)
+	}
+
+	value, err := p.parseMapValue(ctx.withChild(keyText), key, tk)
+	if err != nil {
+		return nil, errors.Wrapf(err, "failed to parse map value")
+	}
+	if err := p.validateMapValue(ctx, key, value); err != nil {
+		return nil, errors.Wrapf(err, "failed to validate map value")
+	}
+
+	mvnode := ast.MappingValue(tk, key, value)
+	mvnode.SetPath(ctx.withChild(keyText).path)
+	node := ast.Mapping(tk, false, mvnode)
+	node.SetPath(ctx.withChild(keyText).path)
+
+	ntk := ctx.nextNotCommentToken()
+	antk := ctx.afterNextNotCommentToken()
+	for antk != nil && antk.Type == token.MappingValueType &&
+		ntk.Position.Column == key.GetToken().Position.Column {
+		ctx.progressIgnoreComment(1)
+		value, err := p.parseToken(ctx, ctx.currentToken())
+		if err != nil {
+			return nil, errors.Wrapf(err, "failed to parse mapping node")
+		}
+		switch value.Type() {
+		case ast.MappingType:
+			c := value.(*ast.MappingNode)
+			comment := c.GetComment()
+			for idx, v := range c.Values {
+				if idx == 0 && comment != nil {
+					if err := v.SetComment(comment); err != nil {
+						return nil, errors.Wrapf(err, "failed to set comment token to node")
+					}
+				}
+				node.Values = append(node.Values, v)
+			}
+		case ast.MappingValueType:
+			node.Values = append(node.Values, value.(*ast.MappingValueNode))
+		default:
+			return nil, xerrors.Errorf("failed to parse mapping value node node is %s", value.Type())
+		}
+		ntk = ctx.nextNotCommentToken()
+		antk = ctx.afterNextNotCommentToken()
+	}
+	if len(node.Values) == 1 {
+		mapKeyCol := mvnode.Key.GetToken().Position.Column
+		commentTk := ctx.nextToken()
+		if commentTk != nil && commentTk.Type == token.CommentType && mapKeyCol <= commentTk.Position.Column {
+			// If the comment is in the same or deeper column as the last element column in map value,
+			// treat it as a footer comment for the last element.
+			comment := p.parseFootComment(ctx, mapKeyCol)
+			mvnode.FootComment = comment
+		}
+		return mvnode, nil
+	}
+	mapCol := node.GetToken().Position.Column
+	commentTk := ctx.nextToken()
+	if commentTk != nil && commentTk.Type == token.CommentType && mapCol <= commentTk.Position.Column {
+		// If the comment is in the same or deeper column as the last element column in map value,
+		// treat it as a footer comment for the last element.
+		comment := p.parseFootComment(ctx, mapCol)
+		node.FootComment = comment
+	}
+	return node, nil
+}
+
+func (p *parser) parseSequenceEntry(ctx *context) (*ast.SequenceNode, error) {
+	tk := ctx.currentToken()
+	sequenceNode := ast.Sequence(tk, false)
+	sequenceNode.SetPath(ctx.path)
+	curColumn := tk.Position.Column
+	for tk.Type == token.SequenceEntryType {
+		ctx.progress(1) // skip sequence token
+		tk = ctx.currentToken()
+		if tk == nil {
+			return nil, errors.ErrSyntax("empty sequence entry", ctx.previousToken())
+		}
+		var comment *ast.CommentGroupNode
+		if tk.Type == token.CommentType {
+			comment = p.parseCommentOnly(ctx)
+			tk = ctx.currentToken()
+			if tk.Type != token.SequenceEntryType {
+				break
+			}
+			ctx.progress(1) // skip sequence token
+		}
+		value, err := p.parseToken(ctx.withIndex(uint(len(sequenceNode.Values))), ctx.currentToken())
+		if err != nil {
+			return nil, errors.Wrapf(err, "failed to parse sequence")
+		}
+		if comment != nil {
+			comment.SetPath(ctx.withIndex(uint(len(sequenceNode.Values))).path)
+			sequenceNode.ValueHeadComments = append(sequenceNode.ValueHeadComments, comment)
+		} else {
+			sequenceNode.ValueHeadComments = append(sequenceNode.ValueHeadComments, nil)
+		}
+		sequenceNode.Values = append(sequenceNode.Values, value)
+		tk = ctx.nextNotCommentToken()
+		if tk == nil {
+			break
+		}
+		if tk.Type != token.SequenceEntryType {
+			break
+		}
+		if tk.Position.Column != curColumn {
+			break
+		}
+		ctx.progressIgnoreComment(1)
+	}
+	commentTk := ctx.nextToken()
+	if commentTk != nil && commentTk.Type == token.CommentType && curColumn <= commentTk.Position.Column {
+		// If the comment is in the same or deeper column as the last element column in sequence value,
+		// treat it as a footer comment for the last element.
+		comment := p.parseFootComment(ctx, curColumn)
+		sequenceNode.FootComment = comment
+	}
+	return sequenceNode, nil
+}
+
+func (p *parser) parseAnchor(ctx *context) (*ast.AnchorNode, error) {
+	tk := ctx.currentToken()
+	anchor := ast.Anchor(tk)
+	anchor.SetPath(ctx.path)
+	ntk := ctx.nextToken()
+	if ntk == nil {
+		return nil, errors.ErrSyntax("unexpected anchor. anchor name is undefined", tk)
+	}
+	ctx.progress(1) // skip anchor token
+	name, err := p.parseToken(ctx, ctx.currentToken())
+	if err != nil {
+		return nil, errors.Wrapf(err, "failed to parser anchor name node")
+	}
+	anchor.Name = name
+	ntk = ctx.nextToken()
+	if ntk == nil {
+		return nil, errors.ErrSyntax("unexpected anchor. anchor value is undefined", ctx.currentToken())
+	}
+	ctx.progress(1)
+	value, err := p.parseToken(ctx, ctx.currentToken())
+	if err != nil {
+		return nil, errors.Wrapf(err, "failed to parser anchor name node")
+	}
+	anchor.Value = value
+	return anchor, nil
+}
+
+func (p *parser) parseAlias(ctx *context) (*ast.AliasNode, error) {
+	tk := ctx.currentToken()
+	alias := ast.Alias(tk)
+	alias.SetPath(ctx.path)
+	ntk := ctx.nextToken()
+	if ntk == nil {
+		return nil, errors.ErrSyntax("unexpected alias. alias name is undefined", tk)
+	}
+	ctx.progress(1) // skip alias token
+	name, err := p.parseToken(ctx, ctx.currentToken())
+	if err != nil {
+		return nil, errors.Wrapf(err, "failed to parser alias name node")
+	}
+	alias.Value = name
+	return alias, nil
+}
+
+func (p *parser) parseMapKey(ctx *context) (ast.MapKeyNode, error) {
+	tk := ctx.currentToken()
+	if value := p.parseScalarValue(tk); value != nil {
+		return value, nil
+	}
+	switch tk.Type {
+	case token.MergeKeyType:
+		return ast.MergeKey(tk), nil
+	case token.MappingKeyType:
+		return p.parseMappingKey(ctx)
+	}
+	return nil, errors.ErrSyntax("unexpected mapping key", tk)
+}
+
+func (p *parser) parseStringValue(tk *token.Token) *ast.StringNode {
+	switch tk.Type {
+	case token.StringType,
+		token.SingleQuoteType,
+		token.DoubleQuoteType:
+		return ast.String(tk)
+	}
+	return nil
+}
+
+func (p *parser) parseScalarValueWithComment(ctx *context, tk *token.Token) (ast.ScalarNode, error) {
+	node := p.parseScalarValue(tk)
+	if node == nil {
+		return nil, nil
+	}
+	node.SetPath(ctx.path)
+	if p.isSameLineComment(ctx.nextToken(), node) {
+		ctx.progress(1)
+		if err := p.setSameLineCommentIfExists(ctx, node); err != nil {
+			return nil, errors.Wrapf(err, "failed to set same line comment to node")
+		}
+	}
+	return node, nil
+}
+
+func (p *parser) parseScalarValue(tk *token.Token) ast.ScalarNode {
+	if node := p.parseStringValue(tk); node != nil {
+		return node
+	}
+	switch tk.Type {
+	case token.NullType:
+		return ast.Null(tk)
+	case token.BoolType:
+		return ast.Bool(tk)
+	case token.IntegerType,
+		token.BinaryIntegerType,
+		token.OctetIntegerType,
+		token.HexIntegerType:
+		return ast.Integer(tk)
+	case token.FloatType:
+		return ast.Float(tk)
+	case token.InfinityType:
+		return ast.Infinity(tk)
+	case token.NanType:
+		return ast.Nan(tk)
+	}
+	return nil
+}
+
+func (p *parser) parseDirective(ctx *context) (*ast.DirectiveNode, error) {
+	node := ast.Directive(ctx.currentToken())
+	ctx.progress(1) // skip directive token
+	value, err := p.parseToken(ctx, ctx.currentToken())
+	if err != nil {
+		return nil, errors.Wrapf(err, "failed to parse directive value")
+	}
+	node.Value = value
+	ctx.progress(1)
+	tk := ctx.currentToken()
+	if tk == nil {
+		// Since current token is nil, use the previous token to specify
+		// the syntax error location.
+		return nil, errors.ErrSyntax("unexpected directive value. document not started", ctx.previousToken())
+	}
+	if tk.Type != token.DocumentHeaderType {
+		return nil, errors.ErrSyntax("unexpected directive value. document not started", ctx.currentToken())
+	}
+	return node, nil
+}
+
+func (p *parser) parseLiteral(ctx *context) (*ast.LiteralNode, error) {
+	node := ast.Literal(ctx.currentToken())
+	ctx.progress(1) // skip literal/folded token
+
+	tk := ctx.currentToken()
+	var comment *ast.CommentGroupNode
+	if tk.Type == token.CommentType {
+		comment = p.parseCommentOnly(ctx)
+		comment.SetPath(ctx.path)
+		if err := node.SetComment(comment); err != nil {
+			return nil, errors.Wrapf(err, "failed to set comment to literal")
+		}
+		tk = ctx.currentToken()
+	}
+	value, err := p.parseToken(ctx, tk)
+	if err != nil {
+		return nil, errors.Wrapf(err, "failed to parse literal/folded value")
+	}
+	snode, ok := value.(*ast.StringNode)
+	if !ok {
+		return nil, errors.ErrSyntax("unexpected token. required string token", value.GetToken())
+	}
+	node.Value = snode
+	return node, nil
+}
+
+func (p *parser) isSameLineComment(tk *token.Token, node ast.Node) bool {
+	if tk == nil {
+		return false
+	}
+	if tk.Type != token.CommentType {
+		return false
+	}
+	return tk.Position.Line == node.GetToken().Position.Line
+}
+
+func (p *parser) setSameLineCommentIfExists(ctx *context, node ast.Node) error {
+	tk := ctx.currentToken()
+	if !p.isSameLineComment(tk, node) {
+		return nil
+	}
+	comment := ast.CommentGroup([]*token.Token{tk})
+	comment.SetPath(ctx.path)
+	if err := node.SetComment(comment); err != nil {
+		return errors.Wrapf(err, "failed to set comment token to ast.Node")
+	}
+	return nil
+}
+
+func (p *parser) parseDocument(ctx *context) (*ast.DocumentNode, error) {
+	startTk := ctx.currentToken()
+	ctx.progress(1) // skip document header token
+	body, err := p.parseToken(ctx, ctx.currentToken())
+	if err != nil {
+		return nil, errors.Wrapf(err, "failed to parse document body")
+	}
+	node := ast.Document(startTk, body)
+	if ntk := ctx.nextToken(); ntk != nil && ntk.Type == token.DocumentEndType {
+		node.End = ntk
+		ctx.progress(1)
+	}
+	return node, nil
+}
+
+func (p *parser) parseCommentOnly(ctx *context) *ast.CommentGroupNode {
+	commentTokens := []*token.Token{}
+	for {
+		tk := ctx.currentToken()
+		if tk == nil {
+			break
+		}
+		if tk.Type != token.CommentType {
+			break
+		}
+		commentTokens = append(commentTokens, tk)
+		ctx.progressIgnoreComment(1) // skip comment token
+	}
+	return ast.CommentGroup(commentTokens)
+}
+
+func (p *parser) parseFootComment(ctx *context, col int) *ast.CommentGroupNode {
+	commentTokens := []*token.Token{}
+	for {
+		ctx.progressIgnoreComment(1)
+		commentTokens = append(commentTokens, ctx.currentToken())
+
+		nextTk := ctx.nextToken()
+		if nextTk == nil {
+			break
+		}
+		if nextTk.Type != token.CommentType {
+			break
+		}
+		if col > nextTk.Position.Column {
+			break
+		}
+	}
+	return ast.CommentGroup(commentTokens)
+}
+
+func (p *parser) parseComment(ctx *context) (ast.Node, error) {
+	group := p.parseCommentOnly(ctx)
+	node, err := p.parseToken(ctx, ctx.currentToken())
+	if err != nil {
+		return nil, errors.Wrapf(err, "failed to parse node after comment")
+	}
+	if node == nil {
+		return group, nil
+	}
+	group.SetPath(node.GetPath())
+	if err := node.SetComment(group); err != nil {
+		return nil, errors.Wrapf(err, "failed to set comment token to node")
+	}
+	return node, nil
+}
+
+func (p *parser) parseMappingKey(ctx *context) (*ast.MappingKeyNode, error) {
+	keyTk := ctx.currentToken()
+	node := ast.MappingKey(keyTk)
+	node.SetPath(ctx.path)
+	ctx.progress(1) // skip mapping key token
+	value, err := p.parseToken(ctx.withChild(keyTk.Value), ctx.currentToken())
+	if err != nil {
+		return nil, errors.Wrapf(err, "failed to parse map key")
+	}
+	node.Value = value
+	return node, nil
+}
+
+func (p *parser) parseToken(ctx *context, tk *token.Token) (ast.Node, error) {
+	node, err := p.createNodeFromToken(ctx, tk)
+	if err != nil {
+		return nil, errors.Wrapf(err, "failed to create node from token")
+	}
+	if node != nil && node.GetPath() == "" {
+		node.SetPath(ctx.path)
+	}
+	return node, nil
+}
+
+func (p *parser) createNodeFromToken(ctx *context, tk *token.Token) (ast.Node, error) {
+	if tk == nil {
+		return nil, nil
+	}
+	if tk.NextType() == token.MappingValueType {
+		node, err := p.parseMappingValue(ctx)
+		return node, err
+	}
+	node, err := p.parseScalarValueWithComment(ctx, tk)
+	if err != nil {
+		return nil, errors.Wrapf(err, "failed to parse scalar value")
+	}
+	if node != nil {
+		return node, nil
+	}
+	switch tk.Type {
+	case token.CommentType:
+		return p.parseComment(ctx)
+	case token.MappingKeyType:
+		return p.parseMappingKey(ctx)
+	case token.DocumentHeaderType:
+		return p.parseDocument(ctx)
+	case token.MappingStartType:
+		return p.parseMapping(ctx)
+	case token.SequenceStartType:
+		return p.parseSequence(ctx)
+	case token.SequenceEntryType:
+		return p.parseSequenceEntry(ctx)
+	case token.AnchorType:
+		return p.parseAnchor(ctx)
+	case token.AliasType:
+		return p.parseAlias(ctx)
+	case token.DirectiveType:
+		return p.parseDirective(ctx)
+	case token.TagType:
+		return p.parseTag(ctx)
+	case token.LiteralType, token.FoldedType:
+		return p.parseLiteral(ctx)
+	}
+	return nil, nil
+}
+
+func (p *parser) parse(tokens token.Tokens, mode Mode) (*ast.File, error) {
+	ctx := newContext(tokens, mode)
+	file := &ast.File{Docs: []*ast.DocumentNode{}}
+	for ctx.next() {
+		node, err := p.parseToken(ctx, ctx.currentToken())
+		if err != nil {
+			return nil, errors.Wrapf(err, "failed to parse")
+		}
+		ctx.progressIgnoreComment(1)
+		if node == nil {
+			continue
+		}
+		if doc, ok := node.(*ast.DocumentNode); ok {
+			file.Docs = append(file.Docs, doc)
+		} else {
+			file.Docs = append(file.Docs, ast.Document(nil, node))
+		}
+	}
+	return file, nil
+}
+
+type Mode uint
+
+const (
+	ParseComments Mode = 1 << iota // parse comments and add them to AST
+)
+
+// ParseBytes parse from byte slice, and returns ast.File
+func ParseBytes(bytes []byte, mode Mode) (*ast.File, error) {
+	tokens := lexer.Tokenize(string(bytes))
+	f, err := Parse(tokens, mode)
+	if err != nil {
+		return nil, errors.Wrapf(err, "failed to parse")
+	}
+	return f, nil
+}
+
+// Parse parse from token instances, and returns ast.File
+func Parse(tokens token.Tokens, mode Mode) (*ast.File, error) {
+	var p parser
+	f, err := p.parse(tokens, mode)
+	if err != nil {
+		return nil, errors.Wrapf(err, "failed to parse")
+	}
+	return f, nil
+}
+
+// Parse parse from filename, and returns ast.File
+func ParseFile(filename string, mode Mode) (*ast.File, error) {
+	file, err := ioutil.ReadFile(filename)
+	if err != nil {
+		return nil, errors.Wrapf(err, "failed to read file: %s", filename)
+	}
+	f, err := ParseBytes(file, mode)
+	if err != nil {
+		return nil, errors.Wrapf(err, "failed to parse")
+	}
+	f.Name = filename
+	return f, nil
+}

