Added XML encoding/decoding (#1067)

* Added XML encoding/decoding

* Minor fixes

* Improve yq doc

* Lint

acceptance_tests/output-format.sh

@@ -142,4 +142,25 @@ EOM
   assertEquals "$expected" "$X"
 }
 
+testOutputXmComplex() {
+  cat >test.yml <<EOL
+a: {b: {c: ["cat", "dog"], +f: meow}}
+EOL
+
+  read -r -d '' expected << EOM
+<a>
+  <b f="meow">
+    <c>cat</c>
+    <c>dog</c>
+  </b>
+</a>
+EOM
+
+  X=$(./yq e --output-format=x test.yml)
+  assertEquals "$expected" "$X"
+
+  X=$(./yq ea --output-format=x test.yml)
+  assertEquals "$expected" "$X"
+}
+
 source ./scripts/shunit2

cmd/evaluate_all_command.go

@@ -85,8 +85,9 @@ func evaluateAll(cmd *cobra.Command, args []string) (cmdError error) {
 	}
 
 	printerWriter := configurePrinterWriter(format, out)
+	encoder := configureEncoder(format)
 
-	printer := yqlib.NewPrinter(printerWriter, format, unwrapScalar, colorsEnabled, indent, !noDocSeparators)
+	printer := yqlib.NewPrinter(encoder, printerWriter)
 
 	if frontMatter != "" {
 		frontMatterHandler := yqlib.NewFrontMatterHandler(args[firstFileIndex])

cmd/evalute_sequence_command.go

@@ -93,8 +93,9 @@ func evaluateSequence(cmd *cobra.Command, args []string) (cmdError error) {
 	}
 
 	printerWriter := configurePrinterWriter(format, out)
+	encoder := configureEncoder(format)
 
-	printer := yqlib.NewPrinter(printerWriter, format, unwrapScalar, colorsEnabled, indent, !noDocSeparators)
+	printer := yqlib.NewPrinter(encoder, printerWriter)
 
 	decoder, err := configureDecoder()
 	if err != nil {

cmd/root.go

@@ -52,7 +52,7 @@ See https://mikefarah.gitbook.io/yq/ for detailed documentation and examples.`,
 	}
 
 	rootCmd.PersistentFlags().StringVarP(&outputFormat, "output-format", "o", "yaml", "[yaml|y|json|j|props|p|xml|x] output format type.")
-	rootCmd.PersistentFlags().StringVarP(&inputFormat, "input-format", "p", "yaml", "[yaml|y|xml|x] input format type. Note that json is a subset of yaml.")
+	rootCmd.PersistentFlags().StringVarP(&inputFormat, "input-format", "p", "yaml", "[yaml|y|xml|x] parse format for input. Note that json is a subset of yaml.")
 
 	rootCmd.PersistentFlags().StringVar(&xmlAttributePrefix, "xml-attribute-prefix", "+", "prefix for xml attributes")
 	rootCmd.PersistentFlags().StringVar(&xmlContentName, "xml-content-name", "+content", "name for xml content (if no attribute name is present).")

cmd/utils.go

@@ -73,3 +73,21 @@ func configurePrinterWriter(format yqlib.PrinterOutputFormat, out io.Writer) yql
 	}
 	return printerWriter
 }
+
+func configureEncoder(format yqlib.PrinterOutputFormat) yqlib.Encoder {
+	switch format {
+	case yqlib.JsonOutputFormat:
+		return yqlib.NewJsonEncoder(indent)
+	case yqlib.PropsOutputFormat:
+		return yqlib.NewPropertiesEncoder()
+	case yqlib.CsvOutputFormat:
+		return yqlib.NewCsvEncoder(',')
+	case yqlib.TsvOutputFormat:
+		return yqlib.NewCsvEncoder('\t')
+	case yqlib.YamlOutputFormat:
+		return yqlib.NewYamlEncoder(indent, colorsEnabled, !noDocSeparators, unwrapScalar)
+	case yqlib.XmlOutputFormat:
+		return yqlib.NewXmlEncoder(indent, xmlAttributePrefix, xmlContentName)
+	}
+	panic("invalid encoder")
+}

examples/data1.yaml

@@ -1,15 +1 @@
-strings:
+cat: purrs
-  - a: banana
-  - a: cat
-  - a: apple
-
-numbers:
-  - a: 12
-  - a: 13
-  - a: 120
-
-
-obj:
-  - a: {cat: "adog"}
-  - a: {cat: "doga"}
-  - a: apple

examples/mike.xml

@@ -1,2 +1,14 @@
-<?xml version="1.0" encoding="UTF-8"?>
+<!-- before cat -->
-<cat legs="4">BiBi</cat>
+<cat>
+	<!-- in cat before -->
+	<x>3<!-- multi
+line comment
+for x --></x>
+	<y>
+		<!-- in y before -->
+		<d><!-- in d before -->4<!-- in d after --></d>
+		<!-- in y after -->
+	</y>
+	<!-- in_cat_after -->
+</cat>
+<!-- after cat -->

go.mod

@@ -6,6 +6,7 @@ require (
 	github.com/goccy/go-yaml v1.9.5
 	github.com/jinzhu/copier v0.3.4
 	github.com/magiconair/properties v1.8.5
+	github.com/pkg/diff v0.0.0-20210226163009-20ebb0f2a09e
 	github.com/spf13/cobra v1.3.0
 	github.com/timtadh/lexmachine v0.2.2
 	golang.org/x/net v0.0.0-20210813160813-60bc85c4be6d

go.sum

@@ -292,6 +292,8 @@ github.com/mwitkow/go-conntrack v0.0.0-20161129095857-cc309e4a2223/go.mod h1:qRW
 github.com/pascaldekloe/goe v0.0.0-20180627143212-57f6aae5913c/go.mod h1:lzWF7FIEvWOWxwDKqyGYQf6ZUaNfKdP144TG7ZOy1lc=
 github.com/pascaldekloe/goe v0.1.0/go.mod h1:lzWF7FIEvWOWxwDKqyGYQf6ZUaNfKdP144TG7ZOy1lc=
 github.com/pelletier/go-toml v1.9.4/go.mod h1:u1nR/EPcESfeI/szUZKdtJ0xRNbUoANCkoOuaOx1Y+c=
+github.com/pkg/diff v0.0.0-20210226163009-20ebb0f2a09e h1:aoZm08cpOy4WuID//EZDgcC4zIxODThtZNPirFr42+A=
+github.com/pkg/diff v0.0.0-20210226163009-20ebb0f2a09e/go.mod h1:pJLUxLENpZxwdsKMEsNbx1VGcRFpLqf3715MtcvvzbA=
 github.com/pkg/errors v0.8.0/go.mod h1:bwawxfHBFNV+L2hUp1rHADufV3IMtnDRdf1r5NINEl0=
 github.com/pkg/errors v0.8.1/go.mod h1:bwawxfHBFNV+L2hUp1rHADufV3IMtnDRdf1r5NINEl0=
 github.com/pkg/sftp v1.10.1/go.mod h1:lYOWFsE0bwd1+KfKJaKeuokY15vzFx25BLbzYYoAxZI=

pkg/yqlib/decoder_xml.go

@@ -4,6 +4,7 @@ import (
 	"encoding/xml"
 	"fmt"
 	"io"
+	"strings"
 	"unicode"
 
 	"golang.org/x/net/html/charset"
@@ -31,15 +32,15 @@ func InputFormatFromString(format string) (InputFormat, error) {
 type xmlDecoder struct {
 	reader          io.Reader
 	attributePrefix string
-	contentPrefix   string
+	contentName     string
 	finished        bool
 }
 
-func NewXmlDecoder(attributePrefix string, contentPrefix string) Decoder {
+func NewXmlDecoder(attributePrefix string, contentName string) Decoder {
-	if contentPrefix == "" {
+	if contentName == "" {
-		contentPrefix = "content"
+		contentName = "content"
 	}
-	return &xmlDecoder{attributePrefix: attributePrefix, contentPrefix: contentPrefix, finished: false}
+	return &xmlDecoder{attributePrefix: attributePrefix, contentName: contentName, finished: false}
 }
 
 func (dec *xmlDecoder) Init(reader io.Reader) {
@@ -60,20 +61,41 @@ func (dec *xmlDecoder) createSequence(nodes []*xmlNode) (*yaml.Node, error) {
 	return yamlNode, nil
 }
 
+func (dec *xmlDecoder) processComment(c string) string {
+	if c == "" {
+		return ""
+	}
+	return "#" + strings.TrimRight(c, " ")
+}
+
 func (dec *xmlDecoder) createMap(n *xmlNode) (*yaml.Node, error) {
-	yamlNode := &yaml.Node{Kind: yaml.MappingNode, HeadComment: n.Comment}
+	log.Debug("createMap: headC: %v, footC: %v", n.HeadComment, n.FootComment)
+	yamlNode := &yaml.Node{Kind: yaml.MappingNode}
 
 	if len(n.Data) > 0 {
-		label := dec.contentPrefix
+		label := dec.contentName
-		yamlNode.Content = append(yamlNode.Content, createScalarNode(label, label), createScalarNode(n.Data, n.Data))
+		labelNode := createScalarNode(label, label)
+		labelNode.HeadComment = dec.processComment(n.HeadComment)
+		labelNode.FootComment = dec.processComment(n.FootComment)
+		yamlNode.Content = append(yamlNode.Content, labelNode, createScalarNode(n.Data, n.Data))
 	}
 
