Add query by MIME type to source.FileResolver (#529)

* add query by MIME type to source.FileResolver

Signed-off-by: Alex Goodman <alex.goodman@anchore.com>

* pull in stereoscope MIME type feature

Signed-off-by: Alex Goodman <alex.goodman@anchore.com>

* fix tests

Signed-off-by: Alex Goodman <alex.goodman@anchore.com>

go.mod

@@ -11,7 +11,7 @@ require (
 	github.com/anchore/go-testutils v0.0.0-20200925183923-d5f45b0d3c04
 	github.com/anchore/go-version v1.2.2-0.20200701162849-18adb9c92b9b
 	github.com/anchore/packageurl-go v0.0.0-20210922164639-b3fa992ebd29
-	github.com/anchore/stereoscope v0.0.0-20210817160504-0f4abc2a5a5a
+	github.com/anchore/stereoscope v0.0.0-20211005213828-538011008578
 	github.com/antihax/optional v1.0.0
 	github.com/bmatcuk/doublestar/v2 v2.0.4
 	github.com/docker/docker v17.12.0-ce-rc1.0.20200309214505-aa6a9891b09c+incompatible

go.sum

@@ -117,8 +117,8 @@ github.com/anchore/go-version v1.2.2-0.20200701162849-18adb9c92b9b h1:e1bmaoJfZV
 github.com/anchore/go-version v1.2.2-0.20200701162849-18adb9c92b9b/go.mod h1:Bkc+JYWjMCF8OyZ340IMSIi2Ebf3uwByOk6ho4wne1E=
 github.com/anchore/packageurl-go v0.0.0-20210922164639-b3fa992ebd29 h1:K9LfnxwhqvihqU0+MF325FNy7fsKV9EGaUxdfR4gnWk=
 github.com/anchore/packageurl-go v0.0.0-20210922164639-b3fa992ebd29/go.mod h1:Oc1UkGaJwY6ND6vtAqPSlYrptKRJngHwkwB6W7l1uP0=
-github.com/anchore/stereoscope v0.0.0-20210817160504-0f4abc2a5a5a h1:RQb+Gft1MKxjDfJCnHP/f1mwfy0Jz50Kp9QGgSWKQiY=
+github.com/anchore/stereoscope v0.0.0-20211005213828-538011008578 h1:gSpftl0RWfdTwlPmsOLgEawHIw16xwwG1mFD/wrVDRE=
-github.com/anchore/stereoscope v0.0.0-20210817160504-0f4abc2a5a5a/go.mod h1:165DfE5jApgEkHTWwu7Bijeml9fofudrgcpuWaD9+tk=
+github.com/anchore/stereoscope v0.0.0-20211005213828-538011008578/go.mod h1:kL7jfbXblrDcBhu5ja/s+VTYL3Mzof4eQNMJiSqcwXQ=
 github.com/andreyvit/diff v0.0.0-20170406064948-c7f18ee00883/go.mod h1:rCTlJbsFo29Kk6CurOXKm700vrz8f0KW0JNfpkRJY/8=
 github.com/antihax/optional v1.0.0 h1:xK2lYat7ZLaVVcIuj82J8kIro4V6kDe0AUDFboUCwcg=
 github.com/antihax/optional v1.0.0/go.mod h1:uupD/76wgC+ih3iEmQUL+0Ugr19nfwCT1kdvxnR2qWY=
@@ -242,6 +242,8 @@ github.com/fortytw2/leaktest v1.3.0/go.mod h1:jDsjWgpAGjm2CA7WthBh/CdZYEPF31XHqu
 github.com/fsnotify/fsnotify v1.4.7/go.mod h1:jwhsz4b93w/PPRr/qN1Yymfu8t87LnFCMoQvtojpjFo=
 github.com/fsnotify/fsnotify v1.4.9 h1:hsms1Qyu0jgnwNXIxa+/V/PDsU6CfLf6CNO8H7IWoS4=