vendor/github.com/goccy/go-yaml/path.go

@@ -0,0 +1,794 @@
+package yaml
+
+import (
+	"bytes"
+	"fmt"
+	"io"
+	"strconv"
+	"strings"
+
+	"github.com/goccy/go-yaml/ast"
+	"github.com/goccy/go-yaml/internal/errors"
+	"github.com/goccy/go-yaml/parser"
+	"github.com/goccy/go-yaml/printer"
+)
+
+// PathString create Path from string
+//
+// YAMLPath rule
+// $     : the root object/element
+// .     : child operator
+// ..    : recursive descent
+// [num] : object/element of array by number
+// [*]   : all objects/elements for array.
+//
+// If you want to use reserved characters such as `.` and `*` as a key name,
+// enclose them in single quotation as follows ( $.foo.'bar.baz-*'.hoge ).
+// If you want to use a single quote with reserved characters, escape it with `\` ( $.foo.'bar.baz\'s value'.hoge ).
+func PathString(s string) (*Path, error) {
+	buf := []rune(s)
+	length := len(buf)
+	cursor := 0
+	builder := &PathBuilder{}
+	for cursor < length {
+		c := buf[cursor]
+		switch c {
+		case '$':
+			builder = builder.Root()
+			cursor++
+		case '.':
+			b, buf, c, err := parsePathDot(builder, buf, cursor)
+			if err != nil {
+				return nil, errors.Wrapf(err, "failed to parse path of dot")
+			}
+			length = len(buf)
+			builder = b
+			cursor = c
+		case '[':
+			b, buf, c, err := parsePathIndex(builder, buf, cursor)
+			if err != nil {
+				return nil, errors.Wrapf(err, "failed to parse path of index")
+			}
+			length = len(buf)
+			builder = b
+			cursor = c
+		default:
+			return nil, errors.Wrapf(ErrInvalidPathString, "invalid path at %d", cursor)
+		}
+	}
+	return builder.Build(), nil
+}
+
+func parsePathRecursive(b *PathBuilder, buf []rune, cursor int) (*PathBuilder, []rune, int, error) {
+	length := len(buf)
+	cursor += 2 // skip .. characters
+	start := cursor
+	for ; cursor < length; cursor++ {
+		c := buf[cursor]
+		switch c {
+		case '$':
+			return nil, nil, 0, errors.Wrapf(ErrInvalidPathString, "specified '$' after '..' character")
+		case '*':
+			return nil, nil, 0, errors.Wrapf(ErrInvalidPathString, "specified '*' after '..' character")
+		case '.', '[':
+			goto end
+		case ']':
+			return nil, nil, 0, errors.Wrapf(ErrInvalidPathString, "specified ']' after '..' character")
+		}
+	}
+end:
+	if start == cursor {
+		return nil, nil, 0, errors.Wrapf(ErrInvalidPathString, "not found recursive selector")
+	}
+	return b.Recursive(string(buf[start:cursor])), buf, cursor, nil
+}
+
+func parsePathDot(b *PathBuilder, buf []rune, cursor int) (*PathBuilder, []rune, int, error) {
+	length := len(buf)
+	if cursor+1 < length && buf[cursor+1] == '.' {
+		b, buf, c, err := parsePathRecursive(b, buf, cursor)
+		if err != nil {
+			return nil, nil, 0, errors.Wrapf(err, "failed to parse path of recursive")
+		}
+		return b, buf, c, nil
+	}
+	cursor++ // skip . character
+	start := cursor
+
+	// if started single quote, looking for end single quote char
+	if cursor < length && buf[cursor] == '\'' {
+		return parseQuotedKey(b, buf, cursor)
+	}
+	for ; cursor < length; cursor++ {
+		c := buf[cursor]
+		switch c {
+		case '$':
+			return nil, nil, 0, errors.Wrapf(ErrInvalidPathString, "specified '$' after '.' character")
+		case '*':
+			return nil, nil, 0, errors.Wrapf(ErrInvalidPathString, "specified '*' after '.' character")
+		case '.', '[':
+			goto end
+		case ']':
+			return nil, nil, 0, errors.Wrapf(ErrInvalidPathString, "specified ']' after '.' character")
+		}
+	}
+end:
+	if start == cursor {
+		return nil, nil, 0, errors.Wrapf(ErrInvalidPathString, "cloud not find by empty key")
+	}
+	return b.child(string(buf[start:cursor])), buf, cursor, nil
+}
+
+func parseQuotedKey(b *PathBuilder, buf []rune, cursor int) (*PathBuilder, []rune, int, error) {
+	cursor++ // skip single quote
+	start := cursor
+	length := len(buf)
+	var foundEndDelim bool
+	for ; cursor < length; cursor++ {
+		switch buf[cursor] {
+		case '\\':
+			buf = append(append([]rune{}, buf[:cursor]...), buf[cursor+1:]...)
+			length = len(buf)
+		case '\'':
+			foundEndDelim = true
+			goto end
+		}
+	}
+end:
+	if !foundEndDelim {
+		return nil, nil, 0, errors.Wrapf(ErrInvalidPathString, "could not find end delimiter for key")
+	}
+	if start == cursor {
+		return nil, nil, 0, errors.Wrapf(ErrInvalidPathString, "could not find by empty key")
+	}
+	selector := buf[start:cursor]
+	cursor++
+	if cursor < length {
+		switch buf[cursor] {
+		case '$':
+			return nil, nil, 0, errors.Wrapf(ErrInvalidPathString, "specified '$' after '.' character")
+		case '*':
+			return nil, nil, 0, errors.Wrapf(ErrInvalidPathString, "specified '*' after '.' character")
+		case ']':
+			return nil, nil, 0, errors.Wrapf(ErrInvalidPathString, "specified ']' after '.' character")
+		}
+	}
+	return b.child(string(selector)), buf, cursor, nil
+}
+
+func parsePathIndex(b *PathBuilder, buf []rune, cursor int) (*PathBuilder, []rune, int, error) {
+	length := len(buf)
+	cursor++ // skip '[' character
+	if length <= cursor {
+		return nil, nil, 0, errors.Wrapf(ErrInvalidPathString, "unexpected end of YAML Path")
+	}
+	c := buf[cursor]
+	switch c {
+	case '0', '1', '2', '3', '4', '5', '6', '7', '8', '9', '*':
+		start := cursor
+		cursor++
+		for ; cursor < length; cursor++ {
+			c := buf[cursor]
+			switch c {
+			case '0', '1', '2', '3', '4', '5', '6', '7', '8', '9':
+				continue
+			}
+			break
+		}
+		if buf[cursor] != ']' {
+			return nil, nil, 0, errors.Wrapf(ErrInvalidPathString, "invalid character %s at %d", string(buf[cursor]), cursor)
+		}
+		numOrAll := string(buf[start:cursor])
+		if numOrAll == "*" {
+			return b.IndexAll(), buf, cursor + 1, nil
+		}
+		num, err := strconv.ParseInt(numOrAll, 10, 64)
+		if err != nil {
+			return nil, nil, 0, errors.Wrapf(err, "failed to parse number")
+		}
+		return b.Index(uint(num)), buf, cursor + 1, nil
+	}
+	return nil, nil, 0, errors.Wrapf(ErrInvalidPathString, "invalid character %s at %d", c, cursor)
+}
+
+// Path represent YAMLPath ( like a JSONPath ).
+type Path struct {
+	node pathNode
+}
+
+// String path to text.
+func (p *Path) String() string {
+	return p.node.String()
+}
+
+// Read decode from r and set extracted value by YAMLPath to v.
+func (p *Path) Read(r io.Reader, v interface{}) error {
+	node, err := p.ReadNode(r)
+	if err != nil {
+		return errors.Wrapf(err, "failed to read node")
+	}
+	if err := Unmarshal([]byte(node.String()), v); err != nil {
+		return errors.Wrapf(err, "failed to unmarshal")
+	}
+	return nil
+}
+
+// ReadNode create AST from r and extract node by YAMLPath.
+func (p *Path) ReadNode(r io.Reader) (ast.Node, error) {
+	if p.node == nil {
+		return nil, ErrInvalidPath
+	}
+	var buf bytes.Buffer
+	if _, err := io.Copy(&buf, r); err != nil {
+		return nil, errors.Wrapf(err, "failed to copy from reader")
+	}
+	f, err := parser.ParseBytes(buf.Bytes(), 0)
+	if err != nil {
+		return nil, errors.Wrapf(err, "failed to parse yaml")
+	}
+	node, err := p.FilterFile(f)
+	if err != nil {
+		return nil, errors.Wrapf(err, "failed to filter from ast.File")
+	}
+	return node, nil
+}
+
+// Filter filter from target by YAMLPath and set it to v.
+func (p *Path) Filter(target, v interface{}) error {
+	b, err := Marshal(target)
+	if err != nil {
+		return errors.Wrapf(err, "failed to marshal target value")
+	}
+	if err := p.Read(bytes.NewBuffer(b), v); err != nil {
+		return errors.Wrapf(err, "failed to read")
+	}
+	return nil
+}
+
+// FilterFile filter from ast.File by YAMLPath.
+func (p *Path) FilterFile(f *ast.File) (ast.Node, error) {
+	for _, doc := range f.Docs {
+		node, err := p.FilterNode(doc.Body)
+		if err != nil {
+			return nil, errors.Wrapf(err, "failed to filter node by path ( %s )", p.node)
+		}
+		if node != nil {
+			return node, nil
+		}
+	}
+	return nil, errors.Wrapf(ErrNotFoundNode, "failed to find path ( %s )", p.node)
+}
+
+// FilterNode filter from node by YAMLPath.
+func (p *Path) FilterNode(node ast.Node) (ast.Node, error) {
+	n, err := p.node.filter(node)
+	if err != nil {
+		return nil, errors.Wrapf(err, "failed to filter node by path ( %s )", p.node)
+	}
+	return n, nil
+}
+
+// MergeFromReader merge YAML text into ast.File.
+func (p *Path) MergeFromReader(dst *ast.File, src io.Reader) error {
+	var buf bytes.Buffer
+	if _, err := io.Copy(&buf, src); err != nil {
+		return errors.Wrapf(err, "failed to copy from reader")
+	}
+	file, err := parser.ParseBytes(buf.Bytes(), 0)
+	if err != nil {
+		return errors.Wrapf(err, "failed to parse")
+	}
+	if err := p.MergeFromFile(dst, file); err != nil {
+		return errors.Wrapf(err, "failed to merge file")
+	}
+	return nil
+}
+
+// MergeFromFile merge ast.File into ast.File.
+func (p *Path) MergeFromFile(dst *ast.File, src *ast.File) error {
+	base, err := p.FilterFile(dst)
+	if err != nil {
+		return errors.Wrapf(err, "failed to filter file")
+	}
+	for _, doc := range src.Docs {
+		if err := ast.Merge(base, doc); err != nil {
+			return errors.Wrapf(err, "failed to merge")
+		}
+	}
+	return nil
+}
+
+// MergeFromNode merge ast.Node into ast.File.
+func (p *Path) MergeFromNode(dst *ast.File, src ast.Node) error {
+	base, err := p.FilterFile(dst)
+	if err != nil {
+		return errors.Wrapf(err, "failed to filter file")
+	}
+	if err := ast.Merge(base, src); err != nil {
+		return errors.Wrapf(err, "failed to merge")
+	}
+	return nil
+}
+
+// ReplaceWithReader replace ast.File with io.Reader.
+func (p *Path) ReplaceWithReader(dst *ast.File, src io.Reader) error {
+	var buf bytes.Buffer
+	if _, err := io.Copy(&buf, src); err != nil {
+		return errors.Wrapf(err, "failed to copy from reader")
+	}
+	file, err := parser.ParseBytes(buf.Bytes(), 0)
+	if err != nil {
+		return errors.Wrapf(err, "failed to parse")
+	}
+	if err := p.ReplaceWithFile(dst, file); err != nil {
+		return errors.Wrapf(err, "failed to replace file")
+	}
+	return nil
+}
+
+// ReplaceWithFile replace ast.File with ast.File.
+func (p *Path) ReplaceWithFile(dst *ast.File, src *ast.File) error {
+	for _, doc := range src.Docs {
+		if err := p.ReplaceWithNode(dst, doc); err != nil {
+			return errors.Wrapf(err, "failed to replace file by path ( %s )", p.node)
+		}
+	}
+	return nil
+}
+
+// ReplaceNode replace ast.File with ast.Node.
+func (p *Path) ReplaceWithNode(dst *ast.File, node ast.Node) error {
+	for _, doc := range dst.Docs {
+		if node.Type() == ast.DocumentType {
+			node = node.(*ast.DocumentNode).Body
+		}
+		if err := p.node.replace(doc.Body, node); err != nil {
+			return errors.Wrapf(err, "failed to replace node by path ( %s )", p.node)
+		}
+	}
+	return nil
+}
+
+// AnnotateSource add annotation to passed source ( see section 5.1 in README.md ).
+func (p *Path) AnnotateSource(source []byte, colored bool) ([]byte, error) {
+	file, err := parser.ParseBytes([]byte(source), 0)
+	if err != nil {
+		return nil, err
+	}
+	node, err := p.FilterFile(file)
+	if err != nil {
+		return nil, err
+	}
+	var pp printer.Printer
+	return []byte(pp.PrintErrorToken(node.GetToken(), colored)), nil
+}
+
+// PathBuilder represent builder for YAMLPath.
+type PathBuilder struct {
+	root *rootNode
+	node pathNode
+}
+
+// Root add '$' to current path.
+func (b *PathBuilder) Root() *PathBuilder {
+	root := newRootNode()
+	return &PathBuilder{root: root, node: root}
+}
+
+// IndexAll add '[*]' to current path.
+func (b *PathBuilder) IndexAll() *PathBuilder {
+	b.node = b.node.chain(newIndexAllNode())
+	return b
+}
+
+// Recursive add '..selector' to current path.
+func (b *PathBuilder) Recursive(selector string) *PathBuilder {
+	b.node = b.node.chain(newRecursiveNode(selector))
+	return b
+}
+
+func (b *PathBuilder) containsReservedPathCharacters(path string) bool {
+	if strings.Contains(path, ".") {
+		return true
+	}
+	if strings.Contains(path, "*") {
+		return true
+	}
+	return false
+}
+
+func (b *PathBuilder) enclosedSingleQuote(name string) bool {
+	return strings.HasPrefix(name, "'") && strings.HasSuffix(name, "'")
+}
+
+func (b *PathBuilder) normalizeSelectorName(name string) string {
+	if b.enclosedSingleQuote(name) {
+		// already escaped name
+		return name
+	}
+	if b.containsReservedPathCharacters(name) {
+		escapedName := strings.ReplaceAll(name, `'`, `\'`)
+		return "'" + escapedName + "'"
+	}
+	return name
+}
+
+func (b *PathBuilder) child(name string) *PathBuilder {
+	b.node = b.node.chain(newSelectorNode(name))
+	return b
+}
+
+// Child add '.name' to current path.
+func (b *PathBuilder) Child(name string) *PathBuilder {
+	return b.child(b.normalizeSelectorName(name))
+}
+
+// Index add '[idx]' to current path.
+func (b *PathBuilder) Index(idx uint) *PathBuilder {
+	b.node = b.node.chain(newIndexNode(idx))
+	return b
+}
+
+// Build build YAMLPath.
+func (b *PathBuilder) Build() *Path {
+	return &Path{node: b.root}
+}
+
+type pathNode interface {
+	fmt.Stringer
+	chain(pathNode) pathNode
+	filter(ast.Node) (ast.Node, error)
+	replace(ast.Node, ast.Node) error
+}
+
+type basePathNode struct {
+	child pathNode
+}
+
+func (n *basePathNode) chain(node pathNode) pathNode {
+	n.child = node
+	return node
+}
+
+type rootNode struct {
+	*basePathNode
+}
+
+func newRootNode() *rootNode {
+	return &rootNode{basePathNode: &basePathNode{}}
+}
+
+func (n *rootNode) String() string {
+	s := "$"
+	if n.child != nil {
+		s += n.child.String()
+	}
+	return s
+}
+
+func (n *rootNode) filter(node ast.Node) (ast.Node, error) {
+	if n.child == nil {
+		return nil, nil
+	}
+	filtered, err := n.child.filter(node)
+	if err != nil {
+		return nil, errors.Wrapf(err, "failed to filter")
+	}
+	return filtered, nil
+}
+
+func (n *rootNode) replace(node ast.Node, target ast.Node) error {
+	if n.child == nil {
+		return nil
+	}
+	if err := n.child.replace(node, target); err != nil {
+		return errors.Wrapf(err, "failed to replace")
+	}
+	return nil
+}
+
+type selectorNode struct {
+	*basePathNode
+	selector string
+}
+
+func newSelectorNode(selector string) *selectorNode {
+	return &selectorNode{
+		basePathNode: &basePathNode{},
+		selector:     selector,
+	}
+}
+
+func (n *selectorNode) filter(node ast.Node) (ast.Node, error) {
+	switch node.Type() {
+	case ast.MappingType:
+		for _, value := range node.(*ast.MappingNode).Values {
+			key := value.Key.GetToken().Value
+			if key == n.selector {
+				if n.child == nil {
+					return value.Value, nil
+				}
+				filtered, err := n.child.filter(value.Value)
+				if err != nil {
+					return nil, errors.Wrapf(err, "failed to filter")
+				}
+				return filtered, nil
+			}
+		}
+	case ast.MappingValueType:
+		value := node.(*ast.MappingValueNode)
+		key := value.Key.GetToken().Value
+		if key == n.selector {
+			if n.child == nil {
+				return value.Value, nil
+			}
+			filtered, err := n.child.filter(value.Value)
+			if err != nil {
+				return nil, errors.Wrapf(err, "failed to filter")
+			}
+			return filtered, nil
+		}
+	default:
+		return nil, errors.Wrapf(ErrInvalidQuery, "expected node type is map or map value. but got %s", node.Type())
+	}
+	return nil, nil
+}
+
+func (n *selectorNode) replaceMapValue(value *ast.MappingValueNode, target ast.Node) error {
+	key := value.Key.GetToken().Value
+	if key != n.selector {
+		return nil
+	}
+	if n.child == nil {
+		if err := value.Replace(target); err != nil {
+			return errors.Wrapf(err, "failed to replace")
+		}
+	} else {
+		if err := n.child.replace(value.Value, target); err != nil {
+			return errors.Wrapf(err, "failed to replace")
+		}
+	}
+	return nil
+}
+
+func (n *selectorNode) replace(node ast.Node, target ast.Node) error {
+	switch node.Type() {
+	case ast.MappingType:
+		for _, value := range node.(*ast.MappingNode).Values {
+			if err := n.replaceMapValue(value, target); err != nil {
+				return errors.Wrapf(err, "failed to replace map value")
+			}
+		}
+	case ast.MappingValueType:
+		value := node.(*ast.MappingValueNode)
+		if err := n.replaceMapValue(value, target); err != nil {
+			return errors.Wrapf(err, "failed to replace map value")
+		}
+	default:
+		return errors.Wrapf(ErrInvalidQuery, "expected node type is map or map value. but got %s", node.Type())
+	}
+	return nil
+}
+
+func (n *selectorNode) String() string {
+	s := fmt.Sprintf(".%s", n.selector)
+	if n.child != nil {
+		s += n.child.String()
+	}
+	return s
+}
+
+type indexNode struct {
+	*basePathNode
+	selector uint
+}
+
+func newIndexNode(selector uint) *indexNode {
+	return &indexNode{
+		basePathNode: &basePathNode{},
+		selector:     selector,
+	}
+}
+
+func (n *indexNode) filter(node ast.Node) (ast.Node, error) {
+	if node.Type() != ast.SequenceType {
+		return nil, errors.Wrapf(ErrInvalidQuery, "expected sequence type node. but got %s", node.Type())
+	}
+	sequence := node.(*ast.SequenceNode)
+	if n.selector >= uint(len(sequence.Values)) {
+		return nil, errors.Wrapf(ErrInvalidQuery, "expected index is %d. but got sequences has %d items", n.selector, sequence.Values)
+	}
+	value := sequence.Values[n.selector]
+	if n.child == nil {
+		return value, nil
+	}
+	filtered, err := n.child.filter(value)
+	if err != nil {
+		return nil, errors.Wrapf(err, "failed to filter")
+	}
+	return filtered, nil
+}
+
+func (n *indexNode) replace(node ast.Node, target ast.Node) error {
+	if node.Type() != ast.SequenceType {
+		return errors.Wrapf(ErrInvalidQuery, "expected sequence type node. but got %s", node.Type())
+	}
+	sequence := node.(*ast.SequenceNode)
+	if n.selector >= uint(len(sequence.Values)) {
+		return errors.Wrapf(ErrInvalidQuery, "expected index is %d. but got sequences has %d items", n.selector, sequence.Values)
+	}
+	if n.child == nil {
+		if err := sequence.Replace(int(n.selector), target); err != nil {
+			return errors.Wrapf(err, "failed to replace")
+		}
+		return nil
+	}
+	if err := n.child.replace(sequence.Values[n.selector], target); err != nil {
+		return errors.Wrapf(err, "failed to replace")
+	}
+	return nil
+}
+
+func (n *indexNode) String() string {
+	s := fmt.Sprintf("[%d]", n.selector)
+	if n.child != nil {
+		s += n.child.String()
+	}
+	return s
+}
+
+type indexAllNode struct {
+	*basePathNode
+}
+
+func newIndexAllNode() *indexAllNode {
+	return &indexAllNode{
+		basePathNode: &basePathNode{},
+	}
+}
+
+func (n *indexAllNode) String() string {
+	s := "[*]"
+	if n.child != nil {
+		s += n.child.String()
+	}
+	return s
+}
+
+func (n *indexAllNode) filter(node ast.Node) (ast.Node, error) {
+	if node.Type() != ast.SequenceType {
+		return nil, errors.Wrapf(ErrInvalidQuery, "expected sequence type node. but got %s", node.Type())
+	}
+	sequence := node.(*ast.SequenceNode)
+	if n.child == nil {
+		return sequence, nil
+	}
+	out := *sequence
+	out.Values = []ast.Node{}
+	for _, value := range sequence.Values {
+		filtered, err := n.child.filter(value)
+		if err != nil {
+			return nil, errors.Wrapf(err, "failed to filter")
+		}
+		out.Values = append(out.Values, filtered)
+	}
+	return &out, nil
+}
+
+func (n *indexAllNode) replace(node ast.Node, target ast.Node) error {
+	if node.Type() != ast.SequenceType {
+		return errors.Wrapf(ErrInvalidQuery, "expected sequence type node. but got %s", node.Type())
+	}
+	sequence := node.(*ast.SequenceNode)
+	if n.child == nil {
+		for idx := range sequence.Values {
+			if err := sequence.Replace(idx, target); err != nil {
+				return errors.Wrapf(err, "failed to replace")
+			}
+		}
+		return nil
+	}
+	for _, value := range sequence.Values {
+		if err := n.child.replace(value, target); err != nil {
+			return errors.Wrapf(err, "failed to replace")
+		}
+	}
+	return nil
+}
+
+type recursiveNode struct {
+	*basePathNode
+	selector string
+}
+
+func newRecursiveNode(selector string) *recursiveNode {
+	return &recursiveNode{
+		basePathNode: &basePathNode{},
+		selector:     selector,
+	}
+}
+
+func (n *recursiveNode) String() string {
+	s := fmt.Sprintf("..%s", n.selector)
+	if n.child != nil {
+		s += n.child.String()
+	}
+	return s
+}
+
+func (n *recursiveNode) filterNode(node ast.Node) (*ast.SequenceNode, error) {
+	sequence := &ast.SequenceNode{BaseNode: &ast.BaseNode{}}
+	switch typedNode := node.(type) {
+	case *ast.MappingNode:
+		for _, value := range typedNode.Values {
+			seq, err := n.filterNode(value)
+			if err != nil {
+				return nil, errors.Wrapf(err, "failed to filter")
+			}
+			sequence.Values = append(sequence.Values, seq.Values...)
+		}
+	case *ast.MappingValueNode:
+		key := typedNode.Key.GetToken().Value
+		if n.selector == key {
+			sequence.Values = append(sequence.Values, typedNode.Value)
+		}
+		seq, err := n.filterNode(typedNode.Value)
+		if err != nil {
+			return nil, errors.Wrapf(err, "failed to filter")
+		}
+		sequence.Values = append(sequence.Values, seq.Values...)
+	case *ast.SequenceNode:
+		for _, value := range typedNode.Values {
+			seq, err := n.filterNode(value)
+			if err != nil {
+				return nil, errors.Wrapf(err, "failed to filter")
+			}
+			sequence.Values = append(sequence.Values, seq.Values...)
+		}
+	}
+	return sequence, nil
+}
+
+func (n *recursiveNode) filter(node ast.Node) (ast.Node, error) {
+	sequence, err := n.filterNode(node)
+	if err != nil {
+		return nil, errors.Wrapf(err, "failed to filter")
+	}
+	sequence.Start = node.GetToken()
+	return sequence, nil
+}
+
+func (n *recursiveNode) replaceNode(node ast.Node, target ast.Node) error {
+	switch typedNode := node.(type) {
+	case *ast.MappingNode:
+		for _, value := range typedNode.Values {
+			if err := n.replaceNode(value, target); err != nil {
+				return errors.Wrapf(err, "failed to replace")
+			}
+		}
+	case *ast.MappingValueNode:
+		key := typedNode.Key.GetToken().Value
+		if n.selector == key {
+			if err := typedNode.Replace(target); err != nil {
+				return errors.Wrapf(err, "failed to replace")
+			}
+		}
+		if err := n.replaceNode(typedNode.Value, target); err != nil {
+			return errors.Wrapf(err, "failed to replace")
+		}
+	case *ast.SequenceNode:
+		for _, value := range typedNode.Values {
+			if err := n.replaceNode(value, target); err != nil {
+				return errors.Wrapf(err, "failed to replace")
+			}
+		}
+	}
+	return nil
+}
+
+func (n *recursiveNode) replace(node ast.Node, target ast.Node) error {
+	if err := n.replaceNode(node, target); err != nil {
+		return errors.Wrapf(err, "failed to replace")
+	}
+	return nil
+}