-	for _, keyValuePair := range n.Children {
+	for i, keyValuePair := range n.Children {
 		label := keyValuePair.K
 		children := keyValuePair.V
 		labelNode := createScalarNode(label, label)
 		var valueNode *yaml.Node
 		var err error
+
+		if i == 0 {
+			labelNode.HeadComment = dec.processComment(n.HeadComment)
+
+		}
+
+		// if i == len(n.Children)-1 {
+		labelNode.FootComment = dec.processComment(keyValuePair.FootComment)
+		// }
+
 		log.Debug("len of children in %v is %v", label, len(children))
 		if len(children) > 1 {
 			valueNode, err = dec.createSequence(children)
@@ -81,6 +103,13 @@ func (dec *xmlDecoder) createMap(n *xmlNode) (*yaml.Node, error) {
 				return nil, err
 			}
 		} else {
+			// comment hack for maps of scalars
+			// if the value is a scalar, the head comment of the scalar needs to go on the key?
+			// add tests for <z/> as well as multiple <ds> of inputXmlWithComments > yaml
+			if len(children[0].Children) == 0 && children[0].HeadComment != "" {
+				labelNode.HeadComment = labelNode.HeadComment + "\n" + strings.TrimSpace(children[0].HeadComment)
+				children[0].HeadComment = ""
+			}
 			valueNode, err = dec.convertToYamlNode(children[0])
 			if err != nil {
 				return nil, err
@@ -97,7 +126,11 @@ func (dec *xmlDecoder) convertToYamlNode(n *xmlNode) (*yaml.Node, error) {
 		return dec.createMap(n)
 	}
 	scalar := createScalarNode(n.Data, n.Data)
-	scalar.HeadComment = n.Comment
+	log.Debug("scalar headC: %v, footC: %v", n.HeadComment, n.FootComment)
+	scalar.HeadComment = dec.processComment(n.HeadComment)
+	scalar.LineComment = dec.processComment(n.LineComment)
+	scalar.FootComment = dec.processComment(n.FootComment)
+
 	return scalar, nil
 }
 
@@ -124,14 +157,17 @@ func (dec *xmlDecoder) Decode(rootYamlNode *yaml.Node) error {
 }
 
 type xmlNode struct {
-	Children []*xmlChildrenKv
+	Children    []*xmlChildrenKv
-	Comment  string
+	HeadComment string
-	Data     string
+	FootComment string
+	LineComment string
+	Data        string
 }
 
 type xmlChildrenKv struct {
-	K string
+	K           string
-	V []*xmlNode
+	V           []*xmlNode
+	FootComment string
 }
 
 // AddChild appends a node to the list of children
@@ -158,6 +194,7 @@ type element struct {
 	parent *element
 	n      *xmlNode
 	label  string
+	state  string
 }
 
 // this code is heavily based on https://github.com/basgys/goxml2json
@@ -183,6 +220,8 @@ func (dec *xmlDecoder) decodeXml(root *xmlNode) error {
 
 		switch se := t.(type) {
 		case xml.StartElement:
+			log.Debug("start element %v", se.Name.Local)
+			elem.state = "started"
 			// Build new a new current element and link it to its parent
 			elem = &element{
 				parent: elem,
@@ -197,7 +236,13 @@ func (dec *xmlDecoder) decodeXml(root *xmlNode) error {
 		case xml.CharData:
 			// Extract XML data (if any)
 			elem.n.Data = trimNonGraphic(string(se))
+			if elem.n.Data != "" {
+				elem.state = "chardata"
+				log.Debug("chardata [%v] for %v", elem.n.Data, elem.label)
+			}
 		case xml.EndElement:
+			log.Debug("end element %v", elem.label)
+			elem.state = "finished"
 			// And add it to its parent list
 			if elem.parent != nil {
 				elem.parent.n.AddChild(elem.label, elem.n)
@@ -206,13 +251,49 @@ func (dec *xmlDecoder) decodeXml(root *xmlNode) error {
 			// Then change the current element to its parent
 			elem = elem.parent
 		case xml.Comment:
-			elem.n.Comment = trimNonGraphic(string(xml.CharData(se)))
+
+			commentStr := string(xml.CharData(se))
+			if elem.state == "started" {
+				applyFootComment(elem, commentStr)
+
+			} else if elem.state == "chardata" {
+				log.Debug("got a line comment for (%v) %v: [%v]", elem.state, elem.label, commentStr)
+				elem.n.LineComment = joinFilter([]string{elem.n.LineComment, commentStr})
+			} else {
+				log.Debug("got a head comment for (%v) %v: [%v]", elem.state, elem.label, commentStr)
+				elem.n.HeadComment = joinFilter([]string{elem.n.HeadComment, commentStr})
+			}
+
 		}
 	}
 
 	return nil
 }
 
+func applyFootComment(elem *element, commentStr string) {
+
+	// first lets try to put the comment on the last child
+	if len(elem.n.Children) > 0 {
+		lastChildIndex := len(elem.n.Children) - 1
+		childKv := elem.n.Children[lastChildIndex]
+		log.Debug("got a foot comment for %v: [%v]", childKv.K, commentStr)
+		childKv.FootComment = joinFilter([]string{elem.n.FootComment, commentStr})
+	} else {
+		log.Debug("got a foot comment for %v: [%v]", elem.label, commentStr)
+		elem.n.FootComment = joinFilter([]string{elem.n.FootComment, commentStr})
+	}
+}
+
+func joinFilter(rawStrings []string) string {
+	stringsToJoin := make([]string, 0)
+	for _, str := range rawStrings {
+		if str != "" {
+			stringsToJoin = append(stringsToJoin, str)
+		}
+	}
+	return strings.Join(stringsToJoin, " ")
+}
+
 // trimNonGraphic returns a slice of the string s, with all leading and trailing
 // non graphic characters and spaces removed.
 //

pkg/yqlib/doc/usage/headers/xml.md

@@ -2,8 +2,6 @@
 
 Encode and decode to and from XML. Whitespace is not conserved for round trips - but the order of the fields are.
 
-As yaml does not have the concept of attributes, xml attributes are converted to regular fields with a prefix to prevent clobbering. This defaults to "+", use the `--xml-attribute-prefix` to change.
-
 Consecutive xml nodes with the same name are assumed to be arrays.
 
 All values in XML are assumed to be strings - but you can use `from_yaml` to parse them into their correct types:
@@ -14,8 +12,6 @@ yq e -p=xml '.myNumberField |= from_yaml' my.xml
 ```
 
 
-XML nodes that have attributes then plain content, e.g:
-
 ```xml
 <cat name="tiger">meow</cat>
 ```

pkg/yqlib/doc/usage/xml.md

@@ -2,8 +2,6 @@
 
 Encode and decode to and from XML. Whitespace is not conserved for round trips - but the order of the fields are.
 
-As yaml does not have the concept of attributes, xml attributes are converted to regular fields with a prefix to prevent clobbering. This defaults to "+", use the `--xml-attribute-prefix` to change.
-
 Consecutive xml nodes with the same name are assumed to be arrays.
 
 All values in XML are assumed to be strings - but you can use `from_yaml` to parse them into their correct types:
@@ -14,8 +12,6 @@ yq e -p=xml '.myNumberField |= from_yaml' my.xml
 ```
 
 
-XML nodes that have attributes then plain content, e.g:
-
 ```xml
 <cat name="tiger">meow</cat>
 ```
@@ -58,7 +54,7 @@ animal:
 ```
 
 ## Parse xml: attributes
-Attributes are converted to fields, with the attribute prefix.
+Attributes are converted to fields, with the default attribute prefix '+'. Use '--xml-attribute-prefix` to set your own.
 
 Given a sample.xml file of:
 ```xml
@@ -79,7 +75,7 @@ cat:
 ```
 
 ## Parse xml: attributes with content
-Content is added as a field, using the content name
+Content is added as a field, using the default content name of '+content'. Use `--xml-content-name` to set your own.
 
 Given a sample.xml file of:
 ```xml
@@ -97,6 +93,53 @@ cat:
   +legs: "4"
 ```
 
+## Parse xml: with comments
+A best attempt is made to preserve comments.
+
+Given a sample.xml file of:
+```xml
+
+<!-- before cat -->
+<cat>
+	<!-- in cat before -->
+	<x>3<!-- multi
+line comment 
+for x --></x>
+	<!-- before y -->
+	<y>
+		<!-- in y before -->
+		<d><!-- in d before -->z<!-- in d after --></d>
+		
+		<!-- in y after -->
+	</y>
+	<!-- in_cat_after -->
+</cat>
+<!-- after cat -->
+
+```
+then
+```bash
+yq e -p=xml '.' sample.xml
+```
+will output
+```yaml
+# before cat
+cat:
+  # in cat before
+  x: "3" # multi
+  # line comment 
+  # for x
+  # before y
+
+  y:
+    # in y before
+    # in d before
+    d: z # in d after
+    # in y after
+  # in_cat_after
+# after cat
+```
+
 ## Encode xml: simple
 Given a sample.yml file of:
 ```yaml
@@ -108,7 +151,8 @@ yq e -o=xml '.' sample.yml
 ```
 will output
 ```xml
-<cat>purrs</cat>```
+<cat>purrs</cat>
+```
 
 ## Encode xml: array
 Given a sample.yml file of:
@@ -127,7 +171,8 @@ will output
 <pets>
   <cat>purrs</cat>
   <cat>meows</cat>
-</pets>```
+</pets>
+```
 