 github.com/fsnotify/fsnotify v1.4.9/go.mod h1:znqG4EE+3YCdAaPaxE2ZRY/06pZUdp0tY4IgpuI1SZQ=
+github.com/gabriel-vasile/mimetype v1.3.0 h1:4YOHITFLyYwF+iqG0ybSLGArRItynpfwdlWRmJnd75E=
+github.com/gabriel-vasile/mimetype v1.3.0/go.mod h1:fA8fi6KUiG7MgQQ+mEWotXoEOvmxRtOJlERCzSmRvr8=
 github.com/ghodss/yaml v0.0.0-20150909031657-73d445a93680/go.mod h1:4dBDuWmgqj2HViK6kFavaiC9ZROes6MMH2rRYeMEF04=
 github.com/ghodss/yaml v1.0.0/go.mod h1:4dBDuWmgqj2HViK6kFavaiC9ZROes6MMH2rRYeMEF04=
 github.com/go-critic/go-critic v0.4.1/go.mod h1:7/14rZGnZbY6E38VEGk2kVhoq6itzc1E68facVDK23g=
@@ -872,7 +874,7 @@ golang.org/x/net v0.0.0-20200324143707-d3edc9973b7e/go.mod h1:qpuaurCH72eLCgpAm/
 golang.org/x/net v0.0.0-20200501053045-e0ff5e5a1de5/go.mod h1:qpuaurCH72eLCgpAm/N6yyVIVM9cpaDIP3A8BGJEC5A=
 golang.org/x/net v0.0.0-20200506145744-7e3656a0809f/go.mod h1:qpuaurCH72eLCgpAm/N6yyVIVM9cpaDIP3A8BGJEC5A=
 golang.org/x/net v0.0.0-20200520182314-0ba52f642ac2/go.mod h1:qpuaurCH72eLCgpAm/N6yyVIVM9cpaDIP3A8BGJEC5A=
-golang.org/x/net v0.0.0-20200602114024-627f9648deb9/go.mod h1:qpuaurCH72eLCgpAm/N6yyVIVM9cpaDIP3A8BGJEC5A=
+golang.org/x/net v0.0.0-20210505024714-0287a6fb4125/go.mod h1:9nx3DQGgdP8bBQD5qxJ1jj9UTztislL4KSBs9R2vV5Y=
 golang.org/x/net v0.0.0-20210813160813-60bc85c4be6d h1:LO7XpTYMwTqxjLcGWPijK3vRXg1aWdlNOVOHRq45d7c=
 golang.org/x/net v0.0.0-20210813160813-60bc85c4be6d/go.mod h1:9nx3DQGgdP8bBQD5qxJ1jj9UTztislL4KSBs9R2vV5Y=
 golang.org/x/oauth2 v0.0.0-20180821212333-d2e6202438be/go.mod h1:N/0e6XlmueqKjAGxoOufVs8QHGRruUQn6yWY3a++T0U=
@@ -939,7 +941,6 @@ golang.org/x/sys v0.0.0-20200331124033-c3d80250170d/go.mod h1:h1NjWce9XRLGQEsW7w
 golang.org/x/sys v0.0.0-20200501052902-10377860bb8e/go.mod h1:h1NjWce9XRLGQEsW7wpKNCjG9DtNlClVuFLEZdDNbEs=
 golang.org/x/sys v0.0.0-20200511232937-7e40ca221e25/go.mod h1:h1NjWce9XRLGQEsW7wpKNCjG9DtNlClVuFLEZdDNbEs=
 golang.org/x/sys v0.0.0-20200523222454-059865788121/go.mod h1:h1NjWce9XRLGQEsW7wpKNCjG9DtNlClVuFLEZdDNbEs=
-golang.org/x/sys v0.0.0-20200602225109-6fdc65e7d980/go.mod h1:h1NjWce9XRLGQEsW7wpKNCjG9DtNlClVuFLEZdDNbEs=
 golang.org/x/sys v0.0.0-20201119102817-f84b799fce68/go.mod h1:h1NjWce9XRLGQEsW7wpKNCjG9DtNlClVuFLEZdDNbEs=
 golang.org/x/sys v0.0.0-20210423082822-04245dca01da h1:b3NXsE2LusjYGGjL5bxEVZZORm/YEFFrWFjR8eFrw/c=
 golang.org/x/sys v0.0.0-20210423082822-04245dca01da/go.mod h1:h1NjWce9XRLGQEsW7wpKNCjG9DtNlClVuFLEZdDNbEs=