vendor/github.com/goccy/go-yaml/printer/printer.go

@@ -0,0 +1,352 @@
+package printer
+
+import (
+	"fmt"
+	"math"
+	"strings"
+
+	"github.com/fatih/color"
+	"github.com/goccy/go-yaml/ast"
+	"github.com/goccy/go-yaml/token"
+)
+
+// Property additional property set for each the token
+type Property struct {
+	Prefix string
+	Suffix string
+}
+
+// PrintFunc returns property instance
+type PrintFunc func() *Property
+
+// Printer create text from token collection or ast
+type Printer struct {
+	LineNumber       bool
+	LineNumberFormat func(num int) string
+	MapKey           PrintFunc
+	Anchor           PrintFunc
+	Alias            PrintFunc
+	Bool             PrintFunc
+	String           PrintFunc
+	Number           PrintFunc
+}
+
+func defaultLineNumberFormat(num int) string {
+	return fmt.Sprintf("%2d | ", num)
+}
+
+func (p *Printer) property(tk *token.Token) *Property {
+	prop := &Property{}
+	switch tk.PreviousType() {
+	case token.AnchorType:
+		if p.Anchor != nil {
+			return p.Anchor()
+		}
+		return prop
+	case token.AliasType:
+		if p.Alias != nil {
+			return p.Alias()
+		}
+		return prop
+	}
+	switch tk.NextType() {
+	case token.MappingValueType:
+		if p.MapKey != nil {
+			return p.MapKey()
+		}
+		return prop
+	}
+	switch tk.Type {
+	case token.BoolType:
+		if p.Bool != nil {
+			return p.Bool()
+		}
+		return prop
+	case token.AnchorType:
+		if p.Anchor != nil {
+			return p.Anchor()
+		}
+		return prop
+	case token.AliasType:
+		if p.Anchor != nil {
+			return p.Alias()
+		}
+		return prop
+	case token.StringType, token.SingleQuoteType, token.DoubleQuoteType:
+		if p.String != nil {
+			return p.String()
+		}
+		return prop
+	case token.IntegerType, token.FloatType:
+		if p.Number != nil {
+			return p.Number()
+		}
+		return prop
+	default:
+	}
+	return prop
+}
+
+// PrintTokens create text from token collection
+func (p *Printer) PrintTokens(tokens token.Tokens) string {
+	if len(tokens) == 0 {
+		return ""
+	}
+	if p.LineNumber {
+		if p.LineNumberFormat == nil {
+			p.LineNumberFormat = defaultLineNumberFormat
+		}
+	}
+	texts := []string{}
+	lineNumber := tokens[0].Position.Line
+	for _, tk := range tokens {
+		lines := strings.Split(tk.Origin, "\n")
+		prop := p.property(tk)
+		header := ""
+		if p.LineNumber {
+			header = p.LineNumberFormat(lineNumber)
+		}
+		if len(lines) == 1 {
+			line := prop.Prefix + lines[0] + prop.Suffix
+			if len(texts) == 0 {
+				texts = append(texts, header+line)
+				lineNumber++
+			} else {
+				text := texts[len(texts)-1]
+				texts[len(texts)-1] = text + line
+			}
+		} else {
+			for idx, src := range lines {
+				if p.LineNumber {
+					header = p.LineNumberFormat(lineNumber)
+				}
+				line := prop.Prefix + src + prop.Suffix
+				if idx == 0 {
+					if len(texts) == 0 {
+						texts = append(texts, header+line)
+						lineNumber++
+					} else {
+						text := texts[len(texts)-1]
+						texts[len(texts)-1] = text + line
+					}
+				} else {
+					texts = append(texts, fmt.Sprintf("%s%s", header, line))
+					lineNumber++
+				}
+			}
+		}
+	}
+	return strings.Join(texts, "\n")
+}
+
+// PrintNode create text from ast.Node
+func (p *Printer) PrintNode(node ast.Node) []byte {
+	return []byte(fmt.Sprintf("%+v\n", node))
+}
+
+const escape = "\x1b"
+
+func format(attr color.Attribute) string {
+	return fmt.Sprintf("%s[%dm", escape, attr)
+}
+
+func (p *Printer) setDefaultColorSet() {
+	p.Bool = func() *Property {
+		return &Property{
+			Prefix: format(color.FgHiMagenta),
+			Suffix: format(color.Reset),
+		}
+	}
+	p.Number = func() *Property {
+		return &Property{
+			Prefix: format(color.FgHiMagenta),
+			Suffix: format(color.Reset),
+		}
+	}
+	p.MapKey = func() *Property {
+		return &Property{
+			Prefix: format(color.FgHiCyan),
+			Suffix: format(color.Reset),
+		}
+	}
+	p.Anchor = func() *Property {
+		return &Property{
+			Prefix: format(color.FgHiYellow),
+			Suffix: format(color.Reset),
+		}
+	}
+	p.Alias = func() *Property {
+		return &Property{
+			Prefix: format(color.FgHiYellow),
+			Suffix: format(color.Reset),
+		}
+	}
+	p.String = func() *Property {
+		return &Property{
+			Prefix: format(color.FgHiGreen),
+			Suffix: format(color.Reset),
+		}
+	}
+}
+
+func (p *Printer) PrintErrorMessage(msg string, isColored bool) string {
+	if isColored {
+		return fmt.Sprintf("%s%s%s",
+			format(color.FgHiRed),
+			msg,
+			format(color.Reset),
+		)
+	}
+	return msg
+}
+
+func (p *Printer) removeLeftSideNewLineChar(src string) string {
+	return strings.TrimLeft(strings.TrimLeft(strings.TrimLeft(src, "\r"), "\n"), "\r\n")
+}
+
+func (p *Printer) removeRightSideNewLineChar(src string) string {
+	return strings.TrimRight(strings.TrimRight(strings.TrimRight(src, "\r"), "\n"), "\r\n")
+}
+
+func (p *Printer) removeRightSideWhiteSpaceChar(src string) string {
+	return p.removeRightSideNewLineChar(strings.TrimRight(src, " "))
+}
+
+func (p *Printer) newLineCount(s string) int {
+	src := []rune(s)
+	size := len(src)
+	cnt := 0
+	for i := 0; i < size; i++ {
+		c := src[i]
+		switch c {
+		case '\r':
+			if i+1 < size && src[i+1] == '\n' {
+				i++
+			}
+			cnt++
+		case '\n':
+			cnt++
+		}
+	}
+	return cnt
+}
+
+func (p *Printer) isNewLineLastChar(s string) bool {
+	for i := len(s) - 1; i > 0; i-- {
+		c := s[i]
+		switch c {
+		case ' ':
+			continue
+		case '\n', '\r':
+			return true
+		}
+		break
+	}
+	return false
+}
+
+func (p *Printer) printBeforeTokens(tk *token.Token, minLine, extLine int) token.Tokens {
+	for {
+		if tk.Prev == nil {
+			break
+		}
+		if tk.Prev.Position.Line < minLine {
+			break
+		}
+		tk = tk.Prev
+	}
+	minTk := tk.Clone()
+	if minTk.Prev != nil {
+		// add white spaces to minTk by prev token
+		prev := minTk.Prev
+		whiteSpaceLen := len(prev.Origin) - len(strings.TrimRight(prev.Origin, " "))
+		minTk.Origin = strings.Repeat(" ", whiteSpaceLen) + minTk.Origin
+	}
+	minTk.Origin = p.removeLeftSideNewLineChar(minTk.Origin)
+	tokens := token.Tokens{minTk}
+	tk = minTk.Next
+	for tk != nil && tk.Position.Line <= extLine {
+		clonedTk := tk.Clone()
+		tokens.Add(clonedTk)
+		tk = clonedTk.Next
+	}
+	lastTk := tokens[len(tokens)-1]
+	trimmedOrigin := p.removeRightSideWhiteSpaceChar(lastTk.Origin)
+	suffix := lastTk.Origin[len(trimmedOrigin):]
+	lastTk.Origin = trimmedOrigin
+
+	if lastTk.Next != nil && len(suffix) > 1 {
+		next := lastTk.Next.Clone()
+		// add suffix to header of next token
+		if suffix[0] == '\n' || suffix[0] == '\r' {
+			suffix = suffix[1:]
+		}
+		next.Origin = suffix + next.Origin
+		lastTk.Next = next
+	}
+	return tokens
+}
+
+func (p *Printer) printAfterTokens(tk *token.Token, maxLine int) token.Tokens {
+	tokens := token.Tokens{}
+	if tk == nil {
+		return tokens
+	}
+	if tk.Position.Line > maxLine {
+		return tokens
+	}
+	minTk := tk.Clone()
+	minTk.Origin = p.removeLeftSideNewLineChar(minTk.Origin)
+	tokens.Add(minTk)
+	tk = minTk.Next
+	for tk != nil && tk.Position.Line <= maxLine {
+		clonedTk := tk.Clone()
+		tokens.Add(clonedTk)
+		tk = clonedTk.Next
+	}
+	return tokens
+}
+
+func (p *Printer) setupErrorTokenFormat(annotateLine int, isColored bool) {
+	prefix := func(annotateLine, num int) string {
+		if annotateLine == num {
+			return fmt.Sprintf("> %2d | ", num)
+		}
+		return fmt.Sprintf("  %2d | ", num)
+	}
+	p.LineNumber = true
+	p.LineNumberFormat = func(num int) string {
+		if isColored {
+			fn := color.New(color.Bold, color.FgHiWhite).SprintFunc()
+			return fn(prefix(annotateLine, num))
+		}
+		return prefix(annotateLine, num)
+	}
+	if isColored {
+		p.setDefaultColorSet()
+	}
+}
+
+func (p *Printer) PrintErrorToken(tk *token.Token, isColored bool) string {
+	errToken := tk
+	curLine := tk.Position.Line
+	curExtLine := curLine + p.newLineCount(p.removeLeftSideNewLineChar(tk.Origin))
+	if p.isNewLineLastChar(tk.Origin) {
+		// if last character ( exclude white space ) is new line character, ignore it.
+		curExtLine--
+	}
+
+	minLine := int(math.Max(float64(curLine-3), 1))
+	maxLine := curExtLine + 3
+	p.setupErrorTokenFormat(curLine, isColored)
+
+	beforeTokens := p.printBeforeTokens(tk, minLine, curExtLine)
+	lastTk := beforeTokens[len(beforeTokens)-1]
+	afterTokens := p.printAfterTokens(lastTk.Next, maxLine)
+
+	beforeSource := p.PrintTokens(beforeTokens)
+	prefixSpaceNum := len(fmt.Sprintf("  %2d | ", curLine))
+	annotateLine := strings.Repeat(" ", prefixSpaceNum+errToken.Position.Column-1) + "^"
+	afterSource := p.PrintTokens(afterTokens)
+	return fmt.Sprintf("%s\n%s\n%s", beforeSource, annotateLine, afterSource)
+}