 ## Encode xml: attributes
 Fields with the matching xml-attribute-prefix are assumed to be attributes.
@@ -147,7 +192,8 @@ will output
 ```xml
 <cat name="tiger">
   <meows>true</meows>
-</cat>```
+</cat>
+```
 
 ## Encode xml: attributes with content
 Fields with the matching xml-content-name is assumed to be content.
@@ -165,5 +211,74 @@ yq e -o=xml '.' sample.yml
 ```
 will output
 ```xml
-<cat name="tiger">cool</cat>```
+<cat name="tiger">cool</cat>
+```
+
+## Encode xml: comments
+A best attempt is made to copy comments to xml.
+
+Given a sample.yml file of:
+```yaml
+# above_cat
+cat: # inline_cat
+  # above_array
+  array: # inline_array
+    - val1 # inline_val1
+    # above_val2
+    - val2 # inline_val2
+# below_cat
+
+```
+then
+```bash
+yq e -o=xml '.' sample.yml
+```
+will output
+```xml
+<!-- above_cat inline_cat --><cat><!-- above_array inline_array -->
+  <array>val1<!-- inline_val1 --></array>
+  <array><!-- above_val2 -->val2<!-- inline_val2 --></array>
+</cat><!-- below_cat -->
+```
+
+## Round trip: with comments
+A best effort is made, but comment positions and white space are not preserved perfectly.
+
+Given a sample.xml file of:
+```xml
+
+<!-- before cat -->
+<cat>
+	<!-- in cat before -->
+	<x>3<!-- multi
+line comment 
+for x --></x>
+	<!-- before y -->
+	<y>
+		<!-- in y before -->
+		<d><!-- in d before -->z<!-- in d after --></d>
+		
+		<!-- in y after -->
+	</y>
+	<!-- in_cat_after -->
+</cat>
+<!-- after cat -->
+
+```
+then
+```bash
+yq e -p=xml -o=xml '.' sample.xml
+```
+will output
+```xml
+<!-- before cat --><cat><!-- in cat before -->
+  <x>3<!-- multi
+line comment 
+for x --></x><!-- before y -->
+  <y><!-- in y before
+in d before -->
+    <d>z<!-- in d after --></d><!-- in y after -->
+  </y><!-- in_cat_after -->
+</cat><!-- after cat -->
+```
 

pkg/yqlib/encoder.go

@@ -11,92 +11,10 @@ import (
 )
 
 type Encoder interface {
-	Encode(node *yaml.Node) error
+	Encode(writer io.Writer, node *yaml.Node) error
-}
+	PrintDocumentSeparator(writer io.Writer) error
-
+	PrintLeadingContent(writer io.Writer, content string) error
-type yamlEncoder struct {
+	CanHandleAliases() bool
-	destination io.Writer
-	indent      int
-	colorise    bool
-	firstDoc    bool
-}
-
-func NewYamlEncoder(destination io.Writer, indent int, colorise bool) Encoder {
-	if indent < 0 {
-		indent = 0
-	}
-	return &yamlEncoder{destination, indent, colorise, true}
-}
-
-func (ye *yamlEncoder) Encode(node *yaml.Node) error {
-
-	destination := ye.destination
-	tempBuffer := bytes.NewBuffer(nil)
-	if ye.colorise {
-		destination = tempBuffer
-	}
-
-	var encoder = yaml.NewEncoder(destination)
-
-	encoder.SetIndent(ye.indent)
-	// TODO: work out if the first doc had a separator or not.
-	if ye.firstDoc {
-		ye.firstDoc = false
-	} else if _, err := destination.Write([]byte("---\n")); err != nil {
-		return err
-	}
-
-	if err := encoder.Encode(node); err != nil {
-		return err
-	}
-
-	if ye.colorise {
-		return colorizeAndPrint(tempBuffer.Bytes(), ye.destination)
-	}
-	return nil
-}
-
-type jsonEncoder struct {
-	encoder *json.Encoder
-}
-
-func mapKeysToStrings(node *yaml.Node) {
-
-	if node.Kind == yaml.MappingNode {
-		for index, child := range node.Content {
-			if index%2 == 0 { // its a map key
-				child.Tag = "!!str"
-			}
-		}
-	}
-
-	for _, child := range node.Content {
-		mapKeysToStrings(child)
-	}
-}
-
-func NewJsonEncoder(destination io.Writer, indent int) Encoder {
-	var encoder = json.NewEncoder(destination)
-	encoder.SetEscapeHTML(false) // do not escape html chars e.g. &, <, >
-
-	var indentString = ""
-
-	for index := 0; index < indent; index++ {
-		indentString = indentString + " "
-	}
-	encoder.SetIndent("", indentString)
-	return &jsonEncoder{encoder}
-}
-
-func (je *jsonEncoder) Encode(node *yaml.Node) error {
-	var dataBucket orderedMap
-	// firstly, convert all map keys to strings
-	mapKeysToStrings(node)
-	errorDecoding := node.Decode(&dataBucket)
-	if errorDecoding != nil {
-		return errorDecoding
-	}
-	return je.encoder.Encode(dataBucket)
 }
 
 // orderedMap allows to marshal and unmarshal JSON and YAML values keeping the

pkg/yqlib/encoder_csv.go

@@ -9,16 +9,26 @@ import (
 )
 
 type csvEncoder struct {
-	destination csv.Writer
+	separator rune
 }
 
-func NewCsvEncoder(destination io.Writer, separator rune) Encoder {
+func NewCsvEncoder(separator rune) Encoder {
-	csvWriter := *csv.NewWriter(destination)
+	return &csvEncoder{separator}
-	csvWriter.Comma = separator
-	return &csvEncoder{csvWriter}
 }
 
-func (e *csvEncoder) encodeRow(contents []*yaml.Node) error {
+func (e *csvEncoder) CanHandleAliases() bool {
+	return false
+}
+
+func (e *csvEncoder) PrintDocumentSeparator(writer io.Writer) error {
+	return nil
+}
+
+func (e *csvEncoder) PrintLeadingContent(writer io.Writer, content string) error {
+	return nil
+}
+
+func (e *csvEncoder) encodeRow(csvWriter *csv.Writer, contents []*yaml.Node) error {
 	stringValues := make([]string, len(contents))
 
 	for i, child := range contents {
@@ -28,10 +38,13 @@ func (e *csvEncoder) encodeRow(contents []*yaml.Node) error {
 		}
 		stringValues[i] = child.Value
 	}
-	return e.destination.Write(stringValues)
+	return csvWriter.Write(stringValues)
 }
 
-func (e *csvEncoder) Encode(originalNode *yaml.Node) error {
+func (e *csvEncoder) Encode(writer io.Writer, originalNode *yaml.Node) error {
+	csvWriter := csv.NewWriter(writer)
+	csvWriter.Comma = e.separator
+
 	// node must be a sequence
 	node := unwrapDoc(originalNode)
 	if node.Kind != yaml.SequenceNode {
@@ -40,7 +53,7 @@ func (e *csvEncoder) Encode(originalNode *yaml.Node) error {
 		return nil
 	}
 	if node.Content[0].Kind == yaml.ScalarNode {
-		return e.encodeRow(node.Content)
+		return e.encodeRow(csvWriter, node.Content)
 	}
 
 	for i, child := range node.Content {
@@ -48,7 +61,7 @@ func (e *csvEncoder) Encode(originalNode *yaml.Node) error {
 		if child.Kind != yaml.SequenceNode {
 			return fmt.Errorf("csv encoding only works for arrays of scalars (string/numbers/booleans), child[%v] is a %v", i, child.Tag)
 		}
-		err := e.encodeRow(child.Content)
+		err := e.encodeRow(csvWriter, child.Content)
 		if err != nil {
 			return err
 		}

pkg/yqlib/encoder_csv_test.go

@@ -13,13 +13,13 @@ func yamlToCsv(sampleYaml string, separator rune) string {
 	var output bytes.Buffer
 	writer := bufio.NewWriter(&output)
 
-	var jsonEncoder = NewCsvEncoder(writer, separator)
+	var jsonEncoder = NewCsvEncoder(separator)
 	inputs, err := readDocuments(strings.NewReader(sampleYaml), "sample.yml", 0, NewYamlDecoder())
 	if err != nil {
 		panic(err)
 	}
 	node := inputs.Front().Value.(*CandidateNode).Node
-	err = jsonEncoder.Encode(node)
+	err = jsonEncoder.Encode(writer, node)
 	if err != nil {
 		panic(err)
 	}

pkg/yqlib/encoder_json.go

@@ -0,0 +1,64 @@
+package yqlib
+
+import (
+	"encoding/json"
+	"io"
+
+	yaml "gopkg.in/yaml.v3"
+)
+
+type jsonEncoder struct {
+	indentString string
+}
+
+func mapKeysToStrings(node *yaml.Node) {
+
+	if node.Kind == yaml.MappingNode {
+		for index, child := range node.Content {
+			if index%2 == 0 { // its a map key
+				child.Tag = "!!str"
+			}
+		}
+	}
+
+	for _, child := range node.Content {
+		mapKeysToStrings(child)
+	}
+}
+
+func NewJsonEncoder(indent int) Encoder {
+	var indentString = ""
+
+	for index := 0; index < indent; index++ {
+		indentString = indentString + " "
+	}
+
+	return &jsonEncoder{indentString}
+}
+
+func (je *jsonEncoder) CanHandleAliases() bool {
+	return false
+}
+
+func (je *jsonEncoder) PrintDocumentSeparator(writer io.Writer) error {
+	return nil
+}
+
+func (je *jsonEncoder) PrintLeadingContent(writer io.Writer, content string) error {
+	return nil
+}
+
+func (je *jsonEncoder) Encode(writer io.Writer, node *yaml.Node) error {
+	var encoder = json.NewEncoder(writer)
+	encoder.SetEscapeHTML(false) // do not escape html chars e.g. &, <, >
+	encoder.SetIndent("", je.indentString)
+
+	var dataBucket orderedMap
+	// firstly, convert all map keys to strings
+	mapKeysToStrings(node)
+	errorDecoding := node.Decode(&dataBucket)
+	if errorDecoding != nil {
+		return errorDecoding
+	}
+	return encoder.Encode(dataBucket)
+}

pkg/yqlib/encoder_properties.go

@@ -1,6 +1,8 @@
 package yqlib
 
 import (
+	"bufio"
+	"errors"
 	"fmt"
 	"io"
