Merge remote-tracking branch 'upstream/main' into add-nested-archive-cataloging

This commit is contained in:
Keith Zantow
2026-10-08 09:29:45 -04:00
19 changed files with 358 additions and 62 deletions
+1 -1
View File
@@ -24,7 +24,7 @@ tools:
# used to parse JSON/YAML annotations on the fixture cache image
- name: yq
version:
want: v4.53.6
want: v4.54.1
method: github-release
with:
repo: mikefarah/yq
+25
View File
@@ -0,0 +1,25 @@
package cli
import (
"io"
"testing"
"github.com/stretchr/testify/require"
"github.com/anchore/clio"
"github.com/anchore/syft/internal/redact"
)
func TestCommandCanRunMultipleTimes(t *testing.T) {
// https://github.com/anchore/syft/issues/2285
t.Cleanup(redact.Reset)
for i := 0; i < 2; i++ {
_, cmd := create(clio.Identification{Name: "syft"}, io.Discard)
cmd.SetArgs([]string{"cataloger", "list", "-o", "json"})
require.NoError(t, cmd.Execute())
// the store is released at the end of each run so the next one can set its own
require.Nil(t, redact.Get())
}
}
+9 -1
View File
@@ -41,6 +41,8 @@ func AppClioSetupConfig(id clio.Identification, out io.Writer) *clio.SetupConfig
stereoscope.SetBus(state.Bus)
bus.Set(state.Bus)
// the redact store is process-global and only cleared when a run ends (see the post-run below), so
// setting it while another run is still in flight panics rather than splitting secrets across stores.
redact.Set(state.RedactStore)
log.Set(state.Logger)
@@ -48,8 +50,14 @@ func AppClioSetupConfig(id clio.Identification, out io.Writer) *clio.SetupConfig
return nil
},
).
WithPostRuns(func(_ *clio.State, _ error) {
WithPostRuns(func(state *clio.State, _ error) {
stereoscope.Cleanup() //nolint:staticcheck // we don't have access to the image object here
// release the redact store now that nothing in this run can report anymore, so a later run in the same
// process (e.g. syft embedded as a library) can set its own. Only release the store this run set.
if redact.Get() == state.RedactStore {
redact.Reset()
}
})
return clioCfg
}
@@ -0,0 +1,32 @@
package internal
import (
"io"
"testing"
"github.com/stretchr/testify/require"
"github.com/anchore/clio"
"github.com/anchore/go-logger/adapter/discard"
gologgerredact "github.com/anchore/go-logger/adapter/redact"
"github.com/anchore/syft/internal/redact"
)
func TestAppClioSetupConfigInitializerPanicsWhileRunInFlight(t *testing.T) {
t.Cleanup(redact.Reset)
cfg := AppClioSetupConfig(clio.Identification{Name: "syft"}, io.Discard)
require.Len(t, cfg.Initializers, 1)
newState := func() *clio.State {
return &clio.State{Logger: discard.New(), RedactStore: gologgerredact.NewStore()}
}
require.NoError(t, cfg.Initializers[0](newState()))
// a second run starting before the first has finished must not silently replace the store, otherwise
// secrets added to the first store would no longer be redacted
require.PanicsWithValue(t, "replace existing redaction store (probably unintentional)", func() {
_ = cfg.Initializers[0](newState())
})
}
+7 -7
View File
@@ -18,11 +18,11 @@ require (
github.com/anchore/go-homedir v0.1.1
github.com/anchore/go-logger v0.2.0
github.com/anchore/go-macholibre v0.1.1
github.com/anchore/go-rpmdb v0.2.2-0.20261002172926-a488674d6c87
github.com/anchore/go-rpmdb v0.2.2
github.com/anchore/go-sync v0.1.2
github.com/anchore/go-version v1.2.2-0.20200701162849-18adb9c92b9b
github.com/anchore/packageurl-go v0.2.0
github.com/anchore/stereoscope v0.3.4-0.20261002170525-59d3a69590a6
github.com/anchore/stereoscope v0.3.4
github.com/anmitsu/go-shlex v0.0.0-20200514113438-38f4b401e2be
github.com/aquasecurity/go-pep440-version v0.0.1
github.com/bitnami/go-version v0.0.0-20250131085805-b1f57a8634ef
@@ -188,7 +188,7 @@ require (
github.com/containers/ocicrypt v1.3.2 // indirect
