Add "pretty" json configuration and change default behavior to be space-efficient (#2275)

* expose underlying format options

Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>

* remove escape html options and address PR feedback

Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>

* incorporate PR feedback

Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>

* fix cli test

Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>

---------

Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>

syft/format/cyclonedxjson/encoder.go

@@ -15,6 +15,7 @@ func SupportedVersions() []string {
 
 type EncoderConfig struct {
 	Version string
+	Pretty  bool // don't include spaces and newlines; same as jq -c
 }
 
 type encoder struct {
@@ -23,7 +24,7 @@ type encoder struct {
 }
 
 func NewFormatEncoderWithConfig(cfg EncoderConfig) (sbom.FormatEncoder, error) {
-	enc, err := cyclonedxutil.NewEncoder(cfg.Version, cyclonedx.BOMFileFormatJSON)
+	enc, err := cyclonedxutil.NewEncoder(cfg.Version, cyclonedx.BOMFileFormatJSON, cfg.Pretty)
 	if err != nil {
 		return nil, err
 	}
@@ -36,6 +37,7 @@ func NewFormatEncoderWithConfig(cfg EncoderConfig) (sbom.FormatEncoder, error) {
 func DefaultEncoderConfig() EncoderConfig {
 	return EncoderConfig{
 		Version: cyclonedxutil.DefaultVersion,
+		Pretty:  false,
 	}
 }
 

syft/format/cyclonedxjson/encoder_test.go

@@ -3,12 +3,15 @@ package cyclonedxjson
 import (
 	"bytes"
 	"flag"
+	"strings"
 	"testing"
 
 	"github.com/stretchr/testify/assert"
 	"github.com/stretchr/testify/require"
 
+	"github.com/anchore/syft/syft/format/internal/cyclonedxutil"
 	"github.com/anchore/syft/syft/format/internal/testutil"
+	"github.com/anchore/syft/syft/pkg"
 	"github.com/anchore/syft/syft/sbom"
 )
 
@@ -16,11 +19,68 @@ var updateSnapshot = flag.Bool("update-cyclonedx-json", false, "update the *.gol
 var updateImage = flag.Bool("update-image", false, "update the golden image used for image encoder testing")
 
 func getEncoder(t testing.TB) sbom.FormatEncoder {
-	enc, err := NewFormatEncoderWithConfig(DefaultEncoderConfig())
+	cfg := DefaultEncoderConfig()
+	cfg.Pretty = true
+
+	enc, err := NewFormatEncoderWithConfig(cfg)
 	require.NoError(t, err)
 	return enc
 }
 
+func TestPrettyOutput(t *testing.T) {
+	run := func(opt bool) string {
+		enc, err := NewFormatEncoderWithConfig(EncoderConfig{
+			Version: cyclonedxutil.DefaultVersion,
+			Pretty:  opt,
+		})
+		require.NoError(t, err)
+
+		dir := t.TempDir()
+		s := testutil.DirectoryInput(t, dir)
+
+		var buffer bytes.Buffer
+		err = enc.Encode(&buffer, s)
+		require.NoError(t, err)
+
+		return strings.TrimSpace(buffer.String())
+	}
+
+	t.Run("pretty", func(t *testing.T) {
+		actual := run(true)
+		assert.Contains(t, actual, "\n")
+	})
+
+	t.Run("compact", func(t *testing.T) {
+		actual := run(false)
+		assert.NotContains(t, actual, "\n")
+	})
+}
+
+func TestEscapeHTML(t *testing.T) {
+	dir := t.TempDir()
+	s := testutil.DirectoryInput(t, dir)
+	s.Artifacts.Packages.Add(pkg.Package{
+		Name: "<html-package>",
+	})
+
+	// by default we do not escape HTML
+	t.Run("default", func(t *testing.T) {
+		cfg := DefaultEncoderConfig()
+
+		enc, err := NewFormatEncoderWithConfig(cfg)
+		require.NoError(t, err)
+
+		var buffer bytes.Buffer
+		err = enc.Encode(&buffer, s)
+		require.NoError(t, err)
+
+		actual := buffer.String()
+		assert.Contains(t, actual, "<html-package>")
+		assert.NotContains(t, actual, "\\u003chtml-package\\u003e")
+	})
+
+}
+
 func TestCycloneDxDirectoryEncoder(t *testing.T) {
 	dir := t.TempDir()
 	testutil.AssertEncoderAgainstGoldenSnapshot(t,