 	"strings"
@@ -10,14 +12,54 @@ import (
 )
 
 type propertiesEncoder struct {
-	destination io.Writer
 }
 
-func NewPropertiesEncoder(destination io.Writer) Encoder {
+func NewPropertiesEncoder() Encoder {
-	return &propertiesEncoder{destination}
+	return &propertiesEncoder{}
 }
 
-func (pe *propertiesEncoder) Encode(node *yaml.Node) error {
+func (pe *propertiesEncoder) CanHandleAliases() bool {
+	return false
+}
+
+func (pe *propertiesEncoder) PrintDocumentSeparator(writer io.Writer) error {
+	return nil
+}
+
+func (pe *propertiesEncoder) PrintLeadingContent(writer io.Writer, content string) error {
+	reader := bufio.NewReader(strings.NewReader(content))
+	for {
+
+		readline, errReading := reader.ReadString('\n')
+		if errReading != nil && !errors.Is(errReading, io.EOF) {
+			return errReading
+		}
+		if strings.Contains(readline, "$yqDocSeperator$") {
+
+			if err := pe.PrintDocumentSeparator(writer); err != nil {
+				return err
+			}
+
+		} else {
+			if err := writeString(writer, readline); err != nil {
+				return err
+			}
+		}
+
+		if errors.Is(errReading, io.EOF) {
+			if readline != "" {
+				// the last comment we read didn't have a new line, put one in
+				if err := writeString(writer, "\n"); err != nil {
+					return err
+				}
+			}
+			break
+		}
+	}
+	return nil
+}
+
+func (pe *propertiesEncoder) Encode(writer io.Writer, node *yaml.Node) error {
 	mapKeysToStrings(node)
 	p := properties.NewProperties()
 	err := pe.doEncode(p, node, "")
@@ -25,14 +67,12 @@ func (pe *propertiesEncoder) Encode(node *yaml.Node) error {
 		return err
 	}
 
-	_, err = p.WriteComment(pe.destination, "#", properties.UTF8)
+	_, err = p.WriteComment(writer, "#", properties.UTF8)
 	return err
 }
 
 func (pe *propertiesEncoder) doEncode(p *properties.Properties, node *yaml.Node, path string) error {
-	p.SetComment(path,
+	p.SetComment(path, headAndLineComment(node))
-		strings.Replace(node.HeadComment, "#", "", 1)+
-			strings.Replace(node.LineComment, "#", "", 1))
 	switch node.Kind {
 	case yaml.ScalarNode:
 		_, _, err := p.Set(path, node.Value)

pkg/yqlib/encoder_properties_test.go

@@ -13,13 +13,13 @@ func yamlToProps(sampleYaml string) string {
 	var output bytes.Buffer
 	writer := bufio.NewWriter(&output)
 
-	var propsEncoder = NewPropertiesEncoder(writer)
+	var propsEncoder = NewPropertiesEncoder()
 	inputs, err := readDocuments(strings.NewReader(sampleYaml), "sample.yml", 0, NewYamlDecoder())
 	if err != nil {
 		panic(err)
 	}
 	node := inputs.Front().Value.(*CandidateNode).Node
-	err = propsEncoder.Encode(node)
+	err = propsEncoder.Encode(writer, node)
 	if err != nil {
 		panic(err)
 	}

pkg/yqlib/encoder_test.go

@@ -13,13 +13,13 @@ func yamlToJson(sampleYaml string, indent int) string {
 	var output bytes.Buffer
 	writer := bufio.NewWriter(&output)
 
-	var jsonEncoder = NewJsonEncoder(writer, indent)
+	var jsonEncoder = NewJsonEncoder(indent)
 	inputs, err := readDocuments(strings.NewReader(sampleYaml), "sample.yml", 0, NewYamlDecoder())
 	if err != nil {
 		panic(err)
 	}
 	node := inputs.Front().Value.(*CandidateNode).Node
-	err = jsonEncoder.Encode(node)
+	err = jsonEncoder.Encode(writer, node)
 	if err != nil {
 		panic(err)
 	}

pkg/yqlib/encoder_xml.go

@@ -9,89 +9,155 @@ import (
 	yaml "gopkg.in/yaml.v3"
 )
 
+var XmlPreferences = xmlPreferences{AttributePrefix: "+", ContentName: "+content"}
+
 type xmlEncoder struct {
-	xmlEncoder      *xml.Encoder
 	attributePrefix string
 	contentName     string
+	indentString    string
 }
 
-func NewXmlEncoder(writer io.Writer, indent int, attributePrefix string, contentName string) Encoder {
+func NewXmlEncoder(indent int, attributePrefix string, contentName string) Encoder {
-	encoder := xml.NewEncoder(writer)
 	var indentString = ""
 
 	for index := 0; index < indent; index++ {
 		indentString = indentString + " "
 	}
-	encoder.Indent("", indentString)
+	return &xmlEncoder{attributePrefix, contentName, indentString}
-	return &xmlEncoder{encoder, attributePrefix, contentName}
-}
-func (e *xmlEncoder) Encode(node *yaml.Node) error {
-	switch node.Kind {
-	case yaml.MappingNode:
-		err := e.encodeTopLevelMap(node)
-		if err != nil {
-			return err
-		}
-		var charData xml.CharData = []byte("\n")
-		err = e.xmlEncoder.EncodeToken(charData)
-		if err != nil {
-			return err
-		}
-		return e.xmlEncoder.Flush()
-	case yaml.DocumentNode:
-		return e.Encode(unwrapDoc(node))
-	case yaml.ScalarNode:
-		var charData xml.CharData = []byte(node.Value)
-		err := e.xmlEncoder.EncodeToken(charData)
-		if err != nil {
-			return err
-		}
-		return e.xmlEncoder.Flush()
-	}
-	return fmt.Errorf("unsupported type %v", node.Tag)
 }
 
-func (e *xmlEncoder) encodeTopLevelMap(node *yaml.Node) error {
+func (e *xmlEncoder) CanHandleAliases() bool {
+	return false
+}
+
+func (e *xmlEncoder) PrintDocumentSeparator(writer io.Writer) error {
+	return nil
+}
+
+func (e *xmlEncoder) PrintLeadingContent(writer io.Writer, content string) error {
+	return nil
+}
+
+func (e *xmlEncoder) Encode(writer io.Writer, node *yaml.Node) error {
+	encoder := xml.NewEncoder(writer)
+	encoder.Indent("", e.indentString)
+
+	switch node.Kind {
+	case yaml.MappingNode:
+		err := e.encodeTopLevelMap(encoder, node)
+		if err != nil {
+			return err
+		}
+	case yaml.DocumentNode:
+		err := e.encodeComment(encoder, headAndLineComment(node))
+		if err != nil {
+			return err
+		}
+		err = e.encodeTopLevelMap(encoder, unwrapDoc(node))
+		if err != nil {
+			return err
+		}
+		err = e.encodeComment(encoder, footComment(node))
+		if err != nil {
+			return err
+		}
+	case yaml.ScalarNode:
+		var charData xml.CharData = []byte(node.Value)
+		err := encoder.EncodeToken(charData)
+		if err != nil {
+			return err
+		}
+		return encoder.Flush()
+	default:
+		return fmt.Errorf("unsupported type %v", node.Tag)
+	}
+	var charData xml.CharData = []byte("\n")
+	return encoder.EncodeToken(charData)
+
+}
+
+func (e *xmlEncoder) encodeTopLevelMap(encoder *xml.Encoder, node *yaml.Node) error {
+	err := e.encodeComment(encoder, headAndLineComment(node))
+	if err != nil {
+		return err
+	}
 	for i := 0; i < len(node.Content); i += 2 {
 		key := node.Content[i]
 		value := node.Content[i+1]
 
 		start := xml.StartElement{Name: xml.Name{Local: key.Value}}
-		err := e.doEncode(value, start)
+		log.Debugf("comments of key %v", key.Value)
+		err := e.encodeComment(encoder, headAndLineComment(key))
+		if err != nil {
+			return err
+		}
+
+		log.Debugf("recursing")
+
+		err = e.doEncode(encoder, value, start)
+		if err != nil {
+			return err
+		}
+		err = e.encodeComment(encoder, footComment(key))
 		if err != nil {
 			return err
 		}
 	}
-	return nil
+	return e.encodeComment(encoder, footComment(node))
 }
 
-func (e *xmlEncoder) doEncode(node *yaml.Node, start xml.StartElement) error {
+func (e *xmlEncoder) encodeStart(encoder *xml.Encoder, node *yaml.Node, start xml.StartElement) error {
+	err := encoder.EncodeToken(start)
+	if err != nil {
+		return err
+	}
+	return e.encodeComment(encoder, headComment(node))
+}
+
+func (e *xmlEncoder) encodeEnd(encoder *xml.Encoder, node *yaml.Node, start xml.StartElement) error {
+	err := encoder.EncodeToken(start.End())
+	if err != nil {
+		return err
+	}
+	return e.encodeComment(encoder, footComment(node))
+}
+
+func (e *xmlEncoder) doEncode(encoder *xml.Encoder, node *yaml.Node, start xml.StartElement) error {
 	switch node.Kind {
 	case yaml.MappingNode:
-		return e.encodeMap(node, start)
+		return e.encodeMap(encoder, node, start)
 	case yaml.SequenceNode:
-		return e.encodeArray(node, start)
+		return e.encodeArray(encoder, node, start)
 	case yaml.ScalarNode:
-		err := e.xmlEncoder.EncodeToken(start)
+		err := e.encodeStart(encoder, node, start)
 		if err != nil {
 			return err
 		}
 
 		var charData xml.CharData = []byte(node.Value)
-		err = e.xmlEncoder.EncodeToken(charData)
+		err = encoder.EncodeToken(charData)
-
 		if err != nil {
 			return err
 		}
-		return e.xmlEncoder.EncodeToken(start.End())
+
+		if err = e.encodeComment(encoder, lineComment(node)); err != nil {
+			return err
+		}
+
+		return e.encodeEnd(encoder, node, start)
 	}
 	return fmt.Errorf("unsupported type %v", node.Tag)
 }
 
-func (e *xmlEncoder) encodeArray(node *yaml.Node, start xml.StartElement) error {