internal/presenter/poweruser/json_file_metadata.go

@@ -22,6 +22,7 @@ type JSONFileMetadataEntry struct {
 	UserID          int             `json:"userID"`
 	GroupID         int             `json:"groupID"`
 	Digests         []file.Digest   `json:"digests,omitempty"`
+	MIMEType        string          `json:"mimeType"`
 }
 
 func NewJSONFileMetadata(data map[source.Location]source.FileMetadata, digests map[source.Location][]file.Digest) ([]JSONFileMetadata, error) {
@@ -46,6 +47,7 @@ func NewJSONFileMetadata(data map[source.Location]source.FileMetadata, digests m
 				UserID:          metadata.UserID,
 				GroupID:         metadata.GroupID,
 				Digests:         digestResults,
+				MIMEType:        metadata.MIMEType,
 			},
 		})
 	}

internal/presenter/poweruser/test-fixtures/snapshot/TestJSONPresenter.golden

@@ -16,7 +16,8 @@
     "mode": 775,
     "type": "directory",
     "userID": 0,
-    "groupID": 0
+    "groupID": 0,
+    "mimeType": ""
    }
   },
   {
@@ -33,7 +34,8 @@
       "algorithm": "sha256",
       "value": "366a3f5653e34673b875891b021647440d0127c2ef041e3b1a22da2a7d4f3703"
      }
-    ]
+    ],
+    "mimeType": ""
    }
   },
   {
@@ -45,7 +47,8 @@
     "type": "symbolicLink",
     "linkDestination": "/c",
     "userID": 0,
-    "groupID": 0
+    "groupID": 0,
+    "mimeType": ""
    }
   },
   {
@@ -62,7 +65,8 @@
       "algorithm": "sha256",
       "value": "1b3722da2a7d90d033b87581a2a3f12021647445653e34666ef041e3b4f3707c"
      }
-    ]
+    ],
+    "mimeType": ""
    }
   }
  ],

syft/file/metadata_cataloger_test.go

@@ -49,11 +49,12 @@ func TestFileMetadataCataloger(t *testing.T) {
 			path:   "/file-1.txt",
 			exists: true,
 			expected: source.FileMetadata{
-				Mode:    0644,
+				Mode:     0644,
-				Type:    "RegularFile",
+				Type:     "RegularFile",
-				UserID:  1,
+				UserID:   1,
-				GroupID: 2,
+				GroupID:  2,
-				Size:    7,
+				Size:     7,
+				MIMEType: "text/plain",
 			},
 		},
 		{
@@ -65,6 +66,7 @@ func TestFileMetadataCataloger(t *testing.T) {
 				LinkDestination: "file-1.txt",
 				UserID:          1,
 				GroupID:         2,
+				MIMEType:        "",
 			},
 		},
 		{
@@ -76,46 +78,51 @@ func TestFileMetadataCataloger(t *testing.T) {
 				LinkDestination: "file-1.txt",
 				UserID:          0,
 				GroupID:         0,
+				MIMEType:        "",
 			},
 		},
 		{
 			path:   "/char-device-1",
 			exists: true,
 			expected: source.FileMetadata{
-				Mode:    0644 | os.ModeDevice | os.ModeCharDevice,
+				Mode:     0644 | os.ModeDevice | os.ModeCharDevice,
-				Type:    "CharacterDevice",
+				Type:     "CharacterDevice",
-				UserID:  0,
+				UserID:   0,
-				GroupID: 0,
+				GroupID:  0,
+				MIMEType: "",
 			},
 		},
 		{
 			path:   "/block-device-1",
 			exists: true,
 			expected: source.FileMetadata{
-				Mode:    0644 | os.ModeDevice,
+				Mode:     0644 | os.ModeDevice,
-				Type:    "BlockDevice",
+				Type:     "BlockDevice",
-				UserID:  0,
+				UserID:   0,
-				GroupID: 0,
+				GroupID:  0,
+				MIMEType: "",
 			},
 		},
 		{
 			path:   "/fifo-1",
 			exists: true,
 			expected: source.FileMetadata{
-				Mode:    0644 | os.ModeNamedPipe,
+				Mode:     0644 | os.ModeNamedPipe,
-				Type:    "FIFONode",
+				Type:     "FIFONode",
-				UserID:  0,
+				UserID:   0,
-				GroupID: 0,
+				GroupID:  0,
+				MIMEType: "",
 			},
 		},
 		{
 			path:   "/bin",
 			exists: true,
 			expected: source.FileMetadata{
-				Mode:    0755 | os.ModeDir,
+				Mode:     0755 | os.ModeDir,
-				Type:    "Directory",
+				Type:     "Directory",
-				UserID:  0,
+				UserID:   0,
-				GroupID: 0,
+				GroupID:  0,
+				MIMEType: "",
 			},
 		},
 	}