vendor/github.com/goccy/go-yaml/scanner/context.go

@@ -0,0 +1,229 @@
+package scanner
+
+import (
+	"sync"
+
+	"github.com/goccy/go-yaml/token"
+)
+
+const whitespace = ' '
+
+// Context context at scanning
+type Context struct {
+	idx                int
+	size               int
+	notSpaceCharPos    int
+	notSpaceOrgCharPos int
+	src                []rune
+	buf                []rune
+	obuf               []rune
+	tokens             token.Tokens
+	isRawFolded        bool
+	isLiteral          bool
+	isFolded           bool
+	isSingleLine       bool
+	literalOpt         string
+}
+
+var (
+	ctxPool = sync.Pool{
+		New: func() interface{} {
+			return createContext()
+		},
+	}
+)
+
+func createContext() *Context {
+	return &Context{
+		idx:          0,
+		tokens:       token.Tokens{},
+		isSingleLine: true,
+	}
+}
+
+func newContext(src []rune) *Context {
+	ctx := ctxPool.Get().(*Context)
+	ctx.reset(src)
+	return ctx
+}
+
+func (c *Context) release() {
+	ctxPool.Put(c)
+}
+
+func (c *Context) reset(src []rune) {
+	c.idx = 0
+	c.size = len(src)
+	c.src = src
+	c.tokens = c.tokens[:0]
+	c.resetBuffer()
+	c.isRawFolded = false
+	c.isSingleLine = true
+	c.isLiteral = false
+	c.isFolded = false
+	c.literalOpt = ""
+}
+
+func (c *Context) resetBuffer() {
+	c.buf = c.buf[:0]
+	c.obuf = c.obuf[:0]
+	c.notSpaceCharPos = 0
+	c.notSpaceOrgCharPos = 0
+}
+
+func (c *Context) isSaveIndentMode() bool {
+	return c.isLiteral || c.isFolded || c.isRawFolded
+}
+
+func (c *Context) breakLiteral() {
+	c.isLiteral = false
+	c.isRawFolded = false
+	c.isFolded = false
+	c.literalOpt = ""
+}
+
+func (c *Context) addToken(tk *token.Token) {
+	if tk == nil {
+		return
+	}
+	c.tokens = append(c.tokens, tk)
+}
+
+func (c *Context) addBuf(r rune) {
+	if len(c.buf) == 0 && r == ' ' {
+		return
+	}
+	c.buf = append(c.buf, r)
+	if r != ' ' && r != '\t' {
+		c.notSpaceCharPos = len(c.buf)
+	}
+}
+
+func (c *Context) addOriginBuf(r rune) {
+	c.obuf = append(c.obuf, r)
+	if r != ' ' && r != '\t' {
+		c.notSpaceOrgCharPos = len(c.obuf)
+	}
+}
+
+func (c *Context) removeRightSpaceFromBuf() int {
+	trimmedBuf := c.obuf[:c.notSpaceOrgCharPos]
+	buflen := len(trimmedBuf)
+	diff := len(c.obuf) - buflen
+	if diff > 0 {
+		c.obuf = c.obuf[:buflen]
+		c.buf = c.bufferedSrc()
+	}
+	return diff
+}
+
+func (c *Context) isDocument() bool {
+	return c.isLiteral || c.isFolded || c.isRawFolded
+}
+
+func (c *Context) isEOS() bool {
+	return len(c.src)-1 <= c.idx
+}
+
+func (c *Context) isNextEOS() bool {
+	return len(c.src)-1 <= c.idx+1
+}
+
+func (c *Context) next() bool {
+	return c.idx < c.size
+}
+
+func (c *Context) source(s, e int) string {
+	return string(c.src[s:e])
+}
+
+func (c *Context) previousChar() rune {
+	if c.idx > 0 {
+		return c.src[c.idx-1]
+	}
+	return rune(0)
+}
+
+func (c *Context) currentChar() rune {
+	if c.size > c.idx {
+		return c.src[c.idx]
+	}
+	return rune(0)
+}
+
+func (c *Context) currentCharWithSkipWhitespace() rune {
+	idx := c.idx
+	for c.size > idx {
+		ch := c.src[idx]
+		if ch != whitespace {
+			return ch
+		}
+		idx++
+	}
+	return rune(0)
+}
+
+func (c *Context) nextChar() rune {
+	if c.size > c.idx+1 {
+		return c.src[c.idx+1]
+	}
+	return rune(0)
+}
+
+func (c *Context) repeatNum(r rune) int {
+	cnt := 0
+	for i := c.idx; i < c.size; i++ {
+		if c.src[i] == r {
+			cnt++
+		} else {
+			break
+		}
+	}
+	return cnt
+}
+
+func (c *Context) progress(num int) {
+	c.idx += num
+}
+
+func (c *Context) nextPos() int {
+	return c.idx + 1
+}
+
+func (c *Context) existsBuffer() bool {
+	return len(c.bufferedSrc()) != 0
+}
+
+func (c *Context) bufferedSrc() []rune {
+	src := c.buf[:c.notSpaceCharPos]
+	if len(src) > 0 && src[len(src)-1] == '\n' && c.isDocument() && c.literalOpt == "-" {
+		// remove end '\n' character
+		src = src[:len(src)-1]
+	}
+	return src
+}
+
+func (c *Context) bufferedToken(pos *token.Position) *token.Token {
+	if c.idx == 0 {
+		return nil
+	}
+	source := c.bufferedSrc()
+	if len(source) == 0 {
+		return nil
+	}
+	var tk *token.Token
+	if c.isDocument() {
+		tk = token.String(string(source), string(c.obuf), pos)
+	} else {
+		tk = token.New(string(source), string(c.obuf), pos)
+	}
+	c.resetBuffer()
+	return tk
+}
+
+func (c *Context) lastToken() *token.Token {
+	if len(c.tokens) != 0 {
+		return c.tokens[len(c.tokens)-1]
+	}
+	return nil
+}