+func (e *xmlEncoder) encodeComment(encoder *xml.Encoder, commentStr string) error {
-	for i := 0; i < len(node.Content); i++ {
+	if commentStr != "" {
-		value := node.Content[i]
+		log.Debugf("encoding comment %v", commentStr)
-		err := e.doEncode(value, start.Copy())
+		if !strings.HasSuffix(commentStr, " ") {
+			commentStr = commentStr + " "
+		}
+
+		var comment xml.Comment = []byte(commentStr)
+		err := encoder.EncodeToken(comment)
 		if err != nil {
 			return err
 		}
@@ -99,7 +165,23 @@ func (e *xmlEncoder) encodeArray(node *yaml.Node, start xml.StartElement) error
 	return nil
 }
 
-func (e *xmlEncoder) encodeMap(node *yaml.Node, start xml.StartElement) error {
+func (e *xmlEncoder) encodeArray(encoder *xml.Encoder, node *yaml.Node, start xml.StartElement) error {
+
+	if err := e.encodeComment(encoder, headAndLineComment(node)); err != nil {
+		return err
+	}
+
+	for i := 0; i < len(node.Content); i++ {
+		value := node.Content[i]
+		if err := e.doEncode(encoder, value, start.Copy()); err != nil {
+			return err
+		}
+	}
+	return e.encodeComment(encoder, footComment(node))
+}
+
+func (e *xmlEncoder) encodeMap(encoder *xml.Encoder, node *yaml.Node, start xml.StartElement) error {
+	log.Debug("its a map")
 
 	//first find all the attributes and put them on the start token
 	for i := 0; i < len(node.Content); i += 2 {
@@ -116,7 +198,7 @@ func (e *xmlEncoder) encodeMap(node *yaml.Node, start xml.StartElement) error {
 		}
 	}
 
-	err := e.xmlEncoder.EncodeToken(start)
+	err := e.encodeStart(encoder, node, start)
 	if err != nil {
 		return err
 	}
@@ -126,21 +208,38 @@ func (e *xmlEncoder) encodeMap(node *yaml.Node, start xml.StartElement) error {
 		key := node.Content[i]
 		value := node.Content[i+1]
 
+		err := e.encodeComment(encoder, headAndLineComment(key))
+		if err != nil {
+			return err
+		}
+
 		if !strings.HasPrefix(key.Value, e.attributePrefix) && key.Value != e.contentName {
 			start := xml.StartElement{Name: xml.Name{Local: key.Value}}
-			err := e.doEncode(value, start)
+			err := e.doEncode(encoder, value, start)
 			if err != nil {
 				return err
 			}
 		} else if key.Value == e.contentName {
 			// directly encode the contents
+			err = e.encodeComment(encoder, headAndLineComment(value))
+			if err != nil {
+				return err
+			}
 			var charData xml.CharData = []byte(value.Value)
-			err = e.xmlEncoder.EncodeToken(charData)
+			err = encoder.EncodeToken(charData)
+			if err != nil {
+				return err
+			}
+			err = e.encodeComment(encoder, footComment(value))
 			if err != nil {
 				return err
 			}
 		}
+		err = e.encodeComment(encoder, footComment(key))
+		if err != nil {
+			return err
+		}
 	}
 
-	return e.xmlEncoder.EncodeToken(start.End())
+	return e.encodeEnd(encoder, node, start)
 }

pkg/yqlib/encoder_yaml.go

@@ -0,0 +1,101 @@
+package yqlib
+
+import (
+	"bufio"
+	"bytes"
+	"errors"
+	"io"
+	"strings"
+
+	yaml "gopkg.in/yaml.v3"
+)
+
+type yamlEncoder struct {
+	indent             int
+	colorise           bool
+	printDocSeparators bool
+	unwrapScalar       bool
+}
+
+func NewYamlEncoder(indent int, colorise bool, printDocSeparators bool, unwrapScalar bool) Encoder {
+	if indent < 0 {
+		indent = 0
+	}
+	return &yamlEncoder{indent, colorise, printDocSeparators, unwrapScalar}
+}
+
+func (ye *yamlEncoder) CanHandleAliases() bool {
+	return true
+}
+
+func (ye *yamlEncoder) PrintDocumentSeparator(writer io.Writer) error {
+	if ye.printDocSeparators {
+		log.Debug("-- writing doc sep")
+		if err := writeString(writer, "---\n"); err != nil {
+			return err
+		}
+	}
+	return nil
+}
+
+func (ye *yamlEncoder) PrintLeadingContent(writer io.Writer, content string) error {
+	// log.Debug("headcommentwas [%v]", content)
+	reader := bufio.NewReader(strings.NewReader(content))
+
+	for {
+
+		readline, errReading := reader.ReadString('\n')
+		if errReading != nil && !errors.Is(errReading, io.EOF) {
+			return errReading
+		}
+		if strings.Contains(readline, "$yqDocSeperator$") {
+
+			if err := ye.PrintDocumentSeparator(writer); err != nil {
+				return err
+			}
+
+		} else {
+			if err := writeString(writer, readline); err != nil {
+				return err
+			}
+		}
+
+		if errors.Is(errReading, io.EOF) {
+			if readline != "" {
+				// the last comment we read didn't have a new line, put one in
+				if err := writeString(writer, "\n"); err != nil {
+					return err
+				}
+			}
+			break
+		}
+	}
+
+	return nil
+}
+
+func (ye *yamlEncoder) Encode(writer io.Writer, node *yaml.Node) error {
+
+	if node.Kind == yaml.ScalarNode && ye.unwrapScalar {
+		return writeString(writer, node.Value+"\n")
+	}
+
+	destination := writer
+	tempBuffer := bytes.NewBuffer(nil)
+	if ye.colorise {
+		destination = tempBuffer
+	}
+
+	var encoder = yaml.NewEncoder(destination)
+
+	encoder.SetIndent(ye.indent)
+
+	if err := encoder.Encode(node); err != nil {
+		return err
+	}
+
+	if ye.colorise {
+		return colorizeAndPrint(tempBuffer.Bytes(), writer)
+	}
+	return nil
+}

pkg/yqlib/lib.go

@@ -18,8 +18,6 @@ type xmlPreferences struct {
 	ContentName     string
 }
 
-var XmlPreferences = xmlPreferences{AttributePrefix: "+", ContentName: "+content"}
-
 var log = logging.MustGetLogger("yq-lib")
 
 // GetLogger returns the yq logger instance.
@@ -236,6 +234,22 @@ func createScalarNode(value interface{}, stringValue string) *yaml.Node {
 	return node
 }
 
+func headAndLineComment(node *yaml.Node) string {
+	return headComment(node) + lineComment(node)
+}
+
+func headComment(node *yaml.Node) string {
+	return strings.Replace(node.HeadComment, "#", "", 1)
+}
+
+func lineComment(node *yaml.Node) string {
+	return strings.Replace(node.LineComment, "#", "", 1)
+}
+
+func footComment(node *yaml.Node) string {
+	return strings.Replace(node.FootComment, "#", "", 1)
+}
+
 func createValueOperation(value interface{}, stringValue string) *Operation {
 	var node *yaml.Node = createScalarNode(value, stringValue)
 

pkg/yqlib/operator_comments.go

@@ -86,7 +86,8 @@ func getCommentsOperator(d *dataTreeNavigator, context Context, expressionNode *
 			var chompRegexp = regexp.MustCompile(`\n$`)
 			var output bytes.Buffer
 			var writer = bufio.NewWriter(&output)
-			if err := processLeadingContent(candidate, writer, false, YamlOutputFormat); err != nil {
+			var encoder = NewYamlEncoder(2, false, false, false)
+			if err := encoder.PrintLeadingContent(writer, candidate.LeadingContent); err != nil {
 				return Context{}, err
 			}
 			if err := writer.Flush(); err != nil {

pkg/yqlib/operator_encoder_decoder.go

@@ -10,11 +10,31 @@ import (
 	"gopkg.in/yaml.v3"
 )
 
-func yamlToString(candidate *CandidateNode, prefs encoderPreferences) (string, error) {
+func configureEncoder(format PrinterOutputFormat, indent int) Encoder {
+	switch format {
+	case JsonOutputFormat:
+		return NewJsonEncoder(indent)
+	case PropsOutputFormat:
+		return NewPropertiesEncoder()
+	case CsvOutputFormat:
+		return NewCsvEncoder(',')
+	case TsvOutputFormat:
+		return NewCsvEncoder('\t')
+	case YamlOutputFormat:
+		return NewYamlEncoder(indent, false, true, true)
+	case XmlOutputFormat:
+		return NewXmlEncoder(indent, XmlPreferences.AttributePrefix, XmlPreferences.ContentName)
+	}
+	panic("invalid encoder")
+}
+
+func encodeToString(candidate *CandidateNode, prefs encoderPreferences) (string, error) {
 	var output bytes.Buffer
 	log.Debug("printing with indent: %v", prefs.indent)
 
-	printer := NewPrinterWithSingleWriter(bufio.NewWriter(&output), prefs.format, true, false, prefs.indent, true)
+	encoder := configureEncoder(prefs.format, prefs.indent)
+
+	printer := NewPrinter(encoder, NewSinglePrinterWriter(bufio.NewWriter(&output)))
 	err := printer.PrintResults(candidate.AsList())
 	return output.String(), err
 }
@@ -36,7 +56,7 @@ func encodeOperator(d *dataTreeNavigator, context Context, expressionNode *Expre
 
 	for el := context.MatchingNodes.Front(); el != nil; el = el.Next() {
 		candidate := el.Value.(*CandidateNode)
-		stringValue, err := yamlToString(candidate, preferences)
+		stringValue, err := encodeToString(candidate, preferences)
 
 		if err != nil {
 			return Context{}, err

pkg/yqlib/operators_test.go

@@ -26,6 +26,10 @@ type expressionScenario struct {
 	dontFormatInputForDoc bool // dont format input doc for documentation generation