github.com/cyberphone/json-canonicalization v0.0.0-20241213102144-19d51d7fe467 // indirect
github.com/cyphar/filepath-securejoin v0.7.0 // indirect
github.com/docker/cli v29.8.1+incompatible // indirect
github.com/docker/cli v29.8.2+incompatible // indirect
github.com/docker/docker-credential-helpers v0.9.8 // indirect
github.com/docker/go-connections v0.8.1 // indirect
github.com/docker/go-units v0.5.0 // indirect
@@ -281,11 +281,11 @@ require (
github.com/rogpeppe/go-internal v1.14.1 // indirect
github.com/sagikazarmark/locafero v0.11.0 // indirect
github.com/sahilm/fuzzy v0.1.1 // indirect
github.com/secure-systems-lab/go-securesystemslib v0.11.0 // indirect
github.com/secure-systems-lab/go-securesystemslib v0.11.1 // indirect
github.com/shopspring/decimal v1.4.0 // indirect
github.com/sigstore/fulcio v1.8.7 // indirect
github.com/sigstore/protobuf-specs v0.5.1 // indirect
github.com/sigstore/sigstore v1.10.8 // indirect
github.com/sigstore/protobuf-specs v0.5.2 // indirect
github.com/sigstore/sigstore v1.10.11 // indirect
github.com/sirupsen/logrus v1.10.2 // indirect
github.com/skeema/knownhosts v1.3.1 // indirect
github.com/smallnest/ringbuffer v0.0.0-20241116012123-461381446e3d // indirect
@@ -299,7 +299,7 @@ require (
github.com/stefanberger/go-pkcs11uri v0.0.0-20230803200340-78284954bff6 // indirect
github.com/stretchr/objx v0.5.3 // indirect
github.com/subosito/gotenv v1.6.0 // indirect
github.com/sylabs/sif/v2 v2.24.1 // indirect
github.com/sylabs/sif/v2 v2.24.2 // indirect
github.com/sylabs/squashfs v1.0.6 // indirect
github.com/tchap/go-patricia/v2 v2.3.3 // indirect
github.com/therootcompany/xz v1.0.1 // indirect
+14 -14
View File
@@ -140,8 +140,8 @@ github.com/anchore/go-lzo v0.1.1 h1:IwL/fvkdtlIrYIXck6WxZ3nb8WjjHziYYmGxlooyOnM=
github.com/anchore/go-lzo v0.1.1/go.mod h1:3kLx0bve2oN1iDwgM1U5zGku1Tfbdb0No5qp1eL1fIk=
github.com/anchore/go-macholibre v0.1.1 h1:sTBHi3S1WVIOCamc6NyT8Sn9mBmD30DCTfjfVm4k6/M=
github.com/anchore/go-macholibre v0.1.1/go.mod h1:YNq2610RlvGCK0Za+Klz/OCMtUaEnwwAGA/VfUAQaro=
github.com/anchore/go-rpmdb v0.2.2-0.20261002172926-a488674d6c87 h1:aY7DcEkjgzMRbCCJsGHu8eQeL5I88V8G9ISEWtfrlCU=
github.com/anchore/go-rpmdb v0.2.2-0.20261002172926-a488674d6c87/go.mod h1:Y64EbetdqhryMjtk8IzXkdCQw/aNDx+feZCT9aj6rjs=
github.com/anchore/go-rpmdb v0.2.2 h1:44VC7mvFAxV9t5JUZYg3S9sqG+E94iZvyWACMDq9hSk=
github.com/anchore/go-rpmdb v0.2.2/go.mod h1:Y64EbetdqhryMjtk8IzXkdCQw/aNDx+feZCT9aj6rjs=
github.com/anchore/go-struct-converter v0.2.0-rc2 h1:q+859fW2/jbHJHB2etbNfRlFwYpknyvbqqk1hUdamQ4=
github.com/anchore/go-struct-converter v0.2.0-rc2/go.mod h1:cDBA5vhcR62nXWo8QH9/Kk2807o65ISaHPNPX66L+Uw=
github.com/anchore/go-sync v0.1.2 h1:RDD5RCanrWqhbtBulqPSZpERadUlgcYjpybYzYd3Uk8=
@@ -150,8 +150,8 @@ 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 v0.2.0 h1:CkrM4RMUwrEGAiE1OVlxaZNzWj0TuHRey7o4T/EAErk=
github.com/anchore/packageurl-go v0.2.0/go.mod h1:2JCgOQMIsqZ7TmliXG4PnUthPJAKE3mWQbsW2XHjAOE=
github.com/anchore/stereoscope v0.3.4-0.20261002170525-59d3a69590a6 h1:6l3YwREGEXX/w4tBhRcPb7efYWo4pf8+fmAy0mND4EY=
github.com/anchore/stereoscope v0.3.4-0.20261002170525-59d3a69590a6/go.mod h1:LxL4iOFtB8XMbA1MykUUTkmtQmDuYdbr4+TkBkCfljU=
github.com/anchore/stereoscope v0.3.4 h1:dT5AfjCTXUJowMiRIc+VBQVwvwnIEwnxxwA1GwAkMxw=
github.com/anchore/stereoscope v0.3.4/go.mod h1:+HxqHTt7dvH8mea0mf3ulqkf084VvHiT1grpAQf5uEs=
github.com/andreyvit/diff v0.0.0-20170406064948-c7f18ee00883/go.mod h1:rCTlJbsFo29Kk6CurOXKm700vrz8f0KW0JNfpkRJY/8=
github.com/andybalholm/brotli v1.2.0 h1:ukwgCxwYrmACq68yiUqwIWnGY0cTPox/M94sVwToPjQ=
github.com/andybalholm/brotli v1.2.0/go.mod h1:rzTDkvFWvIrjDXZHkuS16NPggd91W3kUSvPlQ1pLaKY=
@@ -353,8 +353,8 @@ github.com/distribution/reference v0.6.0 h1:0IXCQ5g4/QMHHkarYzh5l+u8T3t73zM5Qvfr