vendor/github.com/goccy/go-yaml/scanner/scanner.go

@@ -0,0 +1,903 @@
+package scanner
+
+import (
+	"io"
+	"strings"
+
+	"github.com/goccy/go-yaml/token"
+	"golang.org/x/xerrors"
+)
+
+// IndentState state for indent
+type IndentState int
+
+const (
+	// IndentStateEqual equals previous indent
+	IndentStateEqual IndentState = iota
+	// IndentStateUp more indent than previous
+	IndentStateUp
+	// IndentStateDown less indent than previous
+	IndentStateDown
+	// IndentStateKeep uses not indent token
+	IndentStateKeep
+)
+
+// Scanner holds the scanner's internal state while processing a given text.
+// It can be allocated as part of another data structure but must be initialized via Init before use.
+type Scanner struct {
+	source                 []rune
+	sourcePos              int
+	sourceSize             int
+	line                   int
+	column                 int
+	offset                 int
+	prevIndentLevel        int
+	prevIndentNum          int
+	prevIndentColumn       int
+	docStartColumn         int
+	indentLevel            int
+	indentNum              int
+	isFirstCharAtLine      bool
+	isAnchor               bool
+	startedFlowSequenceNum int
+	startedFlowMapNum      int
+	indentState            IndentState
+	savedPos               *token.Position
+}
+
+func (s *Scanner) pos() *token.Position {
+	return &token.Position{
+		Line:        s.line,
+		Column:      s.column,
+		Offset:      s.offset,
+		IndentNum:   s.indentNum,
+		IndentLevel: s.indentLevel,
+	}
+}
+
+func (s *Scanner) bufferedToken(ctx *Context) *token.Token {
+	if s.savedPos != nil {
+		tk := ctx.bufferedToken(s.savedPos)
+		s.savedPos = nil
+		return tk
+	}
+	line := s.line
+	column := s.column - len(ctx.buf)
+	level := s.indentLevel
+	if ctx.isSaveIndentMode() {
+		line -= s.newLineCount(ctx.buf)
+		column = strings.Index(string(ctx.obuf), string(ctx.buf)) + 1
+		// Since we are in a literal, folded or raw folded
+		// we can use the indent level from the last token.
+		last := ctx.lastToken()
+		if last != nil { // The last token should never be nil here.
+			level = last.Position.IndentLevel + 1
+		}
+	}
+	return ctx.bufferedToken(&token.Position{
+		Line:        line,
+		Column:      column,
+		Offset:      s.offset - len(ctx.buf),
+		IndentNum:   s.indentNum,
+		IndentLevel: level,
+	})
+}
+
+func (s *Scanner) progressColumn(ctx *Context, num int) {
+	s.column += num
+	s.offset += num
+	ctx.progress(num)
+}
+
+func (s *Scanner) progressLine(ctx *Context) {
+	s.column = 1
+	s.line++
+	s.offset++
+	s.indentNum = 0
+	s.isFirstCharAtLine = true
+	s.isAnchor = false
+	ctx.progress(1)
+}
+
+func (s *Scanner) isNeededKeepPreviousIndentNum(ctx *Context, c rune) bool {
+	if !s.isChangedToIndentStateUp() {
+		return false
+	}
+	if ctx.isDocument() {
+		return true
+	}
+	if c == '-' && ctx.existsBuffer() {
+		return true
+	}
+	return false
+}
+
+func (s *Scanner) isNewLineChar(c rune) bool {
+	if c == '\n' {
+		return true
+	}
+	if c == '\r' {
+		return true
+	}
+	return false
+}
+
+func (s *Scanner) newLineCount(src []rune) int {
+	size := len(src)
+	cnt := 0
+	for i := 0; i < size; i++ {
+		c := src[i]
+		switch c {
+		case '\r':
+			if i+1 < size && src[i+1] == '\n' {
+				i++
+			}
+			cnt++
+		case '\n':
+			cnt++
+		}
+	}
+	return cnt
+}
+
+func (s *Scanner) updateIndentState(ctx *Context) {
+	indentNumBasedIndentState := s.indentState
+	if s.prevIndentNum < s.indentNum {
+		s.indentLevel = s.prevIndentLevel + 1
+		indentNumBasedIndentState = IndentStateUp
+	} else if s.prevIndentNum == s.indentNum {
+		s.indentLevel = s.prevIndentLevel
+		indentNumBasedIndentState = IndentStateEqual
+	} else {
+		indentNumBasedIndentState = IndentStateDown
+		if s.prevIndentLevel > 0 {
+			s.indentLevel = s.prevIndentLevel - 1
+		}
+	}
+
+	if s.prevIndentColumn > 0 {
+		if s.prevIndentColumn < s.column {
+			s.indentState = IndentStateUp
+		} else if s.prevIndentColumn != s.column || indentNumBasedIndentState != IndentStateEqual {
+			// The following case ( current position is 'd' ), some variables becomes like here
+			// - prevIndentColumn: 1 of 'a'
+			// - indentNumBasedIndentState: IndentStateDown because d's indentNum(1) is less than c's indentNum(3).
+			// Therefore, s.prevIndentColumn(1) == s.column(1) is true, but we want to treat this as IndentStateDown.
+			// So, we look also current indentState value by the above prevIndentNum based logic, and determins finally indentState.
+			// ---
+			// a:
+			//   b
+			//   c
+			// d: e
+			// ^
+			s.indentState = IndentStateDown
+		} else {
+			s.indentState = IndentStateEqual
+		}
+	} else {
+		s.indentState = indentNumBasedIndentState
+	}
+}
+
+func (s *Scanner) updateIndent(ctx *Context, c rune) {
+	if s.isFirstCharAtLine && s.isNewLineChar(c) && ctx.isDocument() {
+		return
+	}
+	if s.isFirstCharAtLine && c == ' ' {
+		s.indentNum++
+		return
+	}
+	if !s.isFirstCharAtLine {
+		s.indentState = IndentStateKeep
+		return
+	}
+	s.updateIndentState(ctx)
+	s.isFirstCharAtLine = false
+	if s.isNeededKeepPreviousIndentNum(ctx, c) {
+		return
+	}
+	if s.indentState != IndentStateUp {
+		s.prevIndentColumn = 0
+	}
+	s.prevIndentNum = s.indentNum
+	s.prevIndentLevel = s.indentLevel
+}
+
+func (s *Scanner) isChangedToIndentStateDown() bool {
+	return s.indentState == IndentStateDown
+}
+
+func (s *Scanner) isChangedToIndentStateUp() bool {
+	return s.indentState == IndentStateUp
+}
+
+func (s *Scanner) isChangedToIndentStateEqual() bool {
+	return s.indentState == IndentStateEqual
+}
+
+func (s *Scanner) addBufferedTokenIfExists(ctx *Context) {
+	ctx.addToken(s.bufferedToken(ctx))
+}
+
+func (s *Scanner) breakLiteral(ctx *Context) {
+	s.docStartColumn = 0
+	ctx.breakLiteral()
+}
+
+func (s *Scanner) scanSingleQuote(ctx *Context) (tk *token.Token, pos int) {
+	ctx.addOriginBuf('\'')
+	srcpos := s.pos()
+	startIndex := ctx.idx + 1
+	src := ctx.src
+	size := len(src)
+	value := []rune{}
+	isFirstLineChar := false
+	isNewLine := false
+	for idx := startIndex; idx < size; idx++ {
+		if !isNewLine {
+			s.progressColumn(ctx, 1)
+		} else {
+			isNewLine = false
+		}
+		c := src[idx]
+		pos = idx + 1
+		ctx.addOriginBuf(c)
+		if s.isNewLineChar(c) {
+			value = append(value, ' ')
+			isFirstLineChar = true
+			isNewLine = true
+			s.progressLine(ctx)
+			continue
+		} else if c == ' ' && isFirstLineChar {
+			continue
+		} else if c != '\'' {
+			value = append(value, c)
+			isFirstLineChar = false
+			continue
+		}
+		if idx+1 < len(ctx.src) && ctx.src[idx+1] == '\'' {
+			// '' handle as ' character
+			value = append(value, c)
+			ctx.addOriginBuf(c)
+			idx++
+			continue
+		}
+		s.progressColumn(ctx, 1)
+		tk = token.SingleQuote(string(value), string(ctx.obuf), srcpos)
+		pos = idx - startIndex + 1
+		return
+	}
+	return
+}
+
+func hexToInt(b rune) int {
+	if b >= 'A' && b <= 'F' {
+		return int(b) - 'A' + 10
+	}
+	if b >= 'a' && b <= 'f' {
+		return int(b) - 'a' + 10
+	}
+	return int(b) - '0'
+}
+
+func hexRunesToInt(b []rune) int {
+	sum := 0
+	for i := 0; i < len(b); i++ {
+		sum += hexToInt(b[i]) << (uint(len(b)-i-1) * 4)
+	}
+	return sum
+}
+
+func (s *Scanner) scanDoubleQuote(ctx *Context) (tk *token.Token, pos int) {
+	ctx.addOriginBuf('"')
+	srcpos := s.pos()
+	startIndex := ctx.idx + 1
+	src := ctx.src
+	size := len(src)
+	value := []rune{}
+	isFirstLineChar := false
+	isNewLine := false
+	for idx := startIndex; idx < size; idx++ {
+		if !isNewLine {
+			s.progressColumn(ctx, 1)
+		} else {
+			isNewLine = false
+		}
+		c := src[idx]
+		pos = idx + 1
+		ctx.addOriginBuf(c)
+		if s.isNewLineChar(c) {
+			value = append(value, ' ')
+			isFirstLineChar = true
+			isNewLine = true
+			s.progressLine(ctx)
+			continue
+		} else if c == ' ' && isFirstLineChar {
+			continue
+		} else if c == '\\' {
+			isFirstLineChar = false
+			if idx+1 < size {
+				nextChar := src[idx+1]
+				switch nextChar {
+				case 'b':
+					ctx.addOriginBuf(nextChar)
+					value = append(value, '\b')
+					idx++
+					continue
+				case 'e':
+					ctx.addOriginBuf(nextChar)
+					value = append(value, '\x1B')
+					idx++
+					continue
+				case 'f':
+					ctx.addOriginBuf(nextChar)
+					value = append(value, '\f')
+					idx++
+					continue
+				case 'n':
+					ctx.addOriginBuf(nextChar)
+					value = append(value, '\n')
+					idx++
+					continue
+				case 'v':
+					ctx.addOriginBuf(nextChar)
+					value = append(value, '\v')
+					idx++
+					continue
+				case 'L': // LS (#x2028)
+					ctx.addOriginBuf(nextChar)
+					value = append(value, []rune{'\xE2', '\x80', '\xA8'}...)
+					idx++
+					continue
+				case 'N': // NEL (#x85)
+					ctx.addOriginBuf(nextChar)
+					value = append(value, []rune{'\xC2', '\x85'}...)
+					idx++
+					continue
+				case 'P': // PS (#x2029)
+					ctx.addOriginBuf(nextChar)
+					value = append(value, []rune{'\xE2', '\x80', '\xA9'}...)
+					idx++
+					continue
+				case '_': // #xA0
+					ctx.addOriginBuf(nextChar)
+					value = append(value, []rune{'\xC2', '\xA0'}...)
+					idx++
+					continue
+				case '"':
+					ctx.addOriginBuf(nextChar)
+					value = append(value, nextChar)
+					idx++
+					continue
+				case 'x':
+					if idx+3 >= size {
+						// TODO: need to return error
+						//err = xerrors.New("invalid escape character \\x")
+						return
+					}
+					codeNum := hexRunesToInt(src[idx+2 : idx+4])
+					value = append(value, rune(codeNum))
+					idx += 3
+					continue
+				case 'u':
+					if idx+5 >= size {
+						// TODO: need to return error
+						//err = xerrors.New("invalid escape character \\u")
+						return
+					}
+					codeNum := hexRunesToInt(src[idx+2 : idx+6])
+					value = append(value, rune(codeNum))
+					idx += 5
+					continue
+				case 'U':
+					if idx+9 >= size {
+						// TODO: need to return error
+						//err = xerrors.New("invalid escape character \\U")
+						return
+					}
+					codeNum := hexRunesToInt(src[idx+2 : idx+10])
+					value = append(value, rune(codeNum))
+					idx += 9
+					continue
+				case '\\':
+					ctx.addOriginBuf(nextChar)
+					idx++
+				}
+			}
+			value = append(value, c)
+			continue
+		} else if c != '"' {
+			value = append(value, c)
+			isFirstLineChar = false
+			continue
+		}
+		s.progressColumn(ctx, 1)
+		tk = token.DoubleQuote(string(value), string(ctx.obuf), srcpos)
+		pos = idx - startIndex + 1
+		return
+	}
+	return
+}
+
+func (s *Scanner) scanQuote(ctx *Context, ch rune) (tk *token.Token, pos int) {
+	if ch == '\'' {
+		return s.scanSingleQuote(ctx)
+	}
+	return s.scanDoubleQuote(ctx)
+}
+
+func (s *Scanner) isMergeKey(ctx *Context) bool {
+	if ctx.repeatNum('<') != 2 {
+		return false
+	}
+	src := ctx.src
+	size := len(src)
+	for idx := ctx.idx + 2; idx < size; idx++ {
+		c := src[idx]
+		if c == ' ' {
+			continue
+		}
+		if c != ':' {
+			return false
+		}
+		if idx+1 < size {
+			nc := src[idx+1]
+			if nc == ' ' || s.isNewLineChar(nc) {
+				return true
+			}
+		}
+	}
+	return false
+}
+
+func (s *Scanner) scanTag(ctx *Context) (tk *token.Token, pos int) {
+	ctx.addOriginBuf('!')
+	ctx.progress(1) // skip '!' character
+	for idx, c := range ctx.src[ctx.idx:] {
+		pos = idx + 1
+		ctx.addOriginBuf(c)
+		switch c {
+		case ' ', '\n', '\r':
+			value := ctx.source(ctx.idx-1, ctx.idx+idx)
+			tk = token.Tag(value, string(ctx.obuf), s.pos())
+			pos = len([]rune(value))
+			return
+		}
+	}
+	return
+}
+
+func (s *Scanner) scanComment(ctx *Context) (tk *token.Token, pos int) {
+	ctx.addOriginBuf('#')
+	ctx.progress(1) // skip '#' character
+	for idx, c := range ctx.src[ctx.idx:] {
+		pos = idx + 1
+		ctx.addOriginBuf(c)
+		switch c {
+		case '\n', '\r':
+			if ctx.previousChar() == '\\' {
+				continue
+			}
+			value := ctx.source(ctx.idx, ctx.idx+idx)
+			tk = token.Comment(value, string(ctx.obuf), s.pos())
+			pos = len([]rune(value)) + 1
+			return
+		}
+	}
+	// document ends with comment.
+	value := string(ctx.src[ctx.idx:])
+	tk = token.Comment(value, string(ctx.obuf), s.pos())
