Refactored decoder responsibilities (#1402)

- improved comment handling
- yaml decoder now responsible for leading content work around

Author: mikefarah

acceptance_tests/empty.sh

@@ -9,7 +9,7 @@ EOL
 
 testEmptyEval() {
   X=$(./yq e test.yml)
-  expected=$(cat test.yml)
+  expected="# comment"
   assertEquals 0 $?
   assertEquals "$expected" "$X"
 }

cmd/constant.go

@@ -1,6 +1,5 @@
 package cmd
 
-var leadingContentPreProcessing = true
 var unwrapScalar = true
 
 var writeInplace = false

cmd/evaluate_all_command.go

@@ -75,7 +75,7 @@ func evaluateAll(cmd *cobra.Command, args []string) (cmdError error) {
 		return err
 	}
 
-	decoder, err := configureDecoder()
+	decoder, err := configureDecoder(true)
 	if err != nil {
 		return err
 	}
@@ -109,13 +109,13 @@ func evaluateAll(cmd *cobra.Command, args []string) (cmdError error) {
 	switch len(args) {
 	case 0:
 		if nullInput {
-			err = yqlib.NewStreamEvaluator().EvaluateNew(processExpression(expression), printer, "")
+			err = yqlib.NewStreamEvaluator().EvaluateNew(processExpression(expression), printer)
 		} else {
 			cmd.Println(cmd.UsageString())
 			return nil
 		}
 	default:
-		err = allAtOnceEvaluator.EvaluateFiles(processExpression(expression), args, printer, leadingContentPreProcessing, decoder)
+		err = allAtOnceEvaluator.EvaluateFiles(processExpression(expression), args, printer, decoder)
 	}
 
 	completedSuccessfully = err == nil

cmd/evalute_sequence_command.go

@@ -97,7 +97,7 @@ func evaluateSequence(cmd *cobra.Command, args []string) (cmdError error) {
 
 	printer := yqlib.NewPrinter(encoder, printerWriter)
 
-	decoder, err := configureDecoder()
+	decoder, err := configureDecoder(false)
 	if err != nil {
 		return err
 	}
@@ -123,13 +123,13 @@ func evaluateSequence(cmd *cobra.Command, args []string) (cmdError error) {
 	switch len(args) {
 	case 0:
 		if nullInput {
-			err = streamEvaluator.EvaluateNew(processExpression(expression), printer, "")
+			err = streamEvaluator.EvaluateNew(processExpression(expression), printer)
 		} else {
 			cmd.Println(cmd.UsageString())
 			return nil
 		}
 	default:
-		err = streamEvaluator.EvaluateFiles(processExpression(expression), args, printer, leadingContentPreProcessing, decoder)
+		err = streamEvaluator.EvaluateFiles(processExpression(expression), args, printer, decoder)
 	}
 	completedSuccessfully = err == nil
 

cmd/root.go

@@ -59,6 +59,11 @@ yq -P sample.json
 					"naming conflicts with the default content name, directive name and proc inst prefix. If you need to keep " +
 					"`+` please set that value explicityly with --xml-attribute-prefix.")
 			}
+
+			//copy preference form global setting
+			yqlib.ConfiguredYamlPreferences.UnwrapScalar = unwrapScalar
+
+			yqlib.ConfiguredYamlPreferences.PrintDocSeparators = !noDocSeparators
 		},
 	}
 
@@ -97,7 +102,7 @@ yq -P sample.json
 	rootCmd.PersistentFlags().BoolVarP(&forceNoColor, "no-colors", "M", false, "force print with no colors")
 	rootCmd.PersistentFlags().StringVarP(&frontMatter, "front-matter", "f", "", "(extract|process) first input as yaml front-matter. Extract will pull out the yaml content, process will run the expression against the yaml content, leaving the remaining data intact")
 	rootCmd.PersistentFlags().StringVarP(&forceExpression, "expression", "", "", "forcibly set the expression argument. Useful when yq argument detection thinks your expression is a file.")
-	rootCmd.PersistentFlags().BoolVarP(&leadingContentPreProcessing, "header-preprocess", "", true, "Slurp any header comments and separators before processing expression.")
+	rootCmd.PersistentFlags().BoolVarP(&yqlib.ConfiguredYamlPreferences.LeadingContentPreProcessing, "header-preprocess", "", true, "Slurp any header comments and separators before processing expression.")
 