syft/pkg/cataloger/rpmdb/parse_rpmdb_test.go

@@ -45,11 +45,16 @@ func (r *rpmdbTestFileResolverMock) FilesByPath(paths ...string) ([]source.Locat
 func (r *rpmdbTestFileResolverMock) FilesByGlob(...string) ([]source.Location, error) {
 	return nil, fmt.Errorf("not implemented")
 }
+
 func (r *rpmdbTestFileResolverMock) RelativeFileByPath(source.Location, string) *source.Location {
 	panic(fmt.Errorf("not implemented"))
 	return nil
 }
 
+func (r rpmdbTestFileResolverMock) FilesByMIMEType(types ...string) ([]source.Location, error) {
+	panic("implement me")
+}
+
 func TestParseRpmDB(t *testing.T) {
 	dbLocation := source.NewLocation("test-path")
 

syft/source/all_layers_resolver.go

@@ -188,6 +188,24 @@ func (r *allLayersResolver) FileContentsByLocation(location Location) (io.ReadCl
 	return r.img.FileContentsByRef(location.ref)
 }
 
+func (r *allLayersResolver) FilesByMIMEType(types ...string) ([]Location, error) {
+	var locations []Location
+	for _, layerIdx := range r.layers {
+		layer := r.img.Layers[layerIdx]
+
+		refs, err := layer.FilesByMIMEType(types...)
+		if err != nil {
+			return nil, err
+		}
+
+		for _, ref := range refs {
+			locations = append(locations, NewLocationFromReference(ref))
+		}
+	}
+
+	return locations, nil
+}
+
 func (r *allLayersResolver) AllLocations() <-chan Location {
 	results := make(chan Location)
 	go func() {

syft/source/all_layers_resolver_test.go

@@ -3,6 +3,8 @@ package source
 import (
 	"testing"
 
+	"github.com/stretchr/testify/assert"
+
 	"github.com/anchore/stereoscope/pkg/imagetest"
 )
 
@@ -239,3 +241,34 @@ func TestAllLayersResolver_FilesByGlob(t *testing.T) {
 		})
 	}
 }
+
+func Test_imageAllLayersResolver_FilesByMIMEType(t *testing.T) {
+
+	tests := []struct {
+		fixtureName   string
+		mimeType      string
+		expectedPaths []string
+	}{
+		{
+			fixtureName:   "image-duplicate-path",
+			mimeType:      "text/plain",
+			expectedPaths: []string{"/somefile-1.txt", "/somefile-1.txt"},
+		},
+	}
+	for _, test := range tests {
+		t.Run(test.fixtureName, func(t *testing.T) {
+			img := imagetest.GetFixtureImage(t, "docker-archive", test.fixtureName)
+
+			resolver, err := newAllLayersResolver(img)
+			assert.NoError(t, err)
+
+			locations, err := resolver.FilesByMIMEType(test.mimeType)
+			assert.NoError(t, err)
+
+			assert.Len(t, test.expectedPaths, len(locations))
+			for idx, l := range locations {
+				assert.Equal(t, test.expectedPaths[idx], l.RealPath, "does not have path %q", l.RealPath)
+			}
+		})
+	}
+}

syft/source/directory_resolver.go

@@ -8,7 +8,6 @@ import (
 	"path"
 	"path/filepath"
 	"strings"
-	"syscall"
 