+	pos = len([]rune(value)) + 1
+	return
+}
+
+func trimCommentFromLiteralOpt(text string) (string, error) {
+	idx := strings.Index(text, "#")
+	if idx < 0 {
+		return text, nil
+	}
+	if idx == 0 {
+		return "", xerrors.New("invalid literal header")
+	}
+	return text[:idx-1], nil
+}
+
+func (s *Scanner) scanLiteral(ctx *Context, c rune) {
+	ctx.addOriginBuf(c)
+	if ctx.isEOS() {
+		if ctx.isLiteral {
+			ctx.addBuf(c)
+		}
+		value := ctx.bufferedSrc()
+		ctx.addToken(token.String(string(value), string(ctx.obuf), s.pos()))
+		ctx.resetBuffer()
+		s.progressColumn(ctx, 1)
+	} else if s.isNewLineChar(c) {
+		if ctx.isLiteral {
+			ctx.addBuf(c)
+		} else {
+			ctx.addBuf(' ')
+		}
+		s.progressLine(ctx)
+	} else if s.isFirstCharAtLine && c == ' ' {
+		if 0 < s.docStartColumn && s.docStartColumn <= s.column {
+			ctx.addBuf(c)
+		}
+		s.progressColumn(ctx, 1)
+	} else {
+		if s.docStartColumn == 0 {
+			s.docStartColumn = s.column
+		}
+		ctx.addBuf(c)
+		s.progressColumn(ctx, 1)
+	}
+}
+
+func (s *Scanner) scanLiteralHeader(ctx *Context) (pos int, err error) {
+	header := ctx.currentChar()
+	ctx.addOriginBuf(header)
+	ctx.progress(1) // skip '|' or '>' character
+	for idx, c := range ctx.src[ctx.idx:] {
+		pos = idx
+		ctx.addOriginBuf(c)
+		switch c {
+		case '\n', '\r':
+			value := ctx.source(ctx.idx, ctx.idx+idx)
+			opt := strings.TrimRight(value, " ")
+			orgOptLen := len(opt)
+			opt, err = trimCommentFromLiteralOpt(opt)
+			if err != nil {
+				return
+			}
+			switch opt {
+			case "", "+", "-",
+				"0", "1", "2", "3", "4", "5", "6", "7", "8", "9":
+				hasComment := len(opt) < orgOptLen
+				if header == '|' {
+					if hasComment {
+						commentLen := orgOptLen - len(opt)
+						headerPos := strings.Index(string(ctx.obuf), "|")
+						litBuf := ctx.obuf[:len(ctx.obuf)-commentLen-headerPos]
+						commentBuf := ctx.obuf[len(litBuf):]
+						ctx.addToken(token.Literal("|"+opt, string(litBuf), s.pos()))
+						s.column += len(litBuf)
+						s.offset += len(litBuf)
+						commentHeader := strings.Index(value, "#")
+						ctx.addToken(token.Comment(string(value[commentHeader+1:]), string(commentBuf), s.pos()))
+					} else {
+						ctx.addToken(token.Literal("|"+opt, string(ctx.obuf), s.pos()))
+					}
+					ctx.isLiteral = true
+				} else if header == '>' {
+					if hasComment {
+						commentLen := orgOptLen - len(opt)
+						headerPos := strings.Index(string(ctx.obuf), ">")
+						foldedBuf := ctx.obuf[:len(ctx.obuf)-commentLen-headerPos]
+						commentBuf := ctx.obuf[len(foldedBuf):]
+						ctx.addToken(token.Folded(">"+opt, string(foldedBuf), s.pos()))
+						s.column += len(foldedBuf)
+						s.offset += len(foldedBuf)
+						commentHeader := strings.Index(value, "#")
+						ctx.addToken(token.Comment(string(value[commentHeader+1:]), string(commentBuf), s.pos()))
+					} else {
+						ctx.addToken(token.Folded(">"+opt, string(ctx.obuf), s.pos()))
+					}
+					ctx.isFolded = true
+				}
+				s.indentState = IndentStateKeep
+				ctx.resetBuffer()
+				ctx.literalOpt = opt
+				return
+			}
+			break
+		}
+	}
+	err = xerrors.New("invalid literal header")
+	return
+}
+
+func (s *Scanner) scanNewLine(ctx *Context, c rune) {
+	if len(ctx.buf) > 0 && s.savedPos == nil {
+		s.savedPos = s.pos()
+		s.savedPos.Column -= len(ctx.bufferedSrc())
+	}
+
+	// if the following case, origin buffer has unnecessary two spaces.
+	// So, `removeRightSpaceFromOriginBuf` remove them, also fix column number too.
+	// ---
+	// a:[space][space]
+	//   b: c
+	removedNum := ctx.removeRightSpaceFromBuf()
+	if removedNum > 0 {
+		s.column -= removedNum
+		s.offset -= removedNum
+		if s.savedPos != nil {
+			s.savedPos.Column -= removedNum
+		}
+	}
+
+	if ctx.isEOS() {
+		s.addBufferedTokenIfExists(ctx)
+	} else if s.isAnchor {
+		s.addBufferedTokenIfExists(ctx)
+	}
+	ctx.addBuf(' ')
+	ctx.addOriginBuf(c)
+	ctx.isSingleLine = false
+	s.progressLine(ctx)
+}
+
+func (s *Scanner) scan(ctx *Context) (pos int) {
+	for ctx.next() {
+		pos = ctx.nextPos()
+		c := ctx.currentChar()
+		s.updateIndent(ctx, c)
+		if ctx.isDocument() {
+			if s.isChangedToIndentStateEqual() ||
+				s.isChangedToIndentStateDown() {
+				s.addBufferedTokenIfExists(ctx)
+				s.breakLiteral(ctx)
+			} else {
+				s.scanLiteral(ctx, c)
+				continue
+			}
+		} else if s.isChangedToIndentStateDown() {
+			s.addBufferedTokenIfExists(ctx)
+		} else if s.isChangedToIndentStateEqual() {
+			// if first character is new line character, buffer expect to raw folded literal
+			if len(ctx.obuf) > 0 && s.newLineCount(ctx.obuf) <= 1 {
+				// doesn't raw folded literal
+				s.addBufferedTokenIfExists(ctx)
+			}
+		}
+		switch c {
+		case '{':
+			if !ctx.existsBuffer() {
+				ctx.addOriginBuf(c)
+				ctx.addToken(token.MappingStart(string(ctx.obuf), s.pos()))
+				s.startedFlowMapNum++
+				s.progressColumn(ctx, 1)
+				return
+			}
+		case '}':
+			if !ctx.existsBuffer() || s.startedFlowMapNum > 0 {
+				ctx.addToken(s.bufferedToken(ctx))
+				ctx.addOriginBuf(c)
+				ctx.addToken(token.MappingEnd(string(ctx.obuf), s.pos()))
+				s.startedFlowMapNum--
+				s.progressColumn(ctx, 1)
+				return
+			}
+		case '.':
+			if s.indentNum == 0 && s.column == 1 && ctx.repeatNum('.') == 3 {
+				ctx.addToken(token.DocumentEnd(string(ctx.obuf)+"...", s.pos()))
+				s.progressColumn(ctx, 3)
+				pos += 2
+				return
+			}
+		case '<':
+			if s.isMergeKey(ctx) {
+				s.prevIndentColumn = s.column
+				ctx.addToken(token.MergeKey(string(ctx.obuf)+"<<", s.pos()))
+				s.progressColumn(ctx, 1)
+				pos++
+				return
+			}
+		case '-':
+			if s.indentNum == 0 && s.column == 1 && ctx.repeatNum('-') == 3 {
+				s.addBufferedTokenIfExists(ctx)
+				ctx.addToken(token.DocumentHeader(string(ctx.obuf)+"---", s.pos()))
+				s.progressColumn(ctx, 3)
+				pos += 2
+				return
+			}
+			if ctx.existsBuffer() && s.isChangedToIndentStateUp() {
+				// raw folded
+				ctx.isRawFolded = true
+				ctx.addBuf(c)
+				ctx.addOriginBuf(c)
+				s.progressColumn(ctx, 1)
+				continue
+			}
+			if ctx.existsBuffer() {
+				// '-' is literal
+				ctx.addBuf(c)
+				ctx.addOriginBuf(c)
+				s.progressColumn(ctx, 1)
+				continue
+			}
+			nc := ctx.nextChar()
+			if nc == ' ' || s.isNewLineChar(nc) {
+				s.addBufferedTokenIfExists(ctx)
+				ctx.addOriginBuf(c)
+				tk := token.SequenceEntry(string(ctx.obuf), s.pos())
+				s.prevIndentColumn = tk.Position.Column
+				ctx.addToken(tk)
+				s.progressColumn(ctx, 1)
+				return
+			}
+		case '[':
+			if !ctx.existsBuffer() {
+				ctx.addOriginBuf(c)
+				ctx.addToken(token.SequenceStart(string(ctx.obuf), s.pos()))
+				s.startedFlowSequenceNum++
+				s.progressColumn(ctx, 1)
+				return
+			}
+		case ']':
+			if !ctx.existsBuffer() || s.startedFlowSequenceNum > 0 {
+				s.addBufferedTokenIfExists(ctx)
+				ctx.addOriginBuf(c)
+				ctx.addToken(token.SequenceEnd(string(ctx.obuf), s.pos()))
+				s.startedFlowSequenceNum--
+				s.progressColumn(ctx, 1)
+				return
+			}
+		case ',':
+			if s.startedFlowSequenceNum > 0 || s.startedFlowMapNum > 0 {
+				s.addBufferedTokenIfExists(ctx)
+				ctx.addOriginBuf(c)
+				ctx.addToken(token.CollectEntry(string(ctx.obuf), s.pos()))
+				s.progressColumn(ctx, 1)
+				return
+			}
+		case ':':
+			nc := ctx.nextChar()
+			if s.startedFlowMapNum > 0 || nc == ' ' || s.isNewLineChar(nc) || ctx.isNextEOS() {
+				// mapping value
+				tk := s.bufferedToken(ctx)
+				if tk != nil {
+					s.prevIndentColumn = tk.Position.Column
+					ctx.addToken(tk)
+				} else if tk := ctx.lastToken(); tk != nil {
+					// If the map key is quote, the buffer does not exist because it has already been cut into tokens.
+					// Therefore, we need to check the last token.
+					if tk.Indicator == token.QuotedScalarIndicator {
+						s.prevIndentColumn = tk.Position.Column
+					}
+				}
+				ctx.addToken(token.MappingValue(s.pos()))
+				s.progressColumn(ctx, 1)
+				return
+			}
+		case '|', '>':
+			if !ctx.existsBuffer() {
+				progress, err := s.scanLiteralHeader(ctx)
+				if err != nil {
+					// TODO: returns syntax error object
+					return
+				}
+				s.progressColumn(ctx, progress)
+				s.progressLine(ctx)
+				continue
+			}
+		case '!':
+			if !ctx.existsBuffer() {
+				token, progress := s.scanTag(ctx)
+				ctx.addToken(token)
+				s.progressColumn(ctx, progress)
+				if c := ctx.previousChar(); s.isNewLineChar(c) {
+					s.progressLine(ctx)
+				}
+				pos += progress
+				return
+			}
+		case '%':
+			if !ctx.existsBuffer() && s.indentNum == 0 {
+				ctx.addToken(token.Directive(string(ctx.obuf)+"%", s.pos()))
+				s.progressColumn(ctx, 1)
+				return
+			}
+		case '?':
+			nc := ctx.nextChar()
+			if !ctx.existsBuffer() && nc == ' ' {
+				ctx.addToken(token.MappingKey(s.pos()))
+				s.progressColumn(ctx, 1)
+				return
+			}
+		case '&':
+			if !ctx.existsBuffer() {
+				s.addBufferedTokenIfExists(ctx)
+				ctx.addOriginBuf(c)
+				ctx.addToken(token.Anchor(string(ctx.obuf), s.pos()))
+				s.progressColumn(ctx, 1)
+				s.isAnchor = true
+				return
+			}
+		case '*':
+			if !ctx.existsBuffer() {
+				s.addBufferedTokenIfExists(ctx)
+				ctx.addOriginBuf(c)
+				ctx.addToken(token.Alias(string(ctx.obuf), s.pos()))
+				s.progressColumn(ctx, 1)
+				return
+			}
+		case '#':
+			if !ctx.existsBuffer() || ctx.previousChar() == ' ' {
+				s.addBufferedTokenIfExists(ctx)
+				token, progress := s.scanComment(ctx)
+				ctx.addToken(token)
+				s.progressColumn(ctx, progress)
+				s.progressLine(ctx)
+				pos += progress
+				return
+			}
+		case '\'', '"':
+			if !ctx.existsBuffer() {
+				token, progress := s.scanQuote(ctx, c)
+				ctx.addToken(token)
+				pos += progress
+				// If the non-whitespace character immediately following the quote is ':', the quote should be treated as a map key.
+				// Therefore, do not return and continue processing as a normal map key.
+				if ctx.currentCharWithSkipWhitespace() == ':' {
+					continue
+				}
+				return
+			}
+		case '\r', '\n':
+			// There is no problem that we ignore CR which followed by LF and normalize it to LF, because of following YAML1.2 spec.
+			// > Line breaks inside scalar content must be normalized by the YAML processor. Each such line break must be parsed into a single line feed character.
+			// > Outside scalar content, YAML allows any line break to be used to terminate lines.
+			// > -- https://yaml.org/spec/1.2/spec.html
+			if c == '\r' && ctx.nextChar() == '\n' {
+				ctx.addOriginBuf('\r')
+				ctx.progress(1)
+				c = '\n'
+			}
+			s.scanNewLine(ctx, c)
+			continue
+		case ' ':
+			if ctx.isSaveIndentMode() || (!s.isAnchor && !s.isFirstCharAtLine) {
+				ctx.addBuf(c)
+				ctx.addOriginBuf(c)
+				s.progressColumn(ctx, 1)
+				continue
+			}
+			if s.isFirstCharAtLine {
+				s.progressColumn(ctx, 1)
+				ctx.addOriginBuf(c)
+				continue
+			}
+			s.addBufferedTokenIfExists(ctx)
+			pos-- // to rescan white space at next scanning for adding white space to next buffer.
+			s.isAnchor = false
+			return
+		}
+		ctx.addBuf(c)
+		ctx.addOriginBuf(c)
+		s.progressColumn(ctx, 1)
+	}
+	s.addBufferedTokenIfExists(ctx)
+	return
+}
+
+// Init prepares the scanner s to tokenize the text src by setting the scanner at the beginning of src.
+func (s *Scanner) Init(text string) {
+	src := []rune(text)
+	s.source = src
+	s.sourcePos = 0
+	s.sourceSize = len(src)
+	s.line = 1
+	s.column = 1
+	s.offset = 1
+	s.prevIndentLevel = 0
+	s.prevIndentNum = 0
+	s.prevIndentColumn = 0
+	s.indentLevel = 0
+	s.indentNum = 0
+	s.isFirstCharAtLine = true
+}
+
+// Scan scans the next token and returns the token collection. The source end is indicated by io.EOF.
+func (s *Scanner) Scan() (token.Tokens, error) {
+	if s.sourcePos >= s.sourceSize {
+		return nil, io.EOF
+	}
+	ctx := newContext(s.source[s.sourcePos:])
+	defer ctx.release()
+	progress := s.scan(ctx)
+	s.sourcePos += progress
+	var tokens token.Tokens
+	tokens = append(tokens, ctx.tokens...)
+	return tokens, nil
+}

