yaml: Error on trailing yaml/json

Turns our yaml is a superset of json

format/csv/csv.go

@@ -23,7 +23,7 @@ func init() {
 	interp.RegisterFormat(decode.Format{
 		Name:        format.CSV,
 		Description: "Comma separated values",
-		ProbeOrder:  format.ProbeOrderText,
+		ProbeOrder:  format.ProbeOrderTextFuzzy,
 		DecodeFn:    decodeCSV,
 		DecodeInArg: format.CSVLIn{
 			Comma:   ",",

format/format.go

@@ -3,8 +3,9 @@ package format
 // TODO: do before-format somehow and topology sort?
 const (
 	ProbeOrderBinUnique = 0   // binary with unlikely overlap
-	ProbeOrderBinFuzzy  = 50  // binary with possible overlap
-	ProbeOrderText      = 100 // text format
+	ProbeOrderBinFuzzy  = 100 // binary with possible overlap
+	ProbeOrderTextJSON  = 200 // text json has prio as yaml overlap
+	ProbeOrderTextFuzzy = 300 // text with possible overlap
 )
 
 // TODO: change to CamelCase?

format/json/json.go

@@ -25,7 +25,7 @@ func init() {
 	interp.RegisterFormat(decode.Format{
 		Name:        format.JSON,
 		Description: "JavaScript Object Notation",
-		ProbeOrder:  format.ProbeOrderText,
+		ProbeOrder:  format.ProbeOrderTextJSON,
 		Groups:      []string{format.PROBE},
 		DecodeFn:    decodeJSON,
 		Functions:   []string{"_todisplay"},

format/json/testdata/json.fqtest

@@ -8,6 +8,18 @@ $ fq . /probe.json
 }
 $ fq . /probe_scalar.json
 123
+# make sure all value types parse to json format
+$ fq -rRs 'fromjson[] | tojson | fromjson | format' variants.json
+json
+json
+json
+json
+json
+json
+json
+json
+json
+json
 $ fq -rRs 'fromjson[] | (tojson | ., fromjson), "----", (tojson({indent:2}) | ., fromjson), "----"' variants.json
 null
 null

format/json/testdata/trailing.fqtest

@@ -2,3 +2,7 @@ $ fq -n '"123 trailing" | fromjson._error.error'
 exitcode: 5
 stderr:
 error: error at position 0xc: trialing data after top-level value
+$ fq -n '`{"a":123}{"b":444}` | fromjson'
+exitcode: 5
+stderr:
+error: error at position 0x12: trialing data after top-level value

format/toml/toml.go

@@ -20,7 +20,7 @@ func init() {
 	interp.RegisterFormat(decode.Format{
 		Name:        format.TOML,
 		Description: "Tom's Obvious, Minimal Language",
-		ProbeOrder:  format.ProbeOrderText,
+		ProbeOrder:  format.ProbeOrderTextFuzzy,
 		Groups:      []string{format.PROBE},
 		DecodeFn:    decodeTOML,
 		Functions:   []string{"_todisplay"},

format/xml/xml.go

@@ -34,7 +34,7 @@ func init() {
 	interp.RegisterFormat(decode.Format{
 		Name:        format.XML,
 		Description: "Extensible Markup Language",
-		ProbeOrder:  format.ProbeOrderText,
+		ProbeOrder:  format.ProbeOrderTextFuzzy,
 		Groups:      []string{format.PROBE},
 		DecodeFn:    decodeXML,
 		DecodeInArg: format.XMLIn{

format/yaml/testdata/trailing.fqtest

@@ -2,3 +2,7 @@ $ fq -n '"- a\ntrailing" | fromyaml._error.error'
 exitcode: 5
 stderr:
 error: error at position 0xc: yaml: line 2: could not find expected ':'
+$ fq -n '`{"a":123}{"b":444}` | fromyaml'
+exitcode: 5
+stderr:
+error: error at position 0x12: trialing data after top-level value

format/yaml/yaml.go

@@ -4,6 +4,8 @@ package yaml
 
 import (
 	"embed"
+	"errors"
+	"io"
 
 	"github.com/wader/fq/format"
 	"github.com/wader/fq/internal/gojqex"
@@ -21,7 +23,7 @@ func init() {
 	interp.RegisterFormat(decode.Format{
 		Name:        format.YAML,
 		Description: "YAML Ain't Markup Language",
-		ProbeOrder:  format.ProbeOrderText,
+		ProbeOrder:  format.ProbeOrderTextFuzzy,
 		Groups:      []string{format.PROBE},
 		DecodeFn:    decodeYAML,
 		Functions:   []string{"_todisplay"},
@@ -34,9 +36,14 @@ func decodeYAML(d *decode.D, _ any) any {
 	br := d.RawLen(d.Len())
 	var r any
 
-	if err := yaml.NewDecoder(bitio.NewIOReader(br)).Decode(&r); err != nil {
+	yd := yaml.NewDecoder(bitio.NewIOReader(br))
+	if err := yd.Decode(&r); err != nil {
 		d.Fatalf("%s", err)
 	}
+	if err := yd.Decode(new(any)); !errors.Is(err, io.EOF) {
+		d.Fatalf("trialing data after top-level value")
+	}
+
 	var s scalar.S
 	s.Actual = r