 	"github.com/anchore/stereoscope/pkg/file"
 	"github.com/anchore/stereoscope/pkg/filetree"
@@ -35,10 +34,11 @@ type directoryResolver struct {
 	path     string
 	cwd      string
 	fileTree *filetree.FileTree
-	infos    map[file.ID]os.FileInfo
+	metadata map[file.ID]FileMetadata
 	// TODO: wire up to report these paths in the json report
-	pathFilterFns []pathFilterFn
+	pathFilterFns  []pathFilterFn
-	errPaths      map[string]error
+	refsByMIMEType map[string][]file.Reference
+	errPaths       map[string]error
 }
 
 func newDirectoryResolver(root string, pathFilters ...pathFilterFn) (*directoryResolver, error) {
@@ -52,12 +52,13 @@ func newDirectoryResolver(root string, pathFilters ...pathFilterFn) (*directoryR
 	}
 
 	resolver := directoryResolver{
-		path:          root,
+		path:           root,
-		cwd:           cwd,
+		cwd:            cwd,
-		fileTree:      filetree.NewFileTree(),
+		fileTree:       filetree.NewFileTree(),
-		infos:         make(map[file.ID]os.FileInfo),
+		metadata:       make(map[file.ID]FileMetadata),
-		pathFilterFns: pathFilters,
+		pathFilterFns:  pathFilters,
-		errPaths:      make(map[string]error),
+		refsByMIMEType: make(map[string][]file.Reference),
+		errPaths:       make(map[string]error),
 	}
 
 	return &resolver, indexAllRoots(root, resolver.indexTree)
@@ -158,7 +159,11 @@ func (r directoryResolver) addPathToIndex(p string, info os.FileInfo) (string, e
 		}
 	}
 
-	r.infos[ref.ID()] = info
+	metadata := fileMetadataFromPath(p, info)
+	if ref != nil {
+		r.refsByMIMEType[metadata.MIMEType] = append(r.refsByMIMEType[metadata.MIMEType], *ref)
+	}
+	r.metadata[ref.ID()] = metadata
 	return newRoot, nil
 }
 
@@ -207,6 +212,7 @@ func (r directoryResolver) FilesByPath(userPaths ...string) ([]Location, error)
 			log.Warnf("unable to get file by path=%q : %+v", userPath, err)
 			continue
 		}
+		// TODO: why not use stored metadata?
 		fileMeta, err := os.Stat(userStrPath)
 		if os.IsNotExist(err) {
 			continue
@@ -280,25 +286,24 @@ func (r *directoryResolver) AllLocations() <-chan Location {
 }
 
 func (r *directoryResolver) FileMetadataByLocation(location Location) (FileMetadata, error) {
-	info, exists := r.infos[location.ref.ID()]
+	metadata, exists := r.metadata[location.ref.ID()]
 	if !exists {
 		return FileMetadata{}, fmt.Errorf("location: %+v : %w", location, os.ErrNotExist)
 	}
 
-	uid := -1
+	return metadata, nil
-	gid := -1
+}
-	if stat, ok := info.Sys().(*syscall.Stat_t); ok {
-		uid = int(stat.Uid)
-		gid = int(stat.Gid)
-	}
 