vendor/github.com/goccy/go-yaml/stdlib_quote.go

@@ -0,0 +1,103 @@
+// Copied and trimmed down from https://github.com/golang/go/blob/e3769299cd3484e018e0e2a6e1b95c2b18ce4f41/src/strconv/quote.go
+// We want to use the standard library's private "quoteWith" function rather than write our own so that we get robust unicode support.
+// Every private function called by quoteWith was copied.
+// There are 2 modifications to simplify the code:
+// 1. The unicode.IsPrint function was substituted for the custom implementation of IsPrint
+// 2. All code paths reachable only when ASCIIonly or grphicOnly are set to true were removed.
+
+// Copyright 2009 The Go Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style
+// license that can be found in the LICENSE file.
+
+package yaml
+
+import (
+	"unicode"
+	"unicode/utf8"
+)
+
+const (
+	lowerhex = "0123456789abcdef"
+)
+
+func quoteWith(s string, quote byte) string {
+	return string(appendQuotedWith(make([]byte, 0, 3*len(s)/2), s, quote))
+}
+
+func appendQuotedWith(buf []byte, s string, quote byte) []byte {
+	// Often called with big strings, so preallocate. If there's quoting,
+	// this is conservative but still helps a lot.
+	if cap(buf)-len(buf) < len(s) {
+		nBuf := make([]byte, len(buf), len(buf)+1+len(s)+1)
+		copy(nBuf, buf)
+		buf = nBuf
+	}
+	buf = append(buf, quote)
+	for width := 0; len(s) > 0; s = s[width:] {
+		r := rune(s[0])
+		width = 1
+		if r >= utf8.RuneSelf {
+			r, width = utf8.DecodeRuneInString(s)
+		}
+		if width == 1 && r == utf8.RuneError {
+			buf = append(buf, `\x`...)
+			buf = append(buf, lowerhex[s[0]>>4])
+			buf = append(buf, lowerhex[s[0]&0xF])
+			continue
+		}
+		buf = appendEscapedRune(buf, r, quote)
+	}
+	buf = append(buf, quote)
+	return buf
+}
+
+func appendEscapedRune(buf []byte, r rune, quote byte) []byte {
+	var runeTmp [utf8.UTFMax]byte
+	if r == rune(quote) || r == '\\' { // always backslashed
+		buf = append(buf, '\\')
+		buf = append(buf, byte(r))
+		return buf
+	}
+	if unicode.IsPrint(r) {
+		n := utf8.EncodeRune(runeTmp[:], r)
+		buf = append(buf, runeTmp[:n]...)
+		return buf
+	}
+	switch r {
+	case '\a':
+		buf = append(buf, `\a`...)
+	case '\b':
+		buf = append(buf, `\b`...)
+	case '\f':
+		buf = append(buf, `\f`...)
+	case '\n':
+		buf = append(buf, `\n`...)
+	case '\r':
+		buf = append(buf, `\r`...)
+	case '\t':
+		buf = append(buf, `\t`...)
+	case '\v':
+		buf = append(buf, `\v`...)
+	default:
+		switch {
+		case r < ' ':
+			buf = append(buf, `\x`...)
+			buf = append(buf, lowerhex[byte(r)>>4])
+			buf = append(buf, lowerhex[byte(r)&0xF])
+		case r > utf8.MaxRune:
+			r = 0xFFFD
+			fallthrough
+		case r < 0x10000:
+			buf = append(buf, `\u`...)
+			for s := 12; s >= 0; s -= 4 {
+				buf = append(buf, lowerhex[r>>uint(s)&0xF])
+			}
+		default:
+			buf = append(buf, `\U`...)
+			for s := 28; s >= 0; s -= 4 {
+				buf = append(buf, lowerhex[r>>uint(s)&0xF])
+			}
+		}
+	}
+	return buf
+}

vendor/github.com/goccy/go-yaml/struct.go

@@ -0,0 +1,130 @@
+package yaml
+
+import (
+	"reflect"
+	"strings"
+
+	"golang.org/x/xerrors"
+)
+
+const (
+	// StructTagName tag keyword for Marshal/Unmarshal
+	StructTagName = "yaml"
+)
+
+// StructField information for each the field in structure
+type StructField struct {
+	FieldName    string
+	RenderName   string
+	AnchorName   string
+	AliasName    string
+	IsAutoAnchor bool
+	IsAutoAlias  bool
+	IsOmitEmpty  bool
+	IsFlow       bool
+	IsInline     bool
+}
+
+func getTag(field reflect.StructField) string {
+	// If struct tag `yaml` exist, use that. If no `yaml`
+	// exists, but `json` does, use that and try the best to
+	// adhere to its rules
+	tag := field.Tag.Get(StructTagName)
+	if tag == "" {
+		tag = field.Tag.Get(`json`)
+	}
+	return tag
+}
+
+func structField(field reflect.StructField) *StructField {
+	tag := getTag(field)
+	fieldName := strings.ToLower(field.Name)
+	options := strings.Split(tag, ",")
+	if len(options) > 0 {
+		if options[0] != "" {
+			fieldName = options[0]
+		}
+	}
+	structField := &StructField{
+		FieldName:  field.Name,
+		RenderName: fieldName,
+	}
+	if len(options) > 1 {
+		for _, opt := range options[1:] {
+			switch {
+			case opt == "omitempty":
+				structField.IsOmitEmpty = true
+			case opt == "flow":
+				structField.IsFlow = true
+			case opt == "inline":
+				structField.IsInline = true
+			case strings.HasPrefix(opt, "anchor"):
+				anchor := strings.Split(opt, "=")
+				if len(anchor) > 1 {
+					structField.AnchorName = anchor[1]
+				} else {
+					structField.IsAutoAnchor = true
+				}
+			case strings.HasPrefix(opt, "alias"):
+				alias := strings.Split(opt, "=")
+				if len(alias) > 1 {
+					structField.AliasName = alias[1]
+				} else {
+					structField.IsAutoAlias = true
+				}
+			default:
+			}
+		}
+	}
+	return structField
+}
+
+func isIgnoredStructField(field reflect.StructField) bool {
+	if field.PkgPath != "" && !field.Anonymous {
+		// private field
+		return true
+	}
+	tag := getTag(field)
+	if tag == "-" {
+		return true
+	}
+	return false
+}
+
+type StructFieldMap map[string]*StructField
+
+func (m StructFieldMap) isIncludedRenderName(name string) bool {
+	for _, v := range m {
+		if !v.IsInline && v.RenderName == name {
+			return true
+		}
+	}
+	return false
+}
+
+func (m StructFieldMap) hasMergeProperty() bool {
+	for _, v := range m {
+		if v.IsOmitEmpty && v.IsInline && v.IsAutoAlias {
+			return true
+		}
+	}
+	return false
+}
+
+func structFieldMap(structType reflect.Type) (StructFieldMap, error) {
+	structFieldMap := StructFieldMap{}
+	renderNameMap := map[string]struct{}{}
+	for i := 0; i < structType.NumField(); i++ {
+		field := structType.Field(i)
+		if isIgnoredStructField(field) {
+			continue
+		}
+		structField := structField(field)
+		if _, exists := renderNameMap[structField.RenderName]; exists {
+			return nil, xerrors.Errorf("duplicated struct field name %s", structField.RenderName)
+		}
+		structFieldMap[structField.FieldName] = structField
+		renderNameMap[structField.RenderName] = struct{}{}
+	}
+	return structFieldMap, nil
+}