github.com/distribution/reference v0.6.0/go.mod h1:BbU0aIcezP1/5jX/8MP0YiH4SdvB5Y4f/wlDRiLyi3E=
github.com/djherbis/times v1.6.0 h1:w2ctJ92J8fBvWPxugmXIv7Nz7Q3iDMKNx9v5ocVH20c=
github.com/djherbis/times v1.6.0/go.mod h1:gOHeRAz2h+VJNZ5Gmc/o7iD9k4wW7NMVqieYCY99oc0=
github.com/docker/cli v29.8.1+incompatible h1:qYL1bCp6cRw2SB1xmLlIOPyV171dilw9W2Jew38vy9c=
github.com/docker/cli v29.8.1+incompatible/go.mod h1:JLrzqnKDaYBop7H2jaqPtU4hHvMKP+vjCwu2uszcLI8=
github.com/docker/cli v29.8.2+incompatible h1:2zgdFuoFst2T80oS42vhKGxkd0JRo305eIEKTsxR7pQ=
github.com/docker/cli v29.8.2+incompatible/go.mod h1:JLrzqnKDaYBop7H2jaqPtU4hHvMKP+vjCwu2uszcLI8=
github.com/docker/distribution v2.8.3+incompatible h1:AtKxIZ36LoNK51+Z6RpzLpddBirtxJnzDrHLEKxTAYk=
github.com/docker/distribution v2.8.3+incompatible/go.mod h1:J2gT2udsDAN96Uj4KfcMRqY0/ypR+oyYUYmja8H+y+w=
github.com/docker/docker-credential-helpers v0.9.8 h1:bIREROb7So6PRlq6KTtdS9MPEjC29OQRkFNlvK2OX8Q=
@@ -893,8 +893,8 @@ github.com/scylladb/go-set v1.0.3-0.20200225121959-cc7b2070d91e/go.mod h1:DkpGd7
github.com/sean-/seed v0.0.0-20170313163322-e2103e2c3529/go.mod h1:DxrIzT+xaE7yg65j358z/aeFdxmN0P9QXhEzd20vsDc=
github.com/sebdah/goldie/v2 v2.8.0 h1:dZb9wR8q5++oplmEiJT+U/5KyotVD+HNGCAc5gNr8rc=
github.com/sebdah/goldie/v2 v2.8.0/go.mod h1:oZ9fp0+se1eapSRjfYbsV/0Hqhbuu3bJVvKI/NNtssI=
github.com/secure-systems-lab/go-securesystemslib v0.11.0 h1:iuCR9kcMFD4QurdKrGvPLoKZLv9YvwPYVr0473BdtFs=
github.com/secure-systems-lab/go-securesystemslib v0.11.0/go.mod h1:+PMOTjUGwHj2vcZ+TFKlb1tXRbrdWE1LYDT5i9JC80Q=
github.com/secure-systems-lab/go-securesystemslib v0.11.1 h1:ayahDPjfSIqKegyt5YVGEvQ7SAi72GHXTy+b2YjUd5w=
github.com/secure-systems-lab/go-securesystemslib v0.11.1/go.mod h1:UyOjhoZLi76ir63u1ptubEevVe6dlNPmNxwgVvKqFFY=
github.com/sergi/go-diff v1.1.0/go.mod h1:STckp+ISIX8hZLjrqAeVduY0gWCT9IjLuqbuNXdaHfM=
github.com/sergi/go-diff v1.2.0/go.mod h1:STckp+ISIX8hZLjrqAeVduY0gWCT9IjLuqbuNXdaHfM=
github.com/sergi/go-diff v1.4.0 h1:n/SP9D5ad1fORl+llWyN+D6qoUETXNZARKjyY2/KVCw=
@@ -903,10 +903,10 @@ github.com/shopspring/decimal v1.4.0 h1:bxl37RwXBklmTi0C79JfXCEBD1cqqHt0bbgBAGFp
github.com/shopspring/decimal v1.4.0/go.mod h1:gawqmDU56v4yIKSwfBSFip1HdCCXN8/+DMd9qYNcwME=
github.com/sigstore/fulcio v1.8.7 h1:d7QJQxukgXarqPCT9uDNeyzt1Bbt/biw58/VfOnwI/o=
github.com/sigstore/fulcio v1.8.7/go.mod h1:7fX+QyigZxRGKTgHvTfOfxwwniCDxK02PVloHy5/vg4=
github.com/sigstore/protobuf-specs v0.5.1 h1:/5OPaNuolRJmQfeZLayJGFXMpsRJEdgC6ah1/+7Px7U=
github.com/sigstore/protobuf-specs v0.5.1/go.mod h1:DRBzpFuE+LnvQMN10/dU6nBeKwVLGEQ6o2FovN2Rats=
github.com/sigstore/sigstore v1.10.8 h1:1Mgkxvkw4AXMfIP1DOjc6kw0GkUgA8pGVpveN/EfOq4=
github.com/sigstore/sigstore v1.10.8/go.mod h1:f9+B/4iaYimvUkySyb2mvc73n3RLqNn24grHZM/ET8M=
github.com/sigstore/protobuf-specs v0.5.2 h1:RSWWUY8QrVTxbYH00jY/jg2e7YnjzrpwP+PeHTMll0E=
github.com/sigstore/protobuf-specs v0.5.2/go.mod h1:DRBzpFuE+LnvQMN10/dU6nBeKwVLGEQ6o2FovN2Rats=
github.com/sigstore/sigstore v1.10.11 h1:fjrwDfR0AWmRONYV/hKGJyqbT0osMtaBz9FTZ6MH9XI=
github.com/sigstore/sigstore v1.10.11/go.mod h1:ecPk+ftLqP8/TmzQFJmn0+QrL2yUQIJvyzVKNmDiw6U=
github.com/sirupsen/logrus v1.2.0/go.mod h1:LxeOpSwHxABJmUn/MG1IvRgCAasNZTLOkJPxbbu5VWo=
github.com/sirupsen/logrus v1.4.2/go.mod h1:tLMulIdttU9McNUspp0xgXVQah82FyeX6MwdIuYE2rE=
github.com/sirupsen/logrus v1.7.0/go.mod h1:yWOB1SBYBC5VeMP7gHvWumXLIWorT60ONWic61uBYv0=
@@ -970,8 +970,8 @@ github.com/stretchr/testify v1.12.1/go.mod h1:MDEgiDPPsNp5cuIrHPPCyornHKgEVbtFUm
github.com/subosito/gotenv v1.2.0/go.mod h1:N0PQaV/YGNqwC0u51sEeR/aUtSLEXKX9iv69rRypqCw=
github.com/subosito/gotenv v1.6.0 h1:9NlTDc1FTs4qu0DDq7AEtTPNw6SVm7uBMsUCUjABIf8=
github.com/subosito/gotenv v1.6.0/go.mod h1:Dk4QP5c2W3ibzajGcXpNraDfq2IrhjMIvMSWPKKo0FU=
