Files
syft/syft/pkg/cataloger/binary/README.md
T
Alex Goodman b5e85c3ea5 chore: migrate fixtures to testdata (#4651)
* migrate fixtures to testdata

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

* fix: correct broken symlinks after testdata migration

The migration from test-fixtures to testdata broke several symlinks:
- elf-test-fixtures symlinks pointed to old test-fixtures paths
- elf-test-fixtures needed to be renamed to elf-testdata
- image-pkg-coverage symlink pointed to test-fixtures instead of testdata

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

* fix: handle missing classifiers/bin directory in Makefile

The clean-fingerprint target was failing when classifiers/bin doesn't
exist (e.g., on fresh clone without downloaded binaries).

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

* fix: add gitignore negation for jar/zip fixtures in test/cli

The jar and zip files in test/cli/testdata/image-unknowns were being
gitignored by the root .gitignore patterns. This caused them to be
untracked and not included when building docker images in CI, resulting
in Test_Unknowns failures since the test expects errors from corrupt
archive files that weren't present.

Add a .gitignore in test/cli/testdata to negate the exclusions for
these specific test fixture files.

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

* switch fixture cache to v2

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

* test: update expected versions for rebuilt fixtures

Update test expectations for packages that have been updated in
upstream repositories when docker images are rebuilt:
- glibc: 2.42-r4 → 2.43-r1 (wolfi)
- php: 8.2.29 → 8.2.30 (ubuntu/apache)

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

* upgrade go

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

* fix: add go-shlex dependency for testdata manager tool

The manager tool in syft/pkg/cataloger/binary/testdata/ imports
go-shlex, but since it's in a testdata directory, Go doesn't track
its dependencies. This caused CI failures when go.mod didn't
explicitly list the dependency.

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

* refactor: move binary classifier manager to internal/

Move the manager tool from testdata/manager to internal/manager so
that Go properly tracks its dependencies. Code in testdata directories
is ignored by Go for dependency tracking, which caused CI failures
when go.mod didn't explicitly list transitive dependencies.

This is a cleaner solution than manually adding dependencies to go.mod
for code that happens to live in testdata.

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

* fix: add gitignore negations for test fixtures blocked by root patterns

Multiple test fixtures were being blocked by root-level gitignore patterns
like bin/, *.jar, *.tar, and *.exe. This adds targeted .gitignore files with
negation patterns to allow these specific test fixtures to be tracked:

- syft/linux/testdata/os/busybox/bin/busybox (blocked by bin/)
- syft/pkg/cataloger/java/testdata/corrupt/example.{jar,tar} (blocked by *.jar, *.tar)
- syft/pkg/cataloger/binary/testdata/classifiers/snippets/go-version-hint/**/bin/go (blocked by bin/)
- syft/pkg/cataloger/bitnami/testdata/no-rel/.../bin/redis-server (blocked by bin/)

Also updates the bitnami test expectation to include the newly required
.gitignore files in the test fixture.

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

* test: update glibc version expectation (2.43-r1 -> 2.43-r2)

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

* add capability drift check as unit step

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

* dont clear test observations before drift detection

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

* bump stereoscope commit to main

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

---------

Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
2026-03-06 19:42:04 +00:00

5.3 KiB

Adding tests for the Binary classifier cataloger

Tip

TL;DR to add a test for a new classifier:

  1. head to the correct directory: cd testdata
  2. add a new entry to config.yaml to track where to get the binary from (verify the entry with make list)
  3. run make download to get the binary
  4. run make add-snippet and follow the prompts (use / to search)
  5. add a new test case to Test_Cataloger_PositiveCases in ../cataloger_test.go

To test the binary cataloger we run it against a set of files ("test fixtures"). There are two kinds of test fixtures:

  • Full binaries: files downloaded and cached at test runtime
  • Snippets: ~100 byte files checked into the repo

The upside with snippets is that they live with the test, don't necessarily require network access or hosting concerns, and are easy to add. The downside is that they are not the entire real binary so modifications may require recreating the snippet entirely.