vendor/github.com/goccy/go-yaml/validate.go

@@ -0,0 +1,13 @@
+package yaml
+
+// StructValidator need to implement Struct method only
+// ( see https://pkg.go.dev/github.com/go-playground/validator/v10#Validate.Struct )
+type StructValidator interface {
+	Struct(interface{}) error
+}
+
+// FieldError need to implement StructField method only
+// ( see https://pkg.go.dev/github.com/go-playground/validator/v10#FieldError )
+type FieldError interface {
+	StructField() string
+}

vendor/github.com/goccy/go-yaml/yaml.go

@@ -0,0 +1,290 @@
+package yaml
+
+import (
+	"bytes"
+	"context"
+	"io"
+	"reflect"
+	"sync"
+
+	"github.com/goccy/go-yaml/ast"
+	"github.com/goccy/go-yaml/internal/errors"
+	"golang.org/x/xerrors"
+)
+
+// BytesMarshaler interface may be implemented by types to customize their
+// behavior when being marshaled into a YAML document. The returned value
+// is marshaled in place of the original value implementing Marshaler.
+//
+// If an error is returned by MarshalYAML, the marshaling procedure stops
+// and returns with the provided error.
+type BytesMarshaler interface {
+	MarshalYAML() ([]byte, error)
+}
+
+// BytesMarshalerContext interface use BytesMarshaler with context.Context.
+type BytesMarshalerContext interface {
+	MarshalYAML(context.Context) ([]byte, error)
+}
+
+// InterfaceMarshaler interface has MarshalYAML compatible with github.com/go-yaml/yaml package.
+type InterfaceMarshaler interface {
+	MarshalYAML() (interface{}, error)
+}
+
+// InterfaceMarshalerContext interface use InterfaceMarshaler with context.Context.
+type InterfaceMarshalerContext interface {
+	MarshalYAML(context.Context) (interface{}, error)
+}
+
+// BytesUnmarshaler interface may be implemented by types to customize their
+// behavior when being unmarshaled from a YAML document.
+type BytesUnmarshaler interface {
+	UnmarshalYAML([]byte) error
+}
+
+// BytesUnmarshalerContext interface use BytesUnmarshaler with context.Context.
+type BytesUnmarshalerContext interface {
+	UnmarshalYAML(context.Context, []byte) error
+}
+
+// InterfaceUnmarshaler interface has UnmarshalYAML compatible with github.com/go-yaml/yaml package.
+type InterfaceUnmarshaler interface {
+	UnmarshalYAML(func(interface{}) error) error
+}
+
+// InterfaceUnmarshalerContext interface use InterfaceUnmarshaler with context.Context.
+type InterfaceUnmarshalerContext interface {
+	UnmarshalYAML(context.Context, func(interface{}) error) error
+}
+
+// MapItem is an item in a MapSlice.
+type MapItem struct {
+	Key, Value interface{}
+}
+
+// MapSlice encodes and decodes as a YAML map.
+// The order of keys is preserved when encoding and decoding.
+type MapSlice []MapItem
+
+// ToMap convert to map[interface{}]interface{}.
+func (s MapSlice) ToMap() map[interface{}]interface{} {
+	v := map[interface{}]interface{}{}
+	for _, item := range s {
+		v[item.Key] = item.Value
+	}
+	return v
+}
+
+// Marshal serializes the value provided into a YAML document. The structure
+// of the generated document will reflect the structure of the value itself.
+// Maps and pointers (to struct, string, int, etc) are accepted as the in value.
+//
+// Struct fields are only marshalled if they are exported (have an upper case
+// first letter), and are marshalled using the field name lowercased as the
+// default key. Custom keys may be defined via the "yaml" name in the field
+// tag: the content preceding the first comma is used as the key, and the
+// following comma-separated options are used to tweak the marshalling process.
+// Conflicting names result in a runtime error.
+//
+// The field tag format accepted is:
+//
+//     `(...) yaml:"[<key>][,<flag1>[,<flag2>]]" (...)`
+//
+// The following flags are currently supported:
+//
+//     omitempty    Only include the field if it's not set to the zero
+//                  value for the type or to empty slices or maps.
+//                  Zero valued structs will be omitted if all their public
+//                  fields are zero, unless they implement an IsZero
+//                  method (see the IsZeroer interface type), in which
+//                  case the field will be included if that method returns true.
+//
+//     flow         Marshal using a flow style (useful for structs,
+//                  sequences and maps).
+//
+//     inline       Inline the field, which must be a struct or a map,
+//                  causing all of its fields or keys to be processed as if
+//                  they were part of the outer struct. For maps, keys must
+//                  not conflict with the yaml keys of other struct fields.
+//
+//     anchor       Marshal with anchor. If want to define anchor name explicitly, use anchor=name style.
+//                  Otherwise, if used 'anchor' name only, used the field name lowercased as the anchor name
+//
+//     alias        Marshal with alias. If want to define alias name explicitly, use alias=name style.
+//                  Otherwise, If omitted alias name and the field type is pointer type,
+//                  assigned anchor name automatically from same pointer address.
+//
+// In addition, if the key is "-", the field is ignored.
+//
+// For example:
+//
+//     type T struct {
+//         F int `yaml:"a,omitempty"`
+//         B int
+//     }
+//     yaml.Marshal(&T{B: 2}) // Returns "b: 2\n"
+//     yaml.Marshal(&T{F: 1}) // Returns "a: 1\nb: 0\n"
+//
+func Marshal(v interface{}) ([]byte, error) {
+	return MarshalWithOptions(v)
+}
+
+// MarshalWithOptions serializes the value provided into a YAML document with EncodeOptions.
+func MarshalWithOptions(v interface{}, opts ...EncodeOption) ([]byte, error) {
+	return MarshalContext(context.Background(), v, opts...)
+}
+
+// MarshalContext serializes the value provided into a YAML document with context.Context and EncodeOptions.
+func MarshalContext(ctx context.Context, v interface{}, opts ...EncodeOption) ([]byte, error) {
+	var buf bytes.Buffer
+	if err := NewEncoder(&buf, opts...).EncodeContext(ctx, v); err != nil {
+		return nil, errors.Wrapf(err, "failed to marshal")
+	}
+	return buf.Bytes(), nil
+}
+
+// ValueToNode convert from value to ast.Node.
+func ValueToNode(v interface{}, opts ...EncodeOption) (ast.Node, error) {
+	var buf bytes.Buffer
+	node, err := NewEncoder(&buf, opts...).EncodeToNode(v)
+	if err != nil {
+		return nil, errors.Wrapf(err, "failed to convert value to node")
+	}
+	return node, nil
+}
+
+// Unmarshal decodes the first document found within the in byte slice
+// and assigns decoded values into the out value.
+//
+// Struct fields are only unmarshalled if they are exported (have an
+// upper case first letter), and are unmarshalled using the field name
+// lowercased as the default key. Custom keys may be defined via the
+// "yaml" name in the field tag: the content preceding the first comma
+// is used as the key, and the following comma-separated options are
+// used to tweak the marshalling process (see Marshal).
+// Conflicting names result in a runtime error.
+//
+// For example:
+//
+//     type T struct {
+//         F int `yaml:"a,omitempty"`
+//         B int
+//     }
+//     var t T
+//     yaml.Unmarshal([]byte("a: 1\nb: 2"), &t)
+//
+// See the documentation of Marshal for the format of tags and a list of
+// supported tag options.
+//
+func Unmarshal(data []byte, v interface{}) error {
+	return UnmarshalWithOptions(data, v)
+}
+
+// UnmarshalWithOptions decodes with DecodeOptions the first document found within the in byte slice
+// and assigns decoded values into the out value.
+func UnmarshalWithOptions(data []byte, v interface{}, opts ...DecodeOption) error {
+	return UnmarshalContext(context.Background(), data, v, opts...)
+}
+
+// UnmarshalContext decodes with context.Context and DecodeOptions.
+func UnmarshalContext(ctx context.Context, data []byte, v interface{}, opts ...DecodeOption) error {
+	dec := NewDecoder(bytes.NewBuffer(data), opts...)
+	if err := dec.DecodeContext(ctx, v); err != nil {
+		if err == io.EOF {
+			return nil
+		}
+		return errors.Wrapf(err, "failed to unmarshal")
+	}
+	return nil
+}
+
+// NodeToValue converts node to the value pointed to by v.
+func NodeToValue(node ast.Node, v interface{}, opts ...DecodeOption) error {
+	var buf bytes.Buffer
+	if err := NewDecoder(&buf, opts...).DecodeFromNode(node, v); err != nil {
+		return errors.Wrapf(err, "failed to convert node to value")
+	}
+	return nil
+}
+
+// FormatError is a utility function that takes advantage of the metadata
+// stored in the errors returned by this package's parser.
+//
+// If the second argument `colored` is true, the error message is colorized.
+// If the third argument `inclSource` is true, the error message will
+// contain snippets of the YAML source that was used.
+func FormatError(e error, colored, inclSource bool) string {
+	var pp errors.PrettyPrinter
+	if xerrors.As(e, &pp) {
+		var buf bytes.Buffer
+		pp.PrettyPrint(&errors.Sink{&buf}, colored, inclSource)
+		return buf.String()
+	}
+
+	return e.Error()
+}
+
+// YAMLToJSON convert YAML bytes to JSON.
+func YAMLToJSON(bytes []byte) ([]byte, error) {
+	var v interface{}
+	if err := UnmarshalWithOptions(bytes, &v, UseOrderedMap()); err != nil {
+		return nil, errors.Wrapf(err, "failed to unmarshal")
+	}
+	out, err := MarshalWithOptions(v, JSON())
+	if err != nil {
+		return nil, errors.Wrapf(err, "failed to marshal with json option")
+	}
+	return out, nil
+}
+
+// JSONToYAML convert JSON bytes to YAML.
+func JSONToYAML(bytes []byte) ([]byte, error) {
+	var v interface{}
+	if err := UnmarshalWithOptions(bytes, &v, UseOrderedMap()); err != nil {
+		return nil, errors.Wrapf(err, "failed to unmarshal from json bytes")
+	}
+	out, err := Marshal(v)
+	if err != nil {
+		return nil, errors.Wrapf(err, "failed to marshal")
+	}
+	return out, nil
+}
+
+var (
+	globalCustomMarshalerMu    sync.Mutex
+	globalCustomUnmarshalerMu  sync.Mutex
+	globalCustomMarshalerMap   = map[reflect.Type]func(interface{}) ([]byte, error){}
+	globalCustomUnmarshalerMap = map[reflect.Type]func(interface{}, []byte) error{}
+)
+
+// RegisterCustomMarshaler overrides any encoding process for the type specified in generics.
+// If you want to switch the behavior for each encoder, use `CustomMarshaler` defined as EncodeOption.
+//
+// NOTE: If type T implements MarshalYAML for pointer receiver, the type specified in RegisterCustomMarshaler must be *T.
+// If RegisterCustomMarshaler and CustomMarshaler of EncodeOption are specified for the same type,
+// the CustomMarshaler specified in EncodeOption takes precedence.
+func RegisterCustomMarshaler[T any](marshaler func(T) ([]byte, error)) {
+	globalCustomMarshalerMu.Lock()
+	defer globalCustomMarshalerMu.Unlock()
+
+	var typ T
+	globalCustomMarshalerMap[reflect.TypeOf(typ)] = func(v interface{}) ([]byte, error) {
+		return marshaler(v.(T))
+	}
+}
+
+// RegisterCustomUnmarshaler overrides any decoding process for the type specified in generics.
+// If you want to switch the behavior for each decoder, use `CustomUnmarshaler` defined as DecodeOption.
+//
+// NOTE: If RegisterCustomUnmarshaler and CustomUnmarshaler of DecodeOption are specified for the same type,
+// the CustomUnmarshaler specified in DecodeOption takes precedence.
+func RegisterCustomUnmarshaler[T any](unmarshaler func(*T, []byte) error) {
+	globalCustomUnmarshalerMu.Lock()
+	defer globalCustomUnmarshalerMu.Unlock()
+
+	var typ *T
+	globalCustomUnmarshalerMap[reflect.TypeOf(typ)] = func(v interface{}, b []byte) error {
+		return unmarshaler(v.(*T), b)
+	}
+}