github.com/sylabs/sif/v2 v2.24.1 h1:OhTOfTwBaGXfbWYWXK9C6Pojma3A1bTaNJ6CEhyuKok=
github.com/sylabs/sif/v2 v2.24.1/go.mod h1:PRq9MoP0g+p0qE5ZRGrOhyVAUdHrPsAVmXHgoiW6VU0=
github.com/sylabs/sif/v2 v2.24.2 h1:rdkFf6z3xLC6yKweU7S9marlKfkS91Ht0ruIri6YWFY=
github.com/sylabs/sif/v2 v2.24.2/go.mod h1:bbrkk0JhTX+xNJcnUtmDkivoO2r4DGWM3LxNR0U2ooA=
github.com/sylabs/squashfs v1.0.6 h1:PvJcDzxr+vIm2kH56mEMbaOzvGu79gK7P7IX+R7BDZI=
github.com/sylabs/squashfs v1.0.6/go.mod h1:DlDeUawVXLWAsSRa085Eo0ZenGzAB32JdAUFaB0LZfE=
github.com/tailscale/hujson v0.0.0-20260302212456-ecc657c15afd h1:Rf9uhF1+VJ7ZHqxrG8pJ6YacmHvVCmByDmGbAWCc/gA=
+8
View File
@@ -13,6 +13,14 @@ func Set(s redact.Store) {
store = s
}
// Reset clears the global redaction store. It must only be called once a command
// run has fully finished (nothing left to Add or Apply), so that a later run can
// Set its own store. Calling it while a run is in flight would let a second Set
// split secrets across two stores and leak the ones in the dropped store.
func Reset() {
store = nil
}
func Get() redact.Store {
return store
}
@@ -75,6 +75,7 @@ catalogers:
method: glob # AUTO-GENERATED
criteria: # AUTO-GENERATED
- '**/*.deb'
- '**/*.ipk'
metadata_types: # AUTO-GENERATED
- pkg.DpkgArchiveEntry
package_types: # AUTO-GENERATED
+2 -2
View File
@@ -18,8 +18,8 @@ func NewDBCataloger() pkg.Cataloger {
WithProcessors(dependency.Processor(dbEntryDependencySpecifier))
}
// NewArchiveCataloger returns a new Debian package cataloger object capable of parsing .deb archive files
// NewArchiveCataloger returns a new Debian package cataloger object capable of parsing .deb and .ipk (opkg) archive files
func NewArchiveCataloger() pkg.Cataloger {
return generic.NewCataloger("deb-archive-cataloger").
WithParserByGlobs(parseDebArchive, "**/*.deb")
WithParserByGlobs(parseDebArchive, "**/*.deb", "**/*.ipk")
}
@@ -276,6 +276,36 @@ func TestDpkgArchiveCataloger(t *testing.T) {
},
},
},
}, {
name: "image-single-ipk",
expected: []pkg.Package{
{
Name: "libz1",
Version: "1.3.2-r0",
FoundBy: "deb-archive-cataloger",
Locations: file.NewLocationSet(
file.NewLocation("/libz1_1.3.2-r0_x86-64-v3.ipk"),
),
Type: pkg.DebPkg,
Metadata: pkg.DpkgArchiveEntry{
Package: "libz1",
Source: "zlib",
SourceVersion: "1.3.2",
Version: "1.3.2-r0",
Architecture: "x86-64-v3",
Maintainer: "Poky Maintainers <poky@lists.yoctoproject.org>",
InstalledSize: 0,
Description: `Zlib Compression Library
Zlib is a general-purpose, patent-free, lossless data compression library
which is used by many different programs.`,
Provides: []string{"zlib"},
Depends: []string{"musl (>= 1.2.6+git0+9fa28ece75)"},
},
Licenses: pkg.NewLicenseSet(
pkg.NewLicenseFromLocationsWithContext(ctx, "Zlib"),
),
},
},
},
}
+61 -15
View File
@@ -9,6 +9,8 @@ import (
"sort"
"strings"
"github.com/scylladb/go-set/strset"
"github.com/anchore/packageurl-go"
"github.com/anchore/syft/internal"
"github.com/anchore/syft/internal/log"
@@ -61,32 +63,65 @@ func newDpkgPackage(ctx context.Context, d dpkgExtractedMetadata, dbLocation fil
return p
}
// a declared License field holds a single expression, a few hundred bytes at most in practice. Anything
// much larger only feeds the expression parser (which recurses on parentheses) attacker-sized input.
const maxDeclaredLicenseLen = 4096
var (
// opkg/ipkg (Yocto via bitbake) declare licenses with "&" for AND and "|" for OR
bitbakeLicenseOperators = strings.NewReplacer("&", " AND ", "|", " OR ")
tightenParens = strings.NewReplacer("( ", "(", " )", ")")
licenseListSeparators = strings.NewReplacer("(", " ", ")", " ")
)
// extractDeclaredLicenses converts a License field from the status DB into a license set. Returns nil
// for empty input so standard dpkg entries (which never declare License inline) incur no allocation.