The upside with full binaries is that they are the "Real McCoy" and allows the business logic to change without needing to update the fixture. The downside is that they require network access and take up a lot of space. For instance, downloading all binaries for testing today requires downloading ~15GB of container images and ends up being ~500MB of disk space.

You can find the test fixtures at the following locations:

syft/pkg/cataloger/binary/testdata/
└── classifiers/
    ├── bin/        # full binaries
    ├── ...
    └── snippets/   # snippets

And use tooling to list and manage the fixtures:

  • make list - list all fixtures
  • make download - download binaries that are not covered by a snippet
  • make download-all - download all binaries
  • go run ./manager add-snippet - add a new snippet based off of a configured binary
  • capture-snippet.sh - add a new snippet based off of a binary on your local machine (not recommended, but allowed)

There is a config.yaml that tracks all binaries that the tests can use. This makes it possible to download it at any time from a hosted source. Today the only method allowed is to download a container image and extract files out.

Here is an example entry in config.yaml for a binary reference from a container image:

from-images:
  
  # note the group-name is assumed from the single binary name extracted: "redis-server"
  - version: 7.2.3
    images:
      # note we're pulling the same binary from multiple images (representing different architectures)
      - ref: redis:7.2.3@sha256:d4c84914b872521e215f77d8845914c2268a96b0e35bacd5691e1f5e1f88b500
        platform: linux/amd64
      - ref: redis:7.2.3@sha256:a0a0c38b31011b813cddf78d997f8ccba13019c27efd386984b0cfc1e4b618ff
        platform: linux/arm64
    # the paths to extract from the binary...
    paths:
      - /usr/local/bin/redis-server

  # since there are multiple binaries in the image, we need to specify the group-name
  - name: ruby-bullseye-shared-libs
    version: 2.7.7
    images:
      - ref: ruby:2.7.7-bullseye@sha256:055191740a063f33fef1f09423e5ed8f91143aae62a3772a90910118464c5120
        platform: linux/amd64
    paths:
      - /usr/local/bin/ruby
      - /usr/local/lib/libruby.so.2.7.7
      - /usr/local/lib/libruby.so.2.7

Note

You will need a system with go, bash, strings, and xxd installed to capture test snippets.

Testing

The test cases have been setup to allow testing against full binaries or a mix of both (default). To force running only against full binaries run with:

go test -must-use-full-binaries ./syft/pkg/cataloger/binary/testdata/...

Adding a new test fixture

Even if you are adding a snippet, it is best practice to:

  • create that snippet from a full binary (not craft a snippet by hand)
  • track where the binary is from and how to download it in config.yaml
  1. Follow the steps above to add a full binary

  2. Run go run ./manager add-snippet and follow the prompts to create a new snippet

    • you should see your binary in the list of binaries to choose from. If not, check step 2
    • if the search results in no matching snippets, you can specify your own search with --search-for <grep-pattern>
    • you should see a new snippet file created in snippets/
  3. Write a test that references your new snippet by <name>/<version>/<architecture>

    • <name> is the name of the binary (e.g. curl) or the name in config.yaml if specified
    • note that your test does not know about if it's running against a snippet or a full binary

Adding a full binary

  1. Add a new entry to config.yaml with the following fields
  • if you are adding a single binary, the name field does not need to be specified
  • the name field is useful for distinguishing a quality about the binary (e.g. java vs java-jre-ibm)
  1. Run make download and ensure your new binary is downloaded

Adding a custom snippet

If you need to add a snippet that is not based off of a full binary, you can use the capture-snippet.sh script.

./capture-snippet.sh <binary-path> <version> [--search-for <pattern>] [--length <length>] [--prefix-length <prefix_length>] [--group <name>]

This is not recommended because it is not reproducible and does not allow for the test to be run against a full binary.