syft/format/cyclonedxxml/encoder.go

@@ -17,6 +17,7 @@ func SupportedVersions() []string {
 
 type EncoderConfig struct {
 	Version string
+	Pretty  bool // set indent to 0
 }
 
 type encoder struct {
@@ -25,7 +26,7 @@ type encoder struct {
 }
 
 func NewFormatEncoderWithConfig(cfg EncoderConfig) (sbom.FormatEncoder, error) {
-	enc, err := cyclonedxutil.NewEncoder(cfg.Version, cyclonedx.BOMFileFormatXML)
+	enc, err := cyclonedxutil.NewEncoder(cfg.Version, cyclonedx.BOMFileFormatXML, cfg.Pretty)
 	if err != nil {
 		return nil, err
 	}
@@ -38,6 +39,7 @@ func NewFormatEncoderWithConfig(cfg EncoderConfig) (sbom.FormatEncoder, error) {
 func DefaultEncoderConfig() EncoderConfig {
 	return EncoderConfig{
 		Version: cyclonedxutil.DefaultVersion,
+		Pretty:  false,
 	}
 }
 

syft/format/cyclonedxxml/encoder_test.go

@@ -3,11 +3,14 @@ package cyclonedxxml
 import (
 	"bytes"
 	"flag"
+	"regexp"
+	"strings"
 	"testing"
 
 	"github.com/stretchr/testify/assert"
 	"github.com/stretchr/testify/require"
 
+	"github.com/anchore/syft/syft/format/internal/cyclonedxutil"
 	"github.com/anchore/syft/syft/format/internal/testutil"
 	"github.com/anchore/syft/syft/sbom"
 )
@@ -16,11 +19,42 @@ var updateSnapshot = flag.Bool("update-cyclonedx-xml", false, "update the *.gold
 var updateImage = flag.Bool("update-image", false, "update the golden image used for image encoder testing")
 
 func getEncoder(t testing.TB) sbom.FormatEncoder {
-	enc, err := NewFormatEncoderWithConfig(DefaultEncoderConfig())
+	cfg := DefaultEncoderConfig()
+	cfg.Pretty = true
+
+	enc, err := NewFormatEncoderWithConfig(cfg)
 	require.NoError(t, err)
 	return enc
 }
 
+func TestPrettyOutput(t *testing.T) {
+	enc, err := NewFormatEncoderWithConfig(EncoderConfig{
+		Version: cyclonedxutil.DefaultVersion,
+		Pretty:  false,
+	})
+	require.NoError(t, err)
+
+	dir := t.TempDir()
+	s := testutil.DirectoryInput(t, dir)
+
+	var buffer bytes.Buffer
+	err = enc.Encode(&buffer, s)
+	require.NoError(t, err)
+
+	actual := buffer.String()
+	lines := strings.Split(actual, "\n")
+	require.NotEmpty(t, lines)
+	whitespace := regexp.MustCompile(`^\s+`)
+	for _, line := range lines {
+		if len(line) == 0 {
+			continue
+		}
+
+		// require a non-whitespace character (tab, space, etc) as the first character of the line
+		require.False(t, whitespace.Match([]byte(line)), "line should not start with whitespace: %q", line)
+	}
+}
+
 func TestCycloneDxDirectoryEncoder(t *testing.T) {
 	dir := t.TempDir()
 	testutil.AssertEncoderAgainstGoldenSnapshot(t,

syft/format/internal/cyclonedxutil/encoder.go

@@ -14,9 +14,10 @@ const DefaultVersion = "1.5"
 type Encoder struct {
 	version cyclonedx.SpecVersion
 	format  cyclonedx.BOMFileFormat
+	pretty  bool
 }
 