// Mirrors the alpine cataloger's approach: keep the value whole if it parses as a valid SPDX expression,
// otherwise split on whitespace to handle space-separated lists.
func extractDeclaredLicenses(ctx context.Context, raw string, dbLocation file.Location) []pkg.License {
values := declaredLicenseValues(raw)
if len(values) == 0 {
return nil
}
return pkg.NewLicensesFromLocationWithContext(ctx, dbLocation, values...)
}
// declaredLicenseValues converts an inline License field (opkg status DB or an archive control file) into
// license values. Bitbake operators are rewritten to SPDX ones and the value is kept whole if it then parses
// as a valid SPDX expression (mirroring the alpine cataloger), otherwise it is treated as a list of license
// names with operators and grouping dropped.
func declaredLicenseValues(raw string) []string {
raw = strings.TrimSpace(raw)
if raw == "" {
return nil
}
licenseStrings := []string{raw}
if _, err := license.ParseExpression(raw); err != nil {
licenseStrings = strings.Fields(raw)
if len(raw) > maxDeclaredLicenseLen {
log.WithFields("length", len(raw)).Debug("ignoring oversized declared license field")
return nil
}
return pkg.NewLicensesFromLocationWithContext(ctx, dbLocation, licenseStrings...)
expr := tightenParens.Replace(strings.Join(strings.Fields(bitbakeLicenseOperators.Replace(raw)), " "))
if _, err := license.ParseExpression(expr); err == nil {
return []string{expr}
}
var values []string
seen := strset.New()
for _, field := range strings.Fields(licenseListSeparators.Replace(expr)) {
if field == "AND" || field == "OR" || seen.Has(field) {
continue
}
seen.Add(field)
values = append(values, field)
}
return values
}
func newDebArchivePackage(ctx context.Context, location file.Location, metadata pkg.DpkgArchiveEntry, licenseStrings []string) pkg.Package {
p := pkg.Package{
Name: metadata.Package,
Version: metadata.Version,
Licenses: pkg.NewLicenseSet(pkg.NewLicensesFromValuesWithContext(ctx, licenseStrings...)...),
Type: pkg.DebPkg,
PURL: packageURL(
pkg.DpkgDBEntry(metadata),
// we don't know the distro information, but since this is a deb file then we can reasonably assume it is a debian-based distro
&linux.Release{IDLike: []string{"debian"}},
),
Name: metadata.Package,
Version: metadata.Version,
Licenses: pkg.NewLicenseSet(pkg.NewLicensesFromValuesWithContext(ctx, licenseStrings...)...),
Type: pkg.DebPkg,
PURL: packageURL(pkg.DpkgDBEntry(metadata), archiveDistro(location)),
Metadata: metadata,
Locations: file.NewLocationSet(location.WithAnnotation(pkg.EvidenceAnnotationKey, pkg.PrimaryEvidenceAnnotation)),
}
@@ -95,6 +130,17 @@ func newDebArchivePackage(ctx context.Context, location file.Location, metadata
return p
}
// archiveDistro guesses the distro an archive was built for. We don't know the distro information, but a .deb
// can reasonably be assumed to target a debian-based distro. An .ipk targets opkg-based distros (Yocto,
// OpenWrt) which have no purl type, so it gets no distro and therefore no PURL, the same as packages found in
// an opkg status DB.
func archiveDistro(location file.Location) *linux.Release {
if strings.EqualFold(path.Ext(location.Path()), ".ipk") {
return nil
}
return &linux.Release{IDLike: []string{"debian"}}
}
// PackageURL returns the PURL for the specific Debian package (see https://github.com/package-url/purl-spec)
func packageURL(m pkg.DpkgDBEntry, distro *linux.Release) string {
if distro == nil {
+31
View File
@@ -3,6 +3,7 @@ package debian
import (
"context"
"sort"
"strings"
"testing"
"github.com/google/go-cmp/cmp"
@@ -145,6 +146,36 @@ func Test_extractDeclaredLicenses(t *testing.T) {
raw: "GPL-2.0 BSD-3-Clause",
expected: []string{"GPL-2.0", "BSD-3-Clause"},
},
{
name: "bitbake operators become an SPDX expression",
raw: "MIT & (BSD-3-Clause | Apache-2.0) & ( GPL-2.0-or-later )",
expected: []string{"MIT AND (BSD-3-Clause OR Apache-2.0) AND (GPL-2.0-or-later)"},
},
{
name: "bitbake grouping is preserved rather than split",
raw: "(MIT & BSD-3-Clause) | Apache-2.0",
expected: []string{"(MIT AND BSD-3-Clause) OR Apache-2.0"},
},
{
name: "parenthesized alternatives are preserved",
raw: "(MIT) | (Apache-2.0)",
expected: []string{"(MIT) OR (Apache-2.0)"},
},
{
name: "non-SPDX names drop operators and grouping",
raw: "GPLv2+ & (LGPLv2.1 | BSD) & GPLv2+",
expected: []string{"GPLv2+", "LGPLv2.1", "BSD"},
},
{
name: "oversized field is ignored",
raw: strings.Repeat("MIT & ", maxDeclaredLicenseLen),
expected: nil,
},
{
name: "deeply nested parens are ignored",
raw: strings.Repeat("(", maxDeclaredLicenseLen) + "MIT" + strings.Repeat(")", maxDeclaredLicenseLen),
expected: nil,
},
}
for _, test := range tests {
+46 -15
View File
@@ -2,6 +2,7 @@ package debian
import (
"archive/tar"
"bytes"
"context"
"errors"
"fmt"
@@ -23,18 +24,27 @@ import (
"github.com/anchore/syft/syft/pkg/cataloger/generic"
)
// parseDebArchive parses a Debian package archive (.deb) file and returns the packages it contains.
