Merge pull request #522 from johnkerl/sort-within-records-recursive

Add a sort-within-record -r option

go/regtest/cases/cli-help/0001/expout

@@ -729,6 +729,7 @@ which is the same as:
 Usage: mlr sort-within-records [options]
 Outputs records sorted lexically ascending by keys.
 Options:
+-r        Recursively sort subobjects/submaps, e.g. for JSON input.
 -h|--help Show this message.
 ================================================================
 Usage: mlr stats1 [options]

go/regtest/cases/verb-sort-within-records/0001/cmd

@@ -1 +1 @@
-mlr        --from regtest/input/sort-within-records.dkvp sort-within-records
+mlr --from regtest/input/sort-within-records.dkvp sort-within-records

go/regtest/cases/verb-sort-within-records/0002/cmd

@@ -1 +1 @@
-mlr --json --from regtest/input/needs-sorting.json       sort-within-records
+mlr --json --from regtest/input/needs-sorting.json sort-within-records

go/regtest/cases/verb-sort-within-records/0003/cmd

@@ -0,0 +1 @@
+mlr --json --from regtest/input/needs-sorting.json sort-within-records -r

go/regtest/cases/verb-sort-within-records/0003/expout

@@ -0,0 +1,13 @@
+{
+  "a": 1,
+  "b": 2
+}
+{
+  "a": 1,
+  "b": 2
+}
+{
+  "a": 1,
+  "b": 2,
+  "c": 3
+}

go/src/transformers/sort-within-records.go

@@ -29,6 +29,7 @@ func transformerSortWithinRecordsUsage(
 	fmt.Fprintf(o, "Usage: %s %s [options]\n", lib.MlrExeName(), verbNameSortWithinRecords)
 	fmt.Fprintln(o, "Outputs records sorted lexically ascending by keys.")
 	fmt.Fprintf(o, "Options:\n")
+	fmt.Fprintf(o, "-r        Recursively sort subobjects/submaps, e.g. for JSON input.\n")
 	fmt.Fprintf(o, "-h|--help Show this message.\n")
 
 	if doExit {
@@ -47,6 +48,7 @@ func transformerSortWithinRecordsParseCLI(
 	// Skip the verb name from the current spot in the mlr command line
 	argi := *pargi
 	argi++
+	doRecurse := false
 
 	for argi < argc /* variable increment: 1 or 2 depending on flag */ {
 		opt := args[argi]
@@ -58,6 +60,9 @@ func transformerSortWithinRecordsParseCLI(
 		if opt == "-h" || opt == "--help" {
 			transformerSortWithinRecordsUsage(os.Stdout, true, 0)
 
+		} else if opt == "-r" {
+			doRecurse = true
+
 		} else {
 			transformerSortWithinRecordsUsage(os.Stderr, true, 1)
 		}
@@ -66,7 +71,7 @@ func transformerSortWithinRecordsParseCLI(
 	// TODO: allow sort by key or value?
 	// TODO: allow sort ascendending/descending?
 
-	transformer, err := NewTransformerSortWithinRecords()
+	transformer, err := NewTransformerSortWithinRecords(doRecurse)
 	if err != nil {
 		fmt.Fprintln(os.Stderr, err)
 		os.Exit(1)
@@ -78,11 +83,19 @@ func transformerSortWithinRecordsParseCLI(
 
 // ----------------------------------------------------------------
 type TransformerSortWithinRecords struct {
+	recordTransformerFunc transforming.RecordTransformerFunc
 }
 
-func NewTransformerSortWithinRecords() (*TransformerSortWithinRecords, error) {
+func NewTransformerSortWithinRecords(
+	doRecurse bool,
+) (*TransformerSortWithinRecords, error) {
 
 	this := &TransformerSortWithinRecords{}
+	if doRecurse {
+		this.recordTransformerFunc = this.transformRecursively
+	} else {
+		this.recordTransformerFunc = this.transformNonrecursively
+	}
 
 	return this, nil
 }
@@ -91,6 +104,14 @@ func NewTransformerSortWithinRecords() (*TransformerSortWithinRecords, error) {
 func (this *TransformerSortWithinRecords) Transform(
 	inrecAndContext *types.RecordAndContext,
 	outputChannel chan<- *types.RecordAndContext,
+) {
+	this.recordTransformerFunc(inrecAndContext, outputChannel)
+}
+
+// ----------------------------------------------------------------
+func (this *TransformerSortWithinRecords) transformNonrecursively(
+	inrecAndContext *types.RecordAndContext,
+	outputChannel chan<- *types.RecordAndContext,
 ) {
 	if !inrecAndContext.EndOfStream {
 		inrec := inrecAndContext.Record
@@ -98,3 +119,15 @@ func (this *TransformerSortWithinRecords) Transform(
 	}
 	outputChannel <- inrecAndContext // including end-of-stream marker
 }
+
+// ----------------------------------------------------------------
+func (this *TransformerSortWithinRecords) transformRecursively(
+	inrecAndContext *types.RecordAndContext,
+	outputChannel chan<- *types.RecordAndContext,
+) {
+	if !inrecAndContext.EndOfStream {
+		inrec := inrecAndContext.Record
+		inrec.SortByKeyRecursively()
+	}
+	outputChannel <- inrecAndContext // including end-of-stream marker
+}

go/src/types/mlrmap_accessors.go

@@ -688,6 +688,26 @@ func (this *Mlrmap) SortByKey() {
 	*this = *that
 }
 
+// ----------------------------------------------------------------
+func (this *Mlrmap) SortByKeyRecursively() {
+	keys := this.GetKeys()
+
+	lib.SortStrings(keys)
+
+	that := NewMlrmapAsRecord()
+
+	for _, key := range keys {
+		// Old record will be GC'ed: just move pointers
+		value := this.Get(key)
+		if value.IsMap() {
+			value.mapval.SortByKeyRecursively()
+		}
+		that.PutReference(key, value)
+	}
+
+	*this = *that
+}
+
 // ================================================================
 // PRIVATE METHODS