-func NewEncoder(version string, format cyclonedx.BOMFileFormat) (Encoder, error) {
+func NewEncoder(version string, format cyclonedx.BOMFileFormat, pretty bool) (Encoder, error) {
 	specVersion, err := SpecVersionFromString(version)
 	if err != nil {
 		return Encoder{}, err
@@ -24,13 +25,14 @@ func NewEncoder(version string, format cyclonedx.BOMFileFormat) (Encoder, error)
 	return Encoder{
 		version: specVersion,
 		format:  format,
+		pretty:  pretty,
 	}, nil
 }
 
 func (e Encoder) Encode(writer io.Writer, s sbom.SBOM) error {
 	bom := cyclonedxhelpers.ToFormatModel(s)
 	enc := cyclonedx.NewBOMEncoder(writer, e.format)
-	enc.SetPretty(true)
+	enc.SetPretty(e.pretty)
 	enc.SetEscapeHTML(false)
 
 	return enc.EncodeVersion(bom, e.version)

syft/format/internal/testutil/snapshot.go

@@ -48,12 +48,8 @@ func AssertEncoderAgainstGoldenSnapshot(t *testing.T, cfg EncoderSnapshotTestCon
 		return
 	}
 
-	var expected []byte
 	if cfg.Redactor != nil {
 		actual = cfg.Redactor.Redact(actual)
-		expected = cfg.Redactor.Redact(testutils.GetGoldenFileContents(t))
-	} else {
-		expected = testutils.GetGoldenFileContents(t)
 	}
 
 	if cfg.UpdateSnapshot && cfg.PersistRedactionsInSnapshot {
@@ -62,6 +58,13 @@ func AssertEncoderAgainstGoldenSnapshot(t *testing.T, cfg EncoderSnapshotTestCon
 		return
 	}
 
+	var expected []byte
+	if cfg.Redactor != nil {
+		expected = cfg.Redactor.Redact(testutils.GetGoldenFileContents(t))
+	} else {
+		expected = testutils.GetGoldenFileContents(t)
+	}
+
 	if cfg.IsJSON {
 		require.JSONEq(t, string(expected), string(actual))
 	} else {

syft/format/spdxjson/encoder.go

@@ -23,6 +23,7 @@ func SupportedVersions() []string {
 
 type EncoderConfig struct {
 	Version string
+	Pretty  bool // don't include spaces and newlines; same as jq -c
 }
 
 type encoder struct {
@@ -38,6 +39,7 @@ func NewFormatEncoderWithConfig(cfg EncoderConfig) (sbom.FormatEncoder, error) {
 func DefaultEncoderConfig() EncoderConfig {
 	return EncoderConfig{
 		Version: spdxutil.DefaultVersion,
+		Pretty:  false,
 	}
 }
 
@@ -84,9 +86,12 @@ func (e encoder) Encode(writer io.Writer, s sbom.SBOM) error {
 	}
 
 	enc := json.NewEncoder(writer)
-	// prevent > and < from being escaped in the payload
+
 	enc.SetEscapeHTML(false)
-	enc.SetIndent("", " ")
+
+	if e.cfg.Pretty {
+		enc.SetIndent("", " ")
+	}
 
 	return enc.Encode(encodeDoc)
 }

syft/format/spdxjson/encoder_test.go

@@ -3,12 +3,15 @@ package spdxjson
 import (
 	"bytes"
 	"flag"
+	"strings"
 	"testing"
 
 	"github.com/stretchr/testify/assert"
 	"github.com/stretchr/testify/require"
 