-	return FileMetadata{
+func (r *directoryResolver) FilesByMIMEType(types ...string) ([]Location, error) {
-		Mode: info.Mode(),
+	var locations []Location
-		Type: newFileTypeFromMode(info.Mode()),
+	for _, ty := range types {
-		// unsupported across platforms
+		if refs, ok := r.refsByMIMEType[ty]; ok {
-		UserID:  uid,
+			for _, ref := range refs {
-		GroupID: gid,
+				locations = append(locations, NewLocationFromDirectory(r.responsePath(string(ref.RealPath)), ref))
-	}, nil
+			}
+		}
+	}
+	return locations, nil
 }
 
 func isUnixSystemRuntimePath(path string) bool {

syft/source/directory_resolver_test.go

@@ -7,6 +7,8 @@ import (
 	"strings"
 	"testing"
 
+	"github.com/scylladb/go-set/strset"
+
 	"github.com/anchore/stereoscope/pkg/file"
 	"github.com/stretchr/testify/assert"
 	"github.com/wagoodman/go-progress"
@@ -303,7 +305,7 @@ func Test_directoryResolver_index(t *testing.T) {
 			if assert.NoError(t, err) {
 				return
 			}
-			assert.Equal(t, info, r.infos[ref.ID()])
+			assert.Equal(t, info, r.metadata[ref.ID()])
 		})
 	}
 }
@@ -429,3 +431,31 @@ func Test_indexAllRoots(t *testing.T) {
 		})
 	}
 }
+
+func Test_directoryResolver_FilesByMIMEType(t *testing.T) {
+
+	tests := []struct {
+		fixturePath   string
+		mimeType      string
+		expectedPaths *strset.Set
+	}{
+		{
+			fixturePath:   "./test-fixtures/image-simple",
+			mimeType:      "text/plain",
+			expectedPaths: strset.New("test-fixtures/image-simple/file-1.txt", "test-fixtures/image-simple/file-2.txt", "test-fixtures/image-simple/target/really/nested/file-3.txt", "test-fixtures/image-simple/Dockerfile"),
+		},
+	}
+	for _, test := range tests {
+		t.Run(test.fixturePath, func(t *testing.T) {
+
+			resolver, err := newDirectoryResolver(test.fixturePath)
+			assert.NoError(t, err)
+			locations, err := resolver.FilesByMIMEType(test.mimeType)
+			assert.NoError(t, err)
+			assert.Equal(t, test.expectedPaths.Size(), len(locations))
+			for _, l := range locations {
+				assert.True(t, test.expectedPaths.Has(l.RealPath), "does not have path %q", l.RealPath)
+			}
+		})
+	}
+}

syft/source/file_metadata.go

@@ -2,6 +2,11 @@ package source
 
 import (
 	"os"
+	"syscall"
+
+	"github.com/anchore/syft/internal/log"
+
+	"github.com/anchore/stereoscope/pkg/file"
 
 	"github.com/anchore/stereoscope/pkg/image"
 )
@@ -13,6 +18,7 @@ type FileMetadata struct {
 	GroupID         int
 	LinkDestination string
 	Size            int64
+	MIMEType        string
 }
 
 func fileMetadataByLocation(img *image.Image, location Location) (FileMetadata, error) {
@@ -28,5 +34,36 @@ func fileMetadataByLocation(img *image.Image, location Location) (FileMetadata,
 		GroupID:         entry.Metadata.GroupID,
 		LinkDestination: entry.Metadata.Linkname,
 		Size:            entry.Metadata.Size,
+		MIMEType:        entry.Metadata.MIMEType,
 	}, nil
 }
+
+func fileMetadataFromPath(path string, info os.FileInfo) FileMetadata {
+	uid := -1
+	gid := -1
+	if stat, ok := info.Sys().(*syscall.Stat_t); ok {
+		uid = int(stat.Uid)
+		gid = int(stat.Gid)
+	}
+
+	f, err := os.Open(path)
+	if err != nil {
+		// TODO: it may be that the file is inaccessible, however, this is not an error or a warning. In the future we need to track these as known-unknowns
+		f = nil
+	} else {
+		defer func() {
+			if err := f.Close(); err != nil {
+				log.Warnf("unable to close file while obtaining metadata: %s", path)
+			}
+		}()
+	}
+
+	return FileMetadata{
+		Mode: info.Mode(),
+		Type: newFileTypeFromMode(info.Mode()),
+		// unsupported across platforms
+		UserID:   uid,
+		GroupID:  gid,
+		MIMEType: file.MIMEType(f),
+	}
+}