 }
 
+func NewSimpleYamlPrinter(writer io.Writer, outputFormat PrinterOutputFormat, unwrapScalar bool, colorsEnabled bool, indent int, printDocSeparators bool) Printer {
+	return NewPrinter(NewYamlEncoder(indent, colorsEnabled, printDocSeparators, unwrapScalar), NewSinglePrinterWriter(writer))
+}
+
 func readDocumentWithLeadingContent(content string, fakefilename string, fakeFileIndex int) (*list.List, error) {
 	reader, firstFileLeadingContent, err := processReadStream(bufio.NewReader(strings.NewReader(content)))
 	if err != nil {
@@ -92,7 +96,7 @@ func testScenario(t *testing.T, s *expressionScenario) {
 
 func resultToString(t *testing.T, n *CandidateNode) string {
 	var valueBuffer bytes.Buffer
-	printer := NewPrinterWithSingleWriter(bufio.NewWriter(&valueBuffer), YamlOutputFormat, true, false, 4, true)
+	printer := NewSimpleYamlPrinter(bufio.NewWriter(&valueBuffer), YamlOutputFormat, true, false, 4, true)
 
 	err := printer.PrintResults(n.AsList())
 	if err != nil {
@@ -145,7 +149,7 @@ func copyFromHeader(folder string, title string, out *os.File) error {
 
 func formatYaml(yaml string, filename string) string {
 	var output bytes.Buffer
-	printer := NewPrinterWithSingleWriter(bufio.NewWriter(&output), YamlOutputFormat, true, false, 2, true)
+	printer := NewSimpleYamlPrinter(bufio.NewWriter(&output), YamlOutputFormat, true, false, 2, true)
 
 	node, err := NewExpressionParser().ParseExpression(".. style= \"\"")
 	if err != nil {
@@ -268,7 +272,7 @@ func documentInput(w *bufio.Writer, s expressionScenario) (string, string) {
 func documentOutput(t *testing.T, w *bufio.Writer, s expressionScenario, formattedDoc string, formattedDoc2 string) {
 	var output bytes.Buffer
 	var err error
-	printer := NewPrinterWithSingleWriter(bufio.NewWriter(&output), YamlOutputFormat, true, false, 2, true)
+	printer := NewSimpleYamlPrinter(bufio.NewWriter(&output), YamlOutputFormat, true, false, 2, true)
 
 	node, err := NewExpressionParser().ParseExpression(s.expression)
 	if err != nil {

pkg/yqlib/printer.go

@@ -48,34 +48,22 @@ func OutputFormatFromString(format string) (PrinterOutputFormat, error) {
 }
 
 type resultsPrinter struct {
-	outputFormat       PrinterOutputFormat
+	encoder           Encoder
-	unwrapScalar       bool
+	printerWriter     PrinterWriter
-	colorsEnabled      bool
+	firstTimePrinting bool
-	indent             int
+	previousDocIndex  uint
-	printDocSeparators bool
+	previousFileIndex int
-	printerWriter      PrinterWriter
+	printedMatches    bool
-	firstTimePrinting  bool
+	treeNavigator     DataTreeNavigator
-	previousDocIndex   uint
+	appendixReader    io.Reader
-	previousFileIndex  int
-	printedMatches     bool
-	treeNavigator      DataTreeNavigator
-	appendixReader     io.Reader
 }
 
-func NewPrinterWithSingleWriter(writer io.Writer, outputFormat PrinterOutputFormat, unwrapScalar bool, colorsEnabled bool, indent int, printDocSeparators bool) Printer {
+func NewPrinter(encoder Encoder, printerWriter PrinterWriter) Printer {
-	return NewPrinter(NewSinglePrinterWriter(writer), outputFormat, unwrapScalar, colorsEnabled, indent, printDocSeparators)
-}
-
-func NewPrinter(printerWriter PrinterWriter, outputFormat PrinterOutputFormat, unwrapScalar bool, colorsEnabled bool, indent int, printDocSeparators bool) Printer {
 	return &resultsPrinter{
-		printerWriter:      printerWriter,
+		encoder:           encoder,
-		outputFormat:       outputFormat,
+		printerWriter:     printerWriter,
-		unwrapScalar:       unwrapScalar,
+		firstTimePrinting: true,
-		colorsEnabled:      colorsEnabled,
+		treeNavigator:     NewDataTreeNavigator(),
-		indent:             indent,
-		printDocSeparators: outputFormat == YamlOutputFormat && printDocSeparators,
-		firstTimePrinting:  true,
-		treeNavigator:      NewDataTreeNavigator(),
 	}
 }
 
@@ -90,28 +78,7 @@ func (p *resultsPrinter) PrintedAnything() bool {
 func (p *resultsPrinter) printNode(node *yaml.Node, writer io.Writer) error {
 	p.printedMatches = p.printedMatches || (node.Tag != "!!null" &&
 		(node.Tag != "!!bool" || node.Value != "false"))
-
+	return p.encoder.Encode(writer, node)
-	var encoder Encoder
-	if node.Kind == yaml.ScalarNode && p.unwrapScalar && p.outputFormat == YamlOutputFormat {
-		return writeString(writer, node.Value+"\n")
-	}
-
-	switch p.outputFormat {
-	case JsonOutputFormat:
-		encoder = NewJsonEncoder(writer, p.indent)
-	case PropsOutputFormat:
-		encoder = NewPropertiesEncoder(writer)
-	case CsvOutputFormat:
-		encoder = NewCsvEncoder(writer, ',')
-	case TsvOutputFormat:
-		encoder = NewCsvEncoder(writer, '\t')
-	case YamlOutputFormat:
-		encoder = NewYamlEncoder(writer, p.indent, p.colorsEnabled)
-	case XmlOutputFormat:
-		encoder = NewXmlEncoder(writer, p.indent, XmlPreferences.AttributePrefix, XmlPreferences.ContentName)
-	}
-
-	return encoder.Encode(node)
 }
 
 func (p *resultsPrinter) PrintResults(matchingNodes *list.List) error {
@@ -122,7 +89,7 @@ func (p *resultsPrinter) PrintResults(matchingNodes *list.List) error {
 		return nil
 	}
 
-	if p.outputFormat != YamlOutputFormat {
+	if !p.encoder.CanHandleAliases() {
 		explodeOp := Operation{OperationType: explodeOpType}
 		explodeNode := ExpressionNode{Operation: &explodeOp}
 		context, err := p.treeNavigator.GetMatchingNodes(Context{MatchingNodes: matchingNodes}, &explodeNode)
@@ -142,7 +109,7 @@ func (p *resultsPrinter) PrintResults(matchingNodes *list.List) error {
 	for el := matchingNodes.Front(); el != nil; el = el.Next() {
 
 		mappedDoc := el.Value.(*CandidateNode)
-		log.Debug("-- print sep logic: p.firstTimePrinting: %v, previousDocIndex: %v, mappedDoc.Document: %v, printDocSeparators: %v", p.firstTimePrinting, p.previousDocIndex, mappedDoc.Document, p.printDocSeparators)
+		log.Debug("-- print sep logic: p.firstTimePrinting: %v, previousDocIndex: %v, mappedDoc.Document: %v", p.firstTimePrinting, p.previousDocIndex, mappedDoc.Document)
 
 		writer, errorWriting := p.printerWriter.GetWriter(mappedDoc)
 		if errorWriting != nil {
@@ -152,14 +119,13 @@ func (p *resultsPrinter) PrintResults(matchingNodes *list.List) error {
 		commentsStartWithSepExp := regexp.MustCompile(`^\$yqDocSeperator\$`)
 		commentStartsWithSeparator := commentsStartWithSepExp.MatchString(mappedDoc.LeadingContent)
 
-		if (p.previousDocIndex != mappedDoc.Document || p.previousFileIndex != mappedDoc.FileIndex) && p.printDocSeparators && !commentStartsWithSeparator {
+		if (p.previousDocIndex != mappedDoc.Document || p.previousFileIndex != mappedDoc.FileIndex) && !commentStartsWithSeparator {
-			log.Debug("-- writing doc sep")
+			if err := p.encoder.PrintDocumentSeparator(writer); err != nil {
-			if err := writeString(writer, "---\n"); err != nil {
 				return err
 			}
 		}
 
-		if err := processLeadingContent(mappedDoc, writer, p.printDocSeparators, p.outputFormat); err != nil {
+		if err := p.encoder.PrintLeadingContent(writer, mappedDoc.LeadingContent); err != nil {
 			return err
 		}
 
@@ -171,9 +137,11 @@ func (p *resultsPrinter) PrintResults(matchingNodes *list.List) error {
 		if err := writer.Flush(); err != nil {
 			return err
 		}
+		log.Debugf("done printing results")
 	}
 
-	if p.appendixReader != nil && p.outputFormat == YamlOutputFormat {
+	// what happens if I remove output format check?
+	if p.appendixReader != nil {
 		writer, err := p.printerWriter.GetWriter(nil)
 		if err != nil {
 			return err

pkg/yqlib/printer_test.go

@@ -33,10 +33,10 @@ func nodeToList(candidate *CandidateNode) *list.List {
 	return elMap
 }
 
-func TestPrinterMultipleDocsInSequence(t *testing.T) {
+func TestPrinterMultipleDocsInSequenceOnly(t *testing.T) {
 	var output bytes.Buffer
 	var writer = bufio.NewWriter(&output)
-	printer := NewPrinterWithSingleWriter(writer, YamlOutputFormat, true, false, 2, true)
+	printer := NewSimpleYamlPrinter(writer, YamlOutputFormat, true, false, 2, true)
 
 	inputs, err := readDocuments(strings.NewReader(multiDocSample), "sample.yml", 0, NewYamlDecoder())
 	if err != nil {
@@ -74,7 +74,7 @@ func TestPrinterMultipleDocsInSequence(t *testing.T) {
 func TestPrinterMultipleDocsInSequenceWithLeadingContent(t *testing.T) {
 	var output bytes.Buffer
 	var writer = bufio.NewWriter(&output)
-	printer := NewPrinterWithSingleWriter(writer, YamlOutputFormat, true, false, 2, true)