+	"github.com/anchore/syft/syft/format/internal/spdxutil"
 	"github.com/anchore/syft/syft/format/internal/testutil"
+	"github.com/anchore/syft/syft/pkg"
 	"github.com/anchore/syft/syft/sbom"
 )
 
@@ -16,11 +19,68 @@ var updateSnapshot = flag.Bool("update-spdx-json", false, "update the *.golden f
 var updateImage = flag.Bool("update-image", false, "update the golden image used for image encoder testing")
 
 func getEncoder(t testing.TB) sbom.FormatEncoder {
-	enc, err := NewFormatEncoderWithConfig(DefaultEncoderConfig())
+	cfg := DefaultEncoderConfig()
+	cfg.Pretty = true
+
+	enc, err := NewFormatEncoderWithConfig(cfg)
 	require.NoError(t, err)
 	return enc
 }
 
+func TestPrettyOutput(t *testing.T) {
+	run := func(opt bool) string {
+		enc, err := NewFormatEncoderWithConfig(EncoderConfig{
+			Version: spdxutil.DefaultVersion,
+			Pretty:  opt,
+		})
+		require.NoError(t, err)
+
+		dir := t.TempDir()
+		s := testutil.DirectoryInput(t, dir)
+
+		var buffer bytes.Buffer
+		err = enc.Encode(&buffer, s)
+		require.NoError(t, err)
+
+		return strings.TrimSpace(buffer.String())
+	}
+
+	t.Run("pretty", func(t *testing.T) {
+		actual := run(true)
+		assert.Contains(t, actual, "\n")
+	})
+
+	t.Run("compact", func(t *testing.T) {
+		actual := run(false)
+		assert.NotContains(t, actual, "\n")
+	})
+}
+
+func TestEscapeHTML(t *testing.T) {
+	dir := t.TempDir()
+	s := testutil.DirectoryInput(t, dir)
+	s.Artifacts.Packages.Add(pkg.Package{
+		Name: "<html-package>",
+	})
+
+	// by default we do not escape HTML
+	t.Run("default", func(t *testing.T) {
+		cfg := DefaultEncoderConfig()
+
+		enc, err := NewFormatEncoderWithConfig(cfg)
+		require.NoError(t, err)
+
+		var buffer bytes.Buffer
+		err = enc.Encode(&buffer, s)
+		require.NoError(t, err)
+
+		actual := buffer.String()
+		assert.Contains(t, actual, "<html-package>")
+		assert.NotContains(t, actual, "\\u003chtml-package\\u003e")
+	})
+
+}
+
 func TestSPDXJSONDirectoryEncoder(t *testing.T) {
 	dir := t.TempDir()
 	testutil.AssertEncoderAgainstGoldenSnapshot(t,

syft/format/syftjson/encoder.go

@@ -14,6 +14,7 @@ const ID sbom.FormatID = "syft-json"
 
 type EncoderConfig struct {
 	Legacy bool // transform the output to the legacy syft-json format (pre v1.0 changes, enumerated in the README.md)
+	Pretty bool // don't include spaces and newlines; same as jq -c
 }
 
 type encoder struct {
@@ -37,6 +38,7 @@ func NewFormatEncoderWithConfig(cfg EncoderConfig) (sbom.FormatEncoder, error) {
 func DefaultEncoderConfig() EncoderConfig {
 	return EncoderConfig{
 		Legacy: false,
+		Pretty: false,
 	}
 }
 
@@ -59,9 +61,12 @@ func (e encoder) Encode(writer io.Writer, s sbom.SBOM) error {
 	doc := ToFormatModel(s, e.cfg)
 
 	enc := json.NewEncoder(writer)
-	// prevent > and < from being escaped in the payload
+
 	enc.SetEscapeHTML(false)
-	enc.SetIndent("", " ")
+
+	if e.cfg.Pretty {
+		enc.SetIndent("", " ")
+	}
 
 	return enc.Encode(&doc)
 }

syft/format/syftjson/encoder_test.go

@@ -1,9 +1,14 @@
 package syftjson
 
 import (
+	"bytes"
 	"flag"
+	"strings"
 	"testing"
 