syft/source/file_resolver.go

@@ -29,6 +29,8 @@ type FilePathResolver interface {
 	FilesByPath(paths ...string) ([]Location, error)
 	// FilesByGlob fetches a set of file references which the given glob matches
 	FilesByGlob(patterns ...string) ([]Location, error)
+	// FilesByGlob fetches a set of file references which the contents have been classified as one of the given MIME Types
+	FilesByMIMEType(types ...string) ([]Location, error)
 	// RelativeFileByPath fetches a single file at the given path relative to the layer squash of the given reference.
 	// This is helpful when attempting to find a file that is in the same layer or lower as another file.
 	RelativeFileByPath(_ Location, path string) *Location

syft/source/image_squash_resolver.go

@@ -151,6 +151,20 @@ func (r *imageSquashResolver) AllLocations() <-chan Location {
 	return results
 }
 
+func (r *imageSquashResolver) FilesByMIMEType(types ...string) ([]Location, error) {
+	refs, err := r.img.FilesByMIMETypeFromSquash(types...)
+	if err != nil {
+		return nil, err
+	}
+
+	var locations []Location
+	for _, ref := range refs {
+		locations = append(locations, NewLocationFromReference(ref))
+	}
+
+	return locations, nil
+}
+
 func (r *imageSquashResolver) FileMetadataByLocation(location Location) (FileMetadata, error) {
 	return fileMetadataByLocation(r.img, location)
 }

syft/source/image_squash_resolver_test.go

@@ -3,6 +3,9 @@ package source
 import (
 	"testing"
 
+	"github.com/scylladb/go-set/strset"
+	"github.com/stretchr/testify/assert"
+
 	"github.com/anchore/stereoscope/pkg/imagetest"
 )
 
@@ -225,3 +228,34 @@ func TestImageSquashResolver_FilesByGlob(t *testing.T) {
 		})
 	}
 }
+
+func Test_imageSquashResolver_FilesByMIMEType(t *testing.T) {
+
+	tests := []struct {
+		fixtureName   string
+		mimeType      string
+		expectedPaths *strset.Set
+	}{
+		{
+			fixtureName:   "image-simple",
+			mimeType:      "text/plain",
+			expectedPaths: strset.New("/somefile-1.txt", "/somefile-2.txt", "/really/nested/file-3.txt"),
+		},
+	}
+	for _, test := range tests {
+		t.Run(test.fixtureName, func(t *testing.T) {
+			img := imagetest.GetFixtureImage(t, "docker-archive", test.fixtureName)
+
+			resolver, err := newImageSquashResolver(img)
+			assert.NoError(t, err)
+
+			locations, err := resolver.FilesByMIMEType(test.mimeType)
+			assert.NoError(t, err)
+
+			assert.Equal(t, test.expectedPaths.Size(), len(locations))
+			for _, l := range locations {
+				assert.True(t, test.expectedPaths.Has(l.RealPath), "does not have path %q", l.RealPath)
+			}
+		})
+	}
+}

syft/source/mock_resolver.go

@@ -14,8 +14,9 @@ var _ FileResolver = (*MockResolver)(nil)
 // It provides an implementation that can resolve local filesystem paths using only a provided discrete list of file
 // paths, which are typically paths to test fixtures.
 type MockResolver struct {
-	Locations []Location
+	locations     []Location
-	Metadata  map[Location]FileMetadata
+	metadata      map[Location]FileMetadata
+	mimeTypeIndex map[string][]Location
 }
 
 // NewMockResolverForPaths creates a new MockResolver, where the only resolvable
@@ -26,21 +27,30 @@ func NewMockResolverForPaths(paths ...string) *MockResolver {
 		locations = append(locations, NewLocation(p))
 	}
 