// A .deb file is an ar archive containing three main files:
// - debian-binary: Version of the .deb format (usually "2.0")
// parseDebArchive parses a Debian package archive (.deb) or opkg package (.ipk) file and returns the packages it
// contains. Both are ar archives containing three main files:
// - debian-binary: Version of the archive format (usually "2.0")
// - control.tar.gz/xz/zst: Contains package metadata (control file, md5sums, conffiles)
// - data.tar.gz/xz/zst: Contains the actual files to be installed (not processed by this cataloger)
//
// This function extracts and processes the control information to create package metadata.
// This function extracts and processes the control information to create package metadata. Licenses come
// from copyright files in the data tar, falling back to the control file's License field (common for .ipk).
// Only ar-framed .ipk files are supported; the older gzipped-tar framing (e.g. OpenWrt's ipkg-build) is
// rejected as an invalid archive.
func parseDebArchive(ctx context.Context, _ file.Resolver, _ *generic.Environment, reader file.LocationReadCloser) ([]pkg.Package, []artifact.Relationship, error) {
arReader := ar.NewReader(reader)
validatedReader, err := newValidatedArReader(reader)
if err != nil {
return nil, nil, fmt.Errorf("invalid ar archive: %w", err)
}
arReader := ar.NewReader(validatedReader)
var metadata *pkg.DpkgArchiveEntry
var licenses []string
var ctrlLicenses []string
var unknownErr error
var sawControl, sawData bool
for {
@@ -56,7 +66,7 @@ func parseDebArchive(ctx context.Context, _ file.Resolver, _ *generic.Environmen
if err != nil {
return nil, nil, unknown.New(reader.Location, fmt.Errorf("failed to decompress control.tar.* file: %w", err))
}
metadata, err = processControlTar(dcReader)
metadata, ctrlLicenses, err = processControlTar(dcReader)
switch {
case err != nil && metadata == nil:
return nil, nil, unknown.New(reader.Location, fmt.Errorf("failed to process control.tar.* file: %w", err))
@@ -87,6 +97,10 @@ func parseDebArchive(ctx context.Context, _ file.Resolver, _ *generic.Environmen
return nil, nil, unknown.New(reader.Location, fmt.Errorf("no application found described in .dpkg archive"))
}
if len(licenses) == 0 && len(ctrlLicenses) > 0 {
licenses = ctrlLicenses
}
// a partial parse still yields a usable package, so report what went wrong alongside it rather than
// dropping either one
return []pkg.Package{
@@ -128,10 +142,10 @@ func processDataTar(dcReader io.ReadCloser) ([]string, error) {
}
// processControlTar always returns whatever metadata it managed to parse, even alongside a non-nil
// error. A non-nil error with non-nil metadata means the package is usable but incomplete (e.g. a
// clipped file listing); nil metadata means nothing usable was found and the caller should treat it
// as fatal.
func processControlTar(dcReader io.ReadCloser) (*pkg.DpkgArchiveEntry, error) {
// error, along with any licenses declared in the control file's License field. A non-nil error with
// non-nil metadata means the package is usable but incomplete (e.g. a clipped file listing); nil
// metadata means nothing usable was found and the caller should treat it as fatal.
func processControlTar(dcReader io.ReadCloser) (*pkg.DpkgArchiveEntry, []string, error) {
defer internal.CloseAndLogError(dcReader, "")
tarReader := tar.NewReader(dcReader)
@@ -140,6 +154,7 @@ func processControlTar(dcReader io.ReadCloser) (*pkg.DpkgArchiveEntry, error) {
var files []pkg.DpkgFileRecord
var confFileRecords []pkg.DpkgFileRecord
var listingErr error
var licenses []string
for {
header, err := tarReader.Next()
@@ -147,7 +162,7 @@ func processControlTar(dcReader io.ReadCloser) (*pkg.DpkgArchiveEntry, error) {
break
}
if err != nil {
return metadata, fmt.Errorf("failed to read control tar: %w", err)
return metadata, nil, fmt.Errorf("failed to read control tar: %w", err)
}
switch filepath.Base(header.Name) {
@@ -155,12 +170,13 @@ func processControlTar(dcReader io.ReadCloser) (*pkg.DpkgArchiveEntry, error) {
// parseDpkgStatus already streams via bufio.Reader
entries, err := parseDpkgStatus(tarReader)
if err != nil {
return nil, fmt.Errorf("failed to parse control file: %w", err)
return nil, nil, fmt.Errorf("failed to parse control file: %w", err)
}
if len(entries) == 0 {
return nil, fmt.Errorf("no package entries found in control file")
return nil, nil, fmt.Errorf("no package entries found in control file")
}
entry := pkg.DpkgArchiveEntry(entries[0].toDpkgEntry())
licenses = declaredLicenseValues(entries[0].License)
metadata = &entry
case "md5sums":
// parseDpkgMD5Info streams via bufio.Scanner and reports its own clipping/scan errors
@@ -176,7 +192,7 @@ func processControlTar(dcReader io.ReadCloser) (*pkg.DpkgArchiveEntry, error) {
}
if metadata == nil {
return nil, fmt.Errorf("control file not found in archive")
return nil, nil, fmt.Errorf("control file not found in archive")
}
if len(confFileRecords) > 0 && len(files) > 0 {
@@ -195,7 +211,22 @@ func processControlTar(dcReader io.ReadCloser) (*pkg.DpkgArchiveEntry, error) {
// a clipped or unreadable file listing leaves the package usable, so hand both back and let the
// caller decide
return metadata, listingErr
return metadata, licenses, listingErr
}
// newValidatedArReader checks for the ar global header before handing the stream to the ar reader, which
// otherwise skips those bytes unchecked and reads garbage headers from non-ar input.