+	printer := NewSimpleYamlPrinter(writer, YamlOutputFormat, true, false, 2, true)
 
 	inputs, err := readDocuments(strings.NewReader(multiDocSample), "sample.yml", 0, NewYamlDecoder())
 	if err != nil {
@@ -116,7 +116,7 @@ func TestPrinterMultipleDocsInSequenceWithLeadingContent(t *testing.T) {
 func TestPrinterMultipleFilesInSequence(t *testing.T) {
 	var output bytes.Buffer
 	var writer = bufio.NewWriter(&output)
-	printer := NewPrinterWithSingleWriter(writer, YamlOutputFormat, true, false, 2, true)
+	printer := NewSimpleYamlPrinter(writer, YamlOutputFormat, true, false, 2, true)
 
 	inputs, err := readDocuments(strings.NewReader(multiDocSample), "sample.yml", 0, NewYamlDecoder())
 	if err != nil {
@@ -163,7 +163,7 @@ func TestPrinterMultipleFilesInSequence(t *testing.T) {
 func TestPrinterMultipleFilesInSequenceWithLeadingContent(t *testing.T) {
 	var output bytes.Buffer
 	var writer = bufio.NewWriter(&output)
-	printer := NewPrinterWithSingleWriter(writer, YamlOutputFormat, true, false, 2, true)
+	printer := NewSimpleYamlPrinter(writer, YamlOutputFormat, true, false, 2, true)
 
 	inputs, err := readDocuments(strings.NewReader(multiDocSample), "sample.yml", 0, NewYamlDecoder())
 	if err != nil {
@@ -213,7 +213,7 @@ func TestPrinterMultipleFilesInSequenceWithLeadingContent(t *testing.T) {
 func TestPrinterMultipleDocsInSinglePrint(t *testing.T) {
 	var output bytes.Buffer
 	var writer = bufio.NewWriter(&output)
-	printer := NewPrinterWithSingleWriter(writer, YamlOutputFormat, true, false, 2, true)
+	printer := NewSimpleYamlPrinter(writer, YamlOutputFormat, true, false, 2, true)
 
 	inputs, err := readDocuments(strings.NewReader(multiDocSample), "sample.yml", 0, NewYamlDecoder())
 	if err != nil {
@@ -232,7 +232,7 @@ func TestPrinterMultipleDocsInSinglePrint(t *testing.T) {
 func TestPrinterMultipleDocsInSinglePrintWithLeadingDoc(t *testing.T) {
 	var output bytes.Buffer
 	var writer = bufio.NewWriter(&output)
-	printer := NewPrinterWithSingleWriter(writer, YamlOutputFormat, true, false, 2, true)
+	printer := NewSimpleYamlPrinter(writer, YamlOutputFormat, true, false, 2, true)
 
 	inputs, err := readDocuments(strings.NewReader(multiDocSample), "sample.yml", 0, NewYamlDecoder())
 	if err != nil {
@@ -261,7 +261,7 @@ a: coconut
 func TestPrinterMultipleDocsInSinglePrintWithLeadingDocTrailing(t *testing.T) {
 	var output bytes.Buffer
 	var writer = bufio.NewWriter(&output)
-	printer := NewPrinterWithSingleWriter(writer, YamlOutputFormat, true, false, 2, true)
+	printer := NewSimpleYamlPrinter(writer, YamlOutputFormat, true, false, 2, true)
 
 	inputs, err := readDocuments(strings.NewReader(multiDocSample), "sample.yml", 0, NewYamlDecoder())
 	if err != nil {
@@ -287,7 +287,7 @@ a: coconut
 func TestPrinterScalarWithLeadingCont(t *testing.T) {
 	var output bytes.Buffer
 	var writer = bufio.NewWriter(&output)
-	printer := NewPrinterWithSingleWriter(writer, YamlOutputFormat, true, false, 2, true)
+	printer := NewSimpleYamlPrinter(writer, YamlOutputFormat, true, false, 2, true)
 
 	node, err := NewExpressionParser().ParseExpression(".a")
 	if err != nil {
@@ -314,7 +314,7 @@ func TestPrinterMultipleDocsJson(t *testing.T) {
 	var writer = bufio.NewWriter(&output)
 	// note printDocSeparators is true, it should still not print document separators
 	// when outputing JSON.
-	printer := NewPrinterWithSingleWriter(writer, JsonOutputFormat, true, false, 0, true)
+	printer := NewPrinter(NewJsonEncoder(0), NewSinglePrinterWriter(writer))
 
 	inputs, err := readDocuments(strings.NewReader(multiDocSample), "sample.yml", 0, NewYamlDecoder())
 	if err != nil {

pkg/yqlib/utils.go

@@ -38,43 +38,6 @@ func writeString(writer io.Writer, txt string) error {
 	return errorWriting
 }
 
-func processLeadingContent(mappedDoc *CandidateNode, writer io.Writer, printDocSeparators bool, outputFormat PrinterOutputFormat) error {
-	log.Debug("headcommentwas %v", mappedDoc.LeadingContent)
-	log.Debug("finished headcomment")
-	reader := bufio.NewReader(strings.NewReader(mappedDoc.LeadingContent))
-
-	for {
-
-		readline, errReading := reader.ReadString('\n')
-		if errReading != nil && !errors.Is(errReading, io.EOF) {
-			return errReading
-		}
-		if strings.Contains(readline, "$yqDocSeperator$") {
-			if printDocSeparators {
-				if err := writeString(writer, "---\n"); err != nil {
-					return err
-				}
-			}
-		} else if outputFormat == YamlOutputFormat {
-			if err := writeString(writer, readline); err != nil {
-				return err
-			}
-		}
-
-		if errors.Is(errReading, io.EOF) {
-			if readline != "" {
-				// the last comment we read didn't have a new line, put one in
-				if err := writeString(writer, "\n"); err != nil {
-					return err
-				}
-			}
-			break
-		}
-	}
-
-	return nil
-}
-
 func processReadStream(reader *bufio.Reader) (io.Reader, string, error) {
 	var commentLineRegEx = regexp.MustCompile(`^\s*#`)
 	var sb strings.Builder

pkg/yqlib/xml_test.go

@@ -24,23 +24,29 @@ func decodeXml(t *testing.T, xml string) *CandidateNode {
 	return &CandidateNode{Node: node}
 }
 
-func yamlToXml(sampleYaml string, indent int) string {
+func processScenario(s xmlScenario) string {
 	var output bytes.Buffer
 	writer := bufio.NewWriter(&output)
 
-	var encoder = NewXmlEncoder(writer, indent, "+", "+content")
+	var encoder = NewXmlEncoder(2, "+", "+content")
-	inputs, err := readDocuments(strings.NewReader(sampleYaml), "sample.yml", 0, NewYamlDecoder())
+
+	var decoder = NewYamlDecoder()
+	if s.scenarioType == "roundtrip" {
+		decoder = NewXmlDecoder("+", "+content")
+	}
+
+	inputs, err := readDocuments(strings.NewReader(s.input), "sample.yml", 0, decoder)
 	if err != nil {
 		panic(err)
 	}
 	node := inputs.Front().Value.(*CandidateNode).Node
-	err = encoder.Encode(node)
+	err = encoder.Encode(writer, node)
 	if err != nil {
 		panic(err)
 	}
 	writer.Flush()
 
-	return strings.TrimSuffix(output.String(), "\n")
+	return output.String()
 }
 
 type xmlScenario struct {
@@ -49,9 +55,154 @@ type xmlScenario struct {
 	description    string
 	subdescription string
 	skipDoc        bool
-	encodeScenario bool
+	scenarioType   string
 }
 
+var inputXmlWithComments = `
+<!-- before cat -->
+<cat>
+	<!-- in cat before -->
+	<x>3<!-- multi
+line comment 
+for x --></x>
+	<!-- before y -->
+	<y>
+		<!-- in y before -->
+		<d><!-- in d before -->z<!-- in d after --></d>
+		
+		<!-- in y after -->
+	</y>
+	<!-- in_cat_after -->
+</cat>
+<!-- after cat -->
+`
+var inputXmlWithCommentsWithSubChild = `
+<!-- before cat -->
+<cat>
+	<!-- in cat before -->
+	<x>3<!-- multi
+line comment 
+for x --></x>
+	<!-- before y -->
+	<y>
+		<!-- in y before -->
+		<d><!-- in d before --><z sweet="cool"/><!-- in d after --></d>
+		
+		<!-- in y after -->
+	</y>
+	<!-- in_cat_after -->
+</cat>
+<!-- after cat -->
+`
+
+var expectedDecodeYamlWithSubChild = `D0, P[], (doc)::# before cat
+cat:
+    # in cat before
+    x: "3" # multi
+    # line comment 
+    # for x
+    # before y
+
+    y:
+        # in y before
+        d:
+            # in d before
+            z:
+                +sweet: cool
+            # in d after
+        # in y after
+    # in_cat_after
+# after cat
+`
+
+var inputXmlWithCommentsWithArray = `
+<!-- before cat -->
+<cat>
+	<!-- in cat before -->
+	<x>3<!-- multi
+line comment 
+for x --></x>
+	<!-- before y -->
+	<y>
+		<!-- in y before -->
+		<d><!-- in d before --><z sweet="cool"/><!-- in d after --></d>
+        <d><!-- in d2 before --><z sweet="cool2"/><!-- in d2 after --></d>
+		
+		<!-- in y after -->
+	</y>
+	<!-- in_cat_after -->
+</cat>
+<!-- after cat -->
+`
+
+var expectedDecodeYamlWithArray = `D0, P[], (doc)::# before cat
+cat:
+    # in cat before
+    x: "3" # multi
+    # line comment 
+    # for x
+    # before y
+
+    y:
+        # in y before
+        d:
+            - # in d before
+              z:
+                +sweet: cool
+              # in d after
+            - # in d2 before
+              z:
+                +sweet: cool2
+              # in d2 after
+        # in y after
+    # in_cat_after
+# after cat
+`
+
+var expectedDecodeYamlWithComments = `D0, P[], (doc)::# before cat
+cat:
+    # in cat before
+    x: "3" # multi
+    # line comment 
+    # for x
+    # before y
+
+    y:
+        # in y before