 	rootCmd.PersistentFlags().StringVarP(&splitFileExp, "split-exp", "s", "", "print each result (or doc) into a file named (exp). [exp] argument must return a string. You can use $index in the expression as the result counter.")
 	rootCmd.PersistentFlags().StringVarP(&splitFileExpFile, "split-exp-file", "", "", "Use a file to specify the split-exp expression.")

cmd/utils.go

@@ -56,7 +56,7 @@ func initCommand(cmd *cobra.Command, args []string) (string, []string, error) {
 	return expression, args, nil
 }
 
-func configureDecoder() (yqlib.Decoder, error) {
+func configureDecoder(evaluateTogether bool) (yqlib.Decoder, error) {
 	yqlibInputFormat, err := yqlib.InputFormatFromString(inputFormat)
 	if err != nil {
 		return nil, err
@@ -73,8 +73,9 @@ func configureDecoder() (yqlib.Decoder, error) {
 	case yqlib.TSVObjectInputFormat:
 		return yqlib.NewCSVObjectDecoder('\t'), nil
 	}
-
-	return yqlib.NewYamlDecoder(), nil
+	prefs := yqlib.ConfiguredYamlPreferences
+	prefs.EvaluateTogether = evaluateTogether
+	return yqlib.NewYamlDecoder(prefs), nil
 }
 
 func configurePrinterWriter(format yqlib.PrinterOutputFormat, out io.Writer) (yqlib.PrinterWriter, error) {
@@ -105,7 +106,7 @@ func configureEncoder(format yqlib.PrinterOutputFormat) yqlib.Encoder {
 	case yqlib.TSVOutputFormat:
 		return yqlib.NewCsvEncoder('\t')
 	case yqlib.YamlOutputFormat:
-		return yqlib.NewYamlEncoder(indent, colorsEnabled, !noDocSeparators, unwrapScalar)
+		return yqlib.NewYamlEncoder(indent, colorsEnabled, yqlib.ConfiguredYamlPreferences)
 	case yqlib.XMLOutputFormat:
 		return yqlib.NewXMLEncoder(indent, yqlib.ConfiguredXMLPreferences)
 	}

examples/empty-no-comment.yaml

@@ -8,7 +8,7 @@ import (
 
 // A yaml expression evaluator that runs the expression once against all files/nodes in memory.
 type Evaluator interface {
-	EvaluateFiles(expression string, filenames []string, printer Printer, leadingContentPreProcessing bool, decoder Decoder) error
+	EvaluateFiles(expression string, filenames []string, printer Printer, decoder Decoder) error
 
 	// EvaluateNodes takes an expression and one or more yaml nodes, returning a list of matching candidate nodes
 	EvaluateNodes(expression string, nodes ...*yaml.Node) (*list.List, error)
@@ -46,21 +46,16 @@ func (e *allAtOnceEvaluator) EvaluateCandidateNodes(expression string, inputCand
 	return context.MatchingNodes, nil
 }
 
-func (e *allAtOnceEvaluator) EvaluateFiles(expression string, filenames []string, printer Printer, leadingContentPreProcessing bool, decoder Decoder) error {
+func (e *allAtOnceEvaluator) EvaluateFiles(expression string, filenames []string, printer Printer, decoder Decoder) error {
 	fileIndex := 0
-	firstFileLeadingContent := ""
 
 	var allDocuments = list.New()
 	for _, filename := range filenames {
-		reader, leadingContent, err := readStream(filename, fileIndex == 0 && leadingContentPreProcessing)
+		reader, err := readStream(filename)
 		if err != nil {
 			return err
 		}
 