func newValidatedArReader(reader io.ReadCloser) (io.ReadCloser, error) {
prefix := make([]byte, len(ar.GLOBAL_HEADER))
if _, err := io.ReadFull(reader, prefix); err != nil {
return nil, fmt.Errorf("failed to read ar header: %w", err)
}
if !bytes.Equal(prefix, []byte(ar.GLOBAL_HEADER)) {
return nil, fmt.Errorf("expected ar header %q, got %q", string(ar.GLOBAL_HEADER), string(prefix))
}
return io.NopCloser(io.MultiReader(bytes.NewReader(prefix), reader)), nil
}
// a .deb's member sizes are bounded by the archive, but what they decompress to is not: xz and zstd
@@ -26,12 +26,13 @@ import (
func TestProcessControlTar(t *testing.T) {
tarBytes := createTestTarWithControlFiles(t)
metadata, err := processControlTar(io.NopCloser(bytes.NewReader(tarBytes)))
metadata, licenses, err := processControlTar(io.NopCloser(bytes.NewReader(tarBytes)))
require.NoError(t, err)
require.NotNil(t, metadata)
assert.Equal(t, "test-package", metadata.Package)
assert.Equal(t, []string{"MIT"}, licenses)
assert.Equal(t, "1.0.0", metadata.Version)
// md5sums should have been parsed into file records
@@ -43,6 +44,14 @@ func TestProcessControlTar(t *testing.T) {
assert.True(t, metadata.Files[0].IsConfigFile, "file listed in conffiles should be marked as config")
}
func TestParseDebArchive_RejectsInvalidArchiveHeader(t *testing.T) {
reader := file.NewLocationReadCloser(file.NewLocation("test.deb"), io.NopCloser(bytes.NewReader([]byte("not-an-ar-archive"))))
_, _, err := parseDebArchive(context.Background(), nil, nil, reader)
require.Error(t, err)
assert.Contains(t, err.Error(), "ar archive")
}
func TestProcessControlTar_ConfigFileMarking(t *testing.T) {
// Create a tar where conffiles lists paths that overlap with md5sums entries
var buf bytes.Buffer
@@ -61,9 +70,10 @@ func TestProcessControlTar_ConfigFileMarking(t *testing.T) {
require.NoError(t, tw.Close())
metadata, err := processControlTar(io.NopCloser(bytes.NewReader(buf.Bytes())))
metadata, licenses, err := processControlTar(io.NopCloser(bytes.NewReader(buf.Bytes())))
require.NoError(t, err)
require.Len(t, metadata.Files, 3)
assert.Empty(t, licenses)
assert.True(t, metadata.Files[0].IsConfigFile, "first file should be marked as config file")
assert.True(t, metadata.Files[1].IsConfigFile, "second file should be marked as config file")
@@ -75,7 +85,7 @@ func createTestTarWithControlFiles(t *testing.T) []byte {
var buf bytes.Buffer
tw := tar.NewWriter(&buf)
controlContent := "Package: test-package\nVersion: 1.0.0\nArchitecture: all\nMaintainer: Test <test@example.com>\nDescription: Test package\n"
controlContent := "Package: test-package\nVersion: 1.0.0\nArchitecture: all\nMaintainer: Test <test@example.com>\nDescription: Test package\nLicense: MIT\n"
writeTarEntry(t, tw, "control", controlContent)
md5Content := "d41d8cd98f00b204e9800998ecf8427e usr/bin/test-command\n"
@@ -165,7 +175,7 @@ func Test_processControlTar_rejectsABombedControlMember(t *testing.T) {
writeTarEntry(t, tw, "md5sums", strings.Repeat("d41d8cd98f00b204e9800998ecf8427e usr/bin/x\n", 20000))
require.NoError(t, tw.Close())
metadata, err := processControlTar(newBoundedReadCloser(io.NopCloser(bytes.NewReader(tarBuf.Bytes())), 4096))
metadata, _, err := processControlTar(newBoundedReadCloser(io.NopCloser(bytes.NewReader(tarBuf.Bytes())), 4096))
require.ErrorIs(t, err, errDecompressedTooLarge)
require.NotNil(t, metadata, "a control tar read failure after \"control\" was parsed must not drop the package")
@@ -181,7 +191,7 @@ func Test_processControlTar_clippedListingIsUsableButReported(t *testing.T) {
writeTarEntry(t, tw, "md5sums", strings.Repeat("d41d8cd98f00b204e9800998ecf8427e usr/bin/x\n", maxDpkgFileRecords+10))
require.NoError(t, tw.Close())
metadata, err := processControlTar(io.NopCloser(bytes.NewReader(tarBuf.Bytes())))
metadata, _, err := processControlTar(io.NopCloser(bytes.NewReader(tarBuf.Bytes())))
require.Error(t, err, "clipping must be reported")
require.ErrorIs(t, err, errClippedFileListing)