+        # in d before
+        d: z # in d after
+        # in y after
+    # in_cat_after
+# after cat
+`
+
+var expectedRoundtripXmlWithComments = `<!-- before cat --><cat><!-- in cat before -->
+  <x>3<!-- multi
+line comment 
+for x --></x><!-- before y -->
+  <y><!-- in y before
+in d before -->
+    <d>z<!-- in d after --></d><!-- in y after -->
+  </y><!-- in_cat_after -->
+</cat><!-- after cat -->
+`
+
+var yamlWithComments = `# above_cat
+cat: # inline_cat
+  # above_array
+  array: # inline_array
+    - val1 # inline_val1
+    # above_val2
+    - val2 # inline_val2
+# below_cat
+`
+
+var expectedXmlWithComments = `<!-- above_cat inline_cat --><cat><!-- above_array inline_array -->
+  <array>val1<!-- inline_val1 --></array>
+  <array><!-- above_val2 -->val2<!-- inline_val2 --></array>
+</cat><!-- below_cat -->
+`
+
 var xmlScenarios = []xmlScenario{
 	{
 		description: "Parse xml: simple",
@@ -66,53 +217,88 @@ var xmlScenarios = []xmlScenario{
 	},
 	{
 		description:    "Parse xml: attributes",
-		subdescription: "Attributes are converted to fields, with the attribute prefix.",
+		subdescription: "Attributes are converted to fields, with the default attribute prefix '+'. Use '--xml-attribute-prefix` to set your own.",
 		input:          "<?xml version=\"1.0\" encoding=\"UTF-8\"?>\n<cat legs=\"4\">\n  <legs>7</legs>\n</cat>",
 		expected:       "D0, P[], (doc)::cat:\n    +legs: \"4\"\n    legs: \"7\"\n",
 	},
 	{
 		description:    "Parse xml: attributes with content",
-		subdescription: "Content is added as a field, using the content name",
+		subdescription: "Content is added as a field, using the default content name of '+content'. Use `--xml-content-name` to set your own.",
 		input:          "<?xml version=\"1.0\" encoding=\"UTF-8\"?>\n<cat legs=\"4\">meow</cat>",
 		expected:       "D0, P[], (doc)::cat:\n    +content: meow\n    +legs: \"4\"\n",
 	},
 	{
-		description:    "Encode xml: simple",
+		description:    "Parse xml: with comments",
-		input:          "cat: purrs",
+		subdescription: "A best attempt is made to preserve comments.",
-		expected:       "<cat>purrs</cat>",
+		input:          inputXmlWithComments,
-		encodeScenario: true,
+		expected:       expectedDecodeYamlWithComments,
+		scenarioType:   "decode",
 	},
 	{
-		description:    "Encode xml: array",
+		description:  "Parse xml: with comments subchild",
-		input:          "pets:\n  cat:\n    - purrs\n    - meows",
+		skipDoc:      true,
-		expected:       "<pets>\n  <cat>purrs</cat>\n  <cat>meows</cat>\n</pets>",
+		input:        inputXmlWithCommentsWithSubChild,
-		encodeScenario: true,
+		expected:     expectedDecodeYamlWithSubChild,
+		scenarioType: "decode",
+	},
+	{
+		description:  "Parse xml: with comments array",
+		skipDoc:      true,
+		input:        inputXmlWithCommentsWithArray,
+		expected:     expectedDecodeYamlWithArray,
+		scenarioType: "decode",
+	},
+	{
+		description:  "Encode xml: simple",
+		input:        "cat: purrs",
+		expected:     "<cat>purrs</cat>\n",
+		scenarioType: "encode",
+	},
+	{
+		description:  "Encode xml: array",
+		input:        "pets:\n  cat:\n    - purrs\n    - meows",
+		expected:     "<pets>\n  <cat>purrs</cat>\n  <cat>meows</cat>\n</pets>\n",
+		scenarioType: "encode",
 	},
 	{
 		description:    "Encode xml: attributes",
 		subdescription: "Fields with the matching xml-attribute-prefix are assumed to be attributes.",
 		input:          "cat:\n  +name: tiger\n  meows: true\n",
-		expected:       "<cat name=\"tiger\">\n  <meows>true</meows>\n</cat>",
+		expected:       "<cat name=\"tiger\">\n  <meows>true</meows>\n</cat>\n",
-		encodeScenario: true,
+		scenarioType:   "encode",
 	},
 	{
-		skipDoc:        true,
+		skipDoc:      true,
-		input:          "cat:\n  ++name: tiger\n  meows: true\n",
+		input:        "cat:\n  ++name: tiger\n  meows: true\n",
-		expected:       "<cat +name=\"tiger\">\n  <meows>true</meows>\n</cat>",
+		expected:     "<cat +name=\"tiger\">\n  <meows>true</meows>\n</cat>\n",
-		encodeScenario: true,
+		scenarioType: "encode",
 	},
 	{
 		description:    "Encode xml: attributes with content",
 		subdescription: "Fields with the matching xml-content-name is assumed to be content.",
 		input:          "cat:\n  +name: tiger\n  +content: cool\n",
-		expected:       "<cat name=\"tiger\">cool</cat>",
+		expected:       "<cat name=\"tiger\">cool</cat>\n",
-		encodeScenario: true,
+		scenarioType:   "encode",
+	},
+	{
+		description:    "Encode xml: comments",
+		subdescription: "A best attempt is made to copy comments to xml.",
+		input:          yamlWithComments,
+		expected:       expectedXmlWithComments,
+		scenarioType:   "encode",
+	},
+	{
+		description:    "Round trip: with comments",
+		subdescription: "A best effort is made, but comment positions and white space are not preserved perfectly.",
+		input:          inputXmlWithComments,
+		expected:       expectedRoundtripXmlWithComments,
+		scenarioType:   "roundtrip",
 	},
 }
 
-func testXmlScenario(t *testing.T, s *xmlScenario) {
+func testXmlScenario(t *testing.T, s xmlScenario) {
-	if s.encodeScenario {
+	if s.scenarioType == "encode" || s.scenarioType == "roundtrip" {
-		test.AssertResultWithContext(t, s.expected, yamlToXml(s.input, 2), s.description)
+		test.AssertResultWithContext(t, s.expected, processScenario(s), s.description)
 	} else {
 		var actual = resultToString(t, decodeXml(t, s.input))
 		test.AssertResultWithContext(t, s.expected, actual, s.description)
@@ -125,8 +311,10 @@ func documentXmlScenario(t *testing.T, w *bufio.Writer, i interface{}) {
 	if s.skipDoc {
 		return
 	}
-	if s.encodeScenario {
+	if s.scenarioType == "encode" {
 		documentXmlEncodeScenario(w, s)
+	} else if s.scenarioType == "roundtrip" {
+		documentXmlRoundTripScenario(w, s)
 	} else {
 		documentXmlDecodeScenario(t, w, s)
 	}
@@ -149,7 +337,7 @@ func documentXmlDecodeScenario(t *testing.T, w *bufio.Writer, s xmlScenario) {
 	writeOrPanic(w, "will output\n")
 
 	var output bytes.Buffer
-	printer := NewPrinterWithSingleWriter(bufio.NewWriter(&output), YamlOutputFormat, true, false, 2, true)
+	printer := NewSimpleYamlPrinter(bufio.NewWriter(&output), YamlOutputFormat, true, false, 2, true)
 
 	node := decodeXml(t, s.input)
 
@@ -177,12 +365,30 @@ func documentXmlEncodeScenario(w *bufio.Writer, s xmlScenario) {
 	writeOrPanic(w, "```bash\nyq e -o=xml '.' sample.yml\n```\n")
 	writeOrPanic(w, "will output\n")
 
-	writeOrPanic(w, fmt.Sprintf("```xml\n%v```\n\n", yamlToXml(s.input, 2)))
+	writeOrPanic(w, fmt.Sprintf("```xml\n%v```\n\n", processScenario(s)))
+}
+
+func documentXmlRoundTripScenario(w *bufio.Writer, s xmlScenario) {
+	writeOrPanic(w, fmt.Sprintf("## %v\n", s.description))
+
+	if s.subdescription != "" {
+		writeOrPanic(w, s.subdescription)
+		writeOrPanic(w, "\n\n")
+	}
+
+	writeOrPanic(w, "Given a sample.xml file of:\n")
+	writeOrPanic(w, fmt.Sprintf("```xml\n%v\n```\n", s.input))
+
+	writeOrPanic(w, "then\n")
+	writeOrPanic(w, "```bash\nyq e -p=xml -o=xml '.' sample.xml\n```\n")
+	writeOrPanic(w, "will output\n")
+
+	writeOrPanic(w, fmt.Sprintf("```xml\n%v```\n\n", processScenario(s)))
 }
 
 func TestXmlScenarios(t *testing.T) {
 	for _, tt := range xmlScenarios {
-		testXmlScenario(t, &tt)
+		testXmlScenario(t, tt)
 	}
 	genericScenarios := make([]interface{}, len(xmlScenarios))
 	for i, s := range xmlScenarios {

test/utils.go

@@ -1,6 +1,7 @@
 package test
 
 import (
+	"bufio"
 	"bytes"
 	"fmt"
 	"os"
@@ -8,6 +9,8 @@ import (
 	"strings"
 	"testing"
 
+	"github.com/pkg/diff"
+	"github.com/pkg/diff/write"
 	"github.com/spf13/cobra"
 	yaml "gopkg.in/yaml.v3"
 )
@@ -63,8 +66,14 @@ func AssertResultComplexWithContext(t *testing.T, expectedValue interface{}, act
 
 func AssertResultWithContext(t *testing.T, expectedValue interface{}, actualValue interface{}, context interface{}) {
 	t.Helper()
+	opts := []write.Option{write.TerminalColor()}
 	if expectedValue != actualValue {
 		t.Error(context)
-		t.Error(": expected <", expectedValue, "> but got <", actualValue, ">")
+		var differenceBuffer bytes.Buffer
+		if err := diff.Text("expected", "actual", expectedValue, actualValue, bufio.NewWriter(&differenceBuffer), opts...); err != nil {
+			t.Error(err)
+		} else {
+			t.Error(differenceBuffer.String())
+		}
 	}
 }