+	"github.com/stretchr/testify/assert"
+	"github.com/stretchr/testify/require"
+
 	stereoFile "github.com/anchore/stereoscope/pkg/file"
 	"github.com/anchore/syft/internal"
 	"github.com/anchore/syft/syft/artifact"
@@ -31,12 +36,69 @@ func TestDefaultNameAndVersion(t *testing.T) {
 	}
 }
 
+func TestPrettyOutput(t *testing.T) {
+	run := func(opt bool) string {
+		enc, err := NewFormatEncoderWithConfig(EncoderConfig{
+			Pretty: opt,
+		})
+		require.NoError(t, err)
+
+		dir := t.TempDir()
+		s := testutil.DirectoryInput(t, dir)
+
+		var buffer bytes.Buffer
+		err = enc.Encode(&buffer, s)
+		require.NoError(t, err)
+
+		return strings.TrimSpace(buffer.String())
+	}
+
+	t.Run("pretty", func(t *testing.T) {
+		actual := run(true)
+		assert.Contains(t, actual, "\n")
+	})
+
+	t.Run("compact", func(t *testing.T) {
+		actual := run(false)
+		assert.NotContains(t, actual, "\n")
+	})
+}
+
+func TestEscapeHTML(t *testing.T) {
+	dir := t.TempDir()
+	s := testutil.DirectoryInput(t, dir)
+	s.Artifacts.Packages.Add(pkg.Package{
+		Name: "<html-package>",
+	})
+
+	// by default we do not escape HTML
+	t.Run("default", func(t *testing.T) {
+		cfg := DefaultEncoderConfig()
+
+		enc, err := NewFormatEncoderWithConfig(cfg)
+		require.NoError(t, err)
+
+		var buffer bytes.Buffer
+		err = enc.Encode(&buffer, s)
+		require.NoError(t, err)
+
+		actual := buffer.String()
+		assert.Contains(t, actual, "<html-package>")
+		assert.NotContains(t, actual, "\\u003chtml-package\\u003e")
+	})
+}
+
 func TestDirectoryEncoder(t *testing.T) {
+	cfg := DefaultEncoderConfig()
+	cfg.Pretty = true
+	enc, err := NewFormatEncoderWithConfig(cfg)
+	require.NoError(t, err)
+
 	dir := t.TempDir()
 	testutil.AssertEncoderAgainstGoldenSnapshot(t,
 		testutil.EncoderSnapshotTestConfig{
 			Subject:                     testutil.DirectoryInput(t, dir),
-			Format:                      NewFormatEncoder(),
+			Format:                      enc,
 			UpdateSnapshot:              *updateSnapshot,
 			PersistRedactionsInSnapshot: true,
 			IsJSON:                      true,
@@ -46,6 +108,11 @@ func TestDirectoryEncoder(t *testing.T) {
 }
 
 func TestImageEncoder(t *testing.T) {
+	cfg := DefaultEncoderConfig()
+	cfg.Pretty = true
+	enc, err := NewFormatEncoderWithConfig(cfg)
+	require.NoError(t, err)
+
 	testImage := "image-simple"
 	testutil.AssertEncoderAgainstGoldenImageSnapshot(t,
 		testutil.ImageSnapshotTestConfig{
@@ -54,7 +121,7 @@ func TestImageEncoder(t *testing.T) {
 		},
 		testutil.EncoderSnapshotTestConfig{
 			Subject:                     testutil.ImageInput(t, testImage, testutil.FromSnapshot()),
-			Format:                      NewFormatEncoder(),
+			Format:                      enc,
 			UpdateSnapshot:              *updateSnapshot,
 			PersistRedactionsInSnapshot: true,
 			IsJSON:                      true,