-	return &MockResolver{Locations: locations}
+	return &MockResolver{
+		locations: locations,
+		metadata:  make(map[Location]FileMetadata),
+	}
 }
 
 func NewMockResolverForPathsWithMetadata(metadata map[Location]FileMetadata) *MockResolver {
 	var locations []Location
-	for p := range metadata {
+	var mimeTypeIndex = make(map[string][]Location)
-		locations = append(locations, p)
+	for l, m := range metadata {
+		locations = append(locations, l)
+		mimeTypeIndex[m.MIMEType] = append(mimeTypeIndex[m.MIMEType], l)
 	}
 
-	return &MockResolver{Locations: locations, Metadata: metadata}
+	return &MockResolver{
+		locations:     locations,
+		metadata:      metadata,
+		mimeTypeIndex: mimeTypeIndex,
+	}
 }
 
 // HasPath indicates if the given path exists in the underlying source.
 func (r MockResolver) HasPath(path string) bool {
-	for _, l := range r.Locations {
+	for _, l := range r.locations {
 		if l.RealPath == path {
 			return true
 		}
@@ -50,13 +60,13 @@ func (r MockResolver) HasPath(path string) bool {
 
 // String returns the string representation of the MockResolver.
 func (r MockResolver) String() string {
-	return fmt.Sprintf("mock:(%s,...)", r.Locations[0].RealPath)
+	return fmt.Sprintf("mock:(%s,...)", r.locations[0].RealPath)
 }
 
 // FileContentsByLocation fetches file contents for a single location. If the
 // path does not exist, an error is returned.
 func (r MockResolver) FileContentsByLocation(location Location) (io.ReadCloser, error) {
-	for _, l := range r.Locations {
+	for _, l := range r.locations {
 		if l == location {
 			return os.Open(location.RealPath)
 		}
@@ -69,7 +79,7 @@ func (r MockResolver) FileContentsByLocation(location Location) (io.ReadCloser,
 func (r MockResolver) FilesByPath(paths ...string) ([]Location, error) {
 	var results []Location
 	for _, p := range paths {
-		for _, location := range r.Locations {
+		for _, location := range r.locations {
 			if p == location.RealPath {
 				results = append(results, NewLocation(p))
 			}
@@ -83,7 +93,7 @@ func (r MockResolver) FilesByPath(paths ...string) ([]Location, error) {
 func (r MockResolver) FilesByGlob(patterns ...string) ([]Location, error) {
 	var results []Location
 	for _, pattern := range patterns {
-		for _, location := range r.Locations {
+		for _, location := range r.locations {
 			matches, err := doublestar.Match(pattern, location.RealPath)
 			if err != nil {
 				return nil, err
@@ -115,7 +125,7 @@ func (r MockResolver) AllLocations() <-chan Location {
 	results := make(chan Location)
 	go func() {
 		defer close(results)
-		for _, l := range r.Locations {
+		for _, l := range r.locations {
 			results <- l
 		}
 	}()
@@ -142,3 +152,11 @@ func (r MockResolver) FileMetadataByLocation(l Location) (FileMetadata, error) {
 		Size:    info.Size(),
 	}, nil
 }
+
+func (r MockResolver) FilesByMIMEType(types ...string) ([]Location, error) {
+	var locations []Location
+	for _, ty := range types {
+		locations = append(r.mimeTypeIndex[ty], locations...)
+	}
+	return locations, nil
+}

syft/source/test-fixtures/image-duplicate-path/Dockerfile

@@ -0,0 +1,4 @@
+# Note: changes to this file will result in updating several test values. Consider making a new image fixture instead of editing this one.
+FROM scratch
+ADD file-1.txt /somefile-1.txt
+ADD file-2.txt /somefile-1.txt

syft/source/test-fixtures/image-duplicate-path/file-1.txt

@@ -0,0 +1 @@
+this file has contents

syft/source/test-fixtures/image-duplicate-path/file-2.txt

@@ -0,0 +1 @@
+file-2 contents!