-		if fileIndex == 0 {
-			firstFileLeadingContent = leadingContent
-		}
-
 		fileDocuments, err := readDocuments(reader, filename, fileIndex, decoder)
 		if err != nil {
 			return err
@@ -75,11 +70,9 @@ func (e *allAtOnceEvaluator) EvaluateFiles(expression string, filenames []string
 			Filename:       "",
 			Node:           &yaml.Node{Kind: yaml.DocumentNode, Content: []*yaml.Node{{Tag: "!!null", Kind: yaml.ScalarNode}}},
 			FileIndex:      0,
-			LeadingContent: firstFileLeadingContent,
+			LeadingContent: "",
 		}
 		allDocuments.PushBack(candidateNode)
-	} else {
-		allDocuments.Front().Value.(*CandidateNode).LeadingContent = firstFileLeadingContent
 	}
 
 	matches, err := e.EvaluateCandidateNodes(expression, allDocuments)

pkg/yqlib/all_at_once_evaluator.go

@@ -136,13 +136,13 @@ var csvScenarios = []formatScenario{
 func testCSVScenario(t *testing.T, s formatScenario) {
 	switch s.scenarioType {
 	case "encode-csv":
-		test.AssertResultWithContext(t, s.expected, processFormatScenario(s, NewYamlDecoder(), NewCsvEncoder(',')), s.description)
+		test.AssertResultWithContext(t, s.expected, processFormatScenario(s, NewYamlDecoder(ConfiguredYamlPreferences), NewCsvEncoder(',')), s.description)
 	case "encode-tsv":
-		test.AssertResultWithContext(t, s.expected, processFormatScenario(s, NewYamlDecoder(), NewCsvEncoder('\t')), s.description)
+		test.AssertResultWithContext(t, s.expected, processFormatScenario(s, NewYamlDecoder(ConfiguredYamlPreferences), NewCsvEncoder('\t')), s.description)
 	case "decode-csv-object":
-		test.AssertResultWithContext(t, s.expected, processFormatScenario(s, NewCSVObjectDecoder(','), NewYamlEncoder(2, false, true, true)), s.description)
+		test.AssertResultWithContext(t, s.expected, processFormatScenario(s, NewCSVObjectDecoder(','), NewYamlEncoder(2, false, ConfiguredYamlPreferences)), s.description)
 	case "decode-tsv-object":
-		test.AssertResultWithContext(t, s.expected, processFormatScenario(s, NewCSVObjectDecoder('\t'), NewYamlEncoder(2, false, true, true)), s.description)
+		test.AssertResultWithContext(t, s.expected, processFormatScenario(s, NewCSVObjectDecoder('\t'), NewYamlEncoder(2, false, ConfiguredYamlPreferences)), s.description)
 	case "roundtrip-csv":
 		test.AssertResultWithContext(t, s.expected, processFormatScenario(s, NewCSVObjectDecoder(','), NewCsvEncoder(',')), s.description)
 	default:
@@ -171,7 +171,7 @@ func documentCSVDecodeObjectScenario(w *bufio.Writer, s formatScenario, formatTy
 	}
 
 	writeOrPanic(w, fmt.Sprintf("```yaml\n%v```\n\n",
-		processFormatScenario(s, NewCSVObjectDecoder(separator), NewYamlEncoder(s.indent, false, true, true))),
+		processFormatScenario(s, NewCSVObjectDecoder(separator), NewYamlEncoder(s.indent, false, ConfiguredYamlPreferences))),
 	)
 }
 
@@ -203,7 +203,7 @@ func documentCSVEncodeScenario(w *bufio.Writer, s formatScenario, formatType str
 	}
 
 	writeOrPanic(w, fmt.Sprintf("```%v\n%v```\n\n", formatType,
-		processFormatScenario(s, NewYamlDecoder(), NewCsvEncoder(separator))),
+		processFormatScenario(s, NewYamlDecoder(ConfiguredYamlPreferences), NewCsvEncoder(separator))),
 	)
 }
 

pkg/yqlib/csv_test.go

@@ -3,8 +3,6 @@ package yqlib
 import (
 	"fmt"
 	"io"
-
-	yaml "gopkg.in/yaml.v3"
 )
 
 type InputFormat uint
@@ -20,8 +18,8 @@ const (
 )
 
 type Decoder interface {
-	Init(reader io.Reader)
-	Decode(node *yaml.Node) error
+	Init(reader io.Reader) error
+	Decode() (*CandidateNode, error)
 }
 
 func InputFormatFromString(format string) (InputFormat, error) {