@@ -418,3 +428,18 @@ func Test_decompressionStream_failsEarlyInsteadOfBuffering(t *testing.T) {
assert.Less(t, bounded, uint64(8*intFile.MB),
"a bounded stream must stop at the cap, not buffer the whole member and check afterwards")
}
func TestProcessControlTar_DeclaredLicense(t *testing.T) {
var buf bytes.Buffer
tw := tar.NewWriter(&buf)
controlContent := "Package: test-package\nVersion: 1.0.0\nArchitecture: all\nLicense: MIT & (BSD-3-Clause | Apache-2.0) & ( GPL-2.0-or-later )\n"
writeTarEntry(t, tw, "control", controlContent)
require.NoError(t, tw.Close())
metadata, licenses, err := processControlTar(io.NopCloser(bytes.NewReader(buf.Bytes())))
require.NoError(t, err)
require.NotNil(t, metadata)
assert.Equal(t, []string{"MIT AND (BSD-3-Clause OR Apache-2.0) AND (GPL-2.0-or-later)"}, licenses)
}
+10 -2
View File
@@ -183,9 +183,11 @@ func parseDpkgStatusEntry(reader *bufio.Reader) (*dpkgExtractedMetadata, error)
}
sourceName, sourceVersion := extractSourceVersion(raw.Source)
if sourceName != "" {
raw.Source = sourceName
}
if sourceVersion != "" {
raw.SourceVersion = sourceVersion
raw.Source = sourceName
}
if raw.Package == "" {
@@ -263,8 +265,14 @@ func extractAllFields(reader *bufio.Reader) (map[string]any, error) {
}
// If the source entry string is of the form "<name> (<version>)" then parse and return the components, if
// of the "<name>" form, then return name and nil
// of the "<name>" form, then return name and nil. Yocto (opkg/ipk) instead writes the bitbake recipe filename
// ("<name>_<version>.bb"), which is split the same way.
func extractSourceVersion(source string) (string, string) {
if recipe, ok := strings.CutSuffix(source, ".bb"); ok && !strings.ContainsAny(recipe, " (") {
// bitbake reserves "_" as the recipe name / version separator
name, version, _ := strings.Cut(recipe, "_")
return name, version
}
// special handling for the Source field since it has formatted data
match := internal.MatchNamedCaptureGroups(sourceRegexp, source)
return match["name"], match["version"]
@@ -366,6 +366,16 @@ func TestSourceVersionExtract(t *testing.T) {
input: "",
expected: []string{"", ""},
},
{
name: "bitbake recipe with version",
input: "zlib_1.3.2.bb",
expected: []string{"zlib", "1.3.2"},
},
{
name: "bitbake recipe without version",
input: "packagegroup-core-boot.bb",
expected: []string{"packagegroup-core-boot", ""},
},
}
for _, test := range tests {
@@ -0,0 +1,21 @@
# assembles a Yocto-style .ipk the same way bitbake's package_ipk does (opkg-build: a GNU ar archive of
# debian-binary, control.tar.gz and data.tar.zst) without paying for a full Yocto build. control/ is copied
# verbatim from libz1_1.3.2-r0_x86-64-v3.ipk built by poky-tiny (wrynose); the library itself is a stub
# since only the archive shape and control metadata matter here.
FROM alpine:3.18.4@sha256:48d9183eb12a05c99bcc0bf44a003607b8e941e1d4f41f9ad12bdcc4b5672f86 AS build
RUN apk add --no-cache binutils zstd
COPY control /pkg/control
RUN mkdir -p /pkg/data/usr/lib \
&& echo stub > /pkg/data/usr/lib/libz.so.1.3.2 \
&& ln -s libz.so.1.3.2 /pkg/data/usr/lib/libz.so.1 \
&& cd /pkg \
&& echo 2.0 > debian-binary \
&& tar -C control -czf control.tar.gz . \
&& tar -C data -cf - . | zstd -q -o data.tar.zst \
&& ar rcD /libz1_1.3.2-r0_x86-64-v3.ipk debian-binary control.tar.gz data.tar.zst
FROM scratch
COPY --from=build /libz1_1.3.2-r0_x86-64-v3.ipk /
@@ -0,0 +1,15 @@
Package: libz1
Version: 1.3.2-r0
Description: Zlib Compression Library
Zlib is a general-purpose, patent-free, lossless data compression library
which is used by many different programs.
Section: libs
Priority: optional
Maintainer: Poky Maintainers <poky@lists.yoctoproject.org>
License: Zlib
Architecture: x86-64-v3
OE: zlib
Homepage: http://zlib.net/
Depends: musl (>= 1.2.6+git0+9fa28ece75)
Provides: zlib
Source: zlib_1.3.2.bb
+5
View File
@@ -0,0 +1,5 @@
#!/bin/sh
set -e
if [ x"$D" = "x" ]; then
if [ -x /sbin/ldconfig ]; then /sbin/ldconfig ; fi
fi