Split the sbom.Format interface by encode and decode use cases (#2186)

* split up sbom.Format into encode and decode ops

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

* update cmd pkg to inject format configs

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

* bump cyclonedx schema to 1.5

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

* redact image metadata from github encoder tests

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

* add more testing around format decoder identify

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

* add test case for format version options

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

* fix cli tests

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

* fix CLI test

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

* [wip] - review comments

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

* keep encoder creation out of post load function

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

* keep decider and identify functions

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

* add a few more doc comments

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

* remove format encoder default function helpers

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

* address PR feedback

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

* move back to streaming based decode functions

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

* with common convention for encoder constructors

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

* fix tests and allow for encoders to be created from cli options

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

* fix cli tests

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

* fix linting

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

* buffer reads from stdin to support seeking

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

---------

Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
This commit is contained in:
Alex Goodman
2023-10-25 13:43:06 +00:00
committed by GitHub
parent 7315f83f9d
commit 7392d607b6
254 changed files with 10450 additions and 2714 deletions
+84
View File
@@ -0,0 +1,84 @@
package internal
import (
"bytes"
"errors"
"io"
"github.com/anchore/syft/internal/log"
)
var _ io.ReadSeekCloser = (*bufferedSeekReader)(nil)
// bufferedSeekReader wraps an io.ReadCloser to provide io.Seeker functionality.
// It only supports seeking from the start and cannot seek past what has already been read.
type bufferedSeekReader struct {
r io.ReadCloser
buf *bytes.Reader
data []byte
pos int64
closed bool
}
func NewBufferedSeeker(rc io.ReadCloser) io.ReadSeekCloser {
return &bufferedSeekReader{
r: rc,
}
}
func (bs *bufferedSeekReader) Read(p []byte) (int, error) {
if bs.closed {
return 0, errors.New("cannot read from closed reader")
}
if bs.pos == int64(len(bs.data)) {
// if we're at the end of our buffer, read more data into it
tmp := make([]byte, len(p))
n, err := bs.r.Read(tmp)
if err != nil && err != io.EOF {
return 0, err
} else if err == io.EOF {
bs.closed = true
}
bs.data = append(bs.data, tmp[:n]...)
bs.buf = bytes.NewReader(bs.data)
}
n, err := bs.buf.ReadAt(p, bs.pos)
if err != nil && err != io.EOF {
log.WithFields("error", err).Trace("buffered seek reader failed to read from underlying reader")
}
bs.pos += int64(n)
return n, nil
}
func (bs *bufferedSeekReader) Seek(offset int64, whence int) (int64, error) {
var abs int64
switch whence {
case io.SeekStart:
abs = offset
case io.SeekCurrent:
abs = bs.pos + offset
case io.SeekEnd:
return 0, errors.New("'SeekEnd' not supported")
default:
return 0, errors.New("invalid seek option")
}
if abs < 0 {
return 0, errors.New("unable to seek before start")
}
if abs > int64(len(bs.data)) {
return 0, errors.New("unable to seek past read data")
}
bs.pos = abs
return bs.pos, nil
}
func (bs *bufferedSeekReader) Close() error {
bs.closed = true
return bs.r.Close()
}
+141
View File
@@ -0,0 +1,141 @@
package internal
import (
"bytes"
"io"
"testing"
"github.com/stretchr/testify/assert"
"github.com/stretchr/testify/require"
)
func TestBufferedSeeker_Read(t *testing.T) {
tests := []struct {
name string
initialData string
readLengths []int
expectedReads []string
expectError bool
}{
{
name: "go case (read)",
initialData: "Hello, World!",
readLengths: []int{5},
expectedReads: []string{"Hello"},
},
{
name: "multiple reads",
initialData: "Hello, World!",
readLengths: []int{5, 8},
expectedReads: []string{"Hello", ", World!"},
},
}
for _, tt := range tests {
t.Run(tt.name, func(t *testing.T) {
bs := NewBufferedSeeker(io.NopCloser(bytes.NewBufferString(tt.initialData)))
for i, length := range tt.readLengths {
buf := make([]byte, length)
n, err := bs.Read(buf)
if !tt.expectError {
assert.NoError(t, err)
assert.Equalf(t, tt.expectedReads[i], string(buf[:n]), "read index %d", i)
} else {
assert.Error(t, err)
}
}
})
}
}
func TestBufferedSeeker_Seek(t *testing.T) {
tests := []struct {
name string
initialData string
readLengths []int
seekOffsets []int64
seekWhence []int
expectedReads []string
seekError require.ErrorAssertionFunc
readError require.ErrorAssertionFunc
}{
{
name: "seek start 0 without read first",
initialData: "Hello, World!",
readLengths: []int{5},
seekOffsets: []int64{0},
seekWhence: []int{io.SeekStart},
expectedReads: []string{"Hello"},
},
{
name: "read + seek back",
initialData: "Hello, World!",
readLengths: []int{5, 8, 8},
seekOffsets: []int64{-1, -1, 2},
seekWhence: []int{io.SeekStart, io.SeekStart, io.SeekStart},
expectedReads: []string{"Hello", ", World!", "llo, Wor"},
},
{
name: "seek past read data",
initialData: "Hello, World!",
readLengths: []int{5},
seekOffsets: []int64{20},
seekWhence: []int{io.SeekStart},
expectedReads: []string{""},
seekError: require.Error,
},
{
name: "seek to end",
initialData: "Hello, World!",
readLengths: []int{-1},
seekOffsets: []int64{20},
seekWhence: []int{io.SeekEnd},
expectedReads: []string{""},
seekError: require.Error,
},
}
for _, tt := range tests {
t.Run(tt.name, func(t *testing.T) {
if tt.seekError == nil {
tt.seekError = require.NoError
}
if tt.readError == nil {
tt.readError = require.NoError
}
bs := NewBufferedSeeker(io.NopCloser(bytes.NewBufferString(tt.initialData)))
for i, length := range tt.readLengths {
if len(tt.seekOffsets) > i && tt.seekOffsets[i] >= 0 {
_, err := bs.Seek(tt.seekOffsets[i], tt.seekWhence[i])
tt.seekError(t, err)
if err != nil {
continue
}
}
if length >= 0 {
buf := make([]byte, length)
n, err := bs.Read(buf)
tt.readError(t, err)
if err != nil {
continue
}
assert.Equalf(t, tt.expectedReads[i], string(buf[:n]), "read index %d", i)
}
}
})
}
}
func TestBufferedSeeker_Close(t *testing.T) {
bs := NewBufferedSeeker(io.NopCloser(bytes.NewBufferString("Hello, World!")))
err := bs.Close()
assert.NoError(t, err)
n, err := bs.Read(make([]byte, 1))
assert.Equal(t, 0, n)
assert.Error(t, err)
}