From 1fc57aa33902293c6ba22c2e27b596d7fa0d5bfb Mon Sep 17 00:00:00 2001
From: Michael Vogt <michael.vogt@gmail.com>
Date: Tue, 17 Sep 2024 13:15:42 +0200
Subject: [PATCH] stages(tar): expose new `transform` option to tar stage

This commit adds a new `transform` option to the tar stages that
maps directly to the `--transform=` comamndline argument of tar(1).

This allows to transform the names while files/dirs are added to
a tarfile. This is useful for the `gcp` pipeline for
bootc-image-builder where we want to create a gcp tar file that
expects the disk image filename in the tar to be exactly `disk.raw`.

Note that tar allows only a single `--transform` and we leave it
to the user to construct `sed` expressions if multiple renames
are required.
---
 stages/org.osbuild.tar           |  4 +++
 stages/org.osbuild.tar.meta.json |  4 +++
 stages/test/test_tar.py          | 46 ++++++++++++++++++++++++++++++++
 3 files changed, 54 insertions(+)

diff --git a/stages/org.osbuild.tar b/stages/org.osbuild.tar
index 40663aac0..beb97a997 100755
--- a/stages/org.osbuild.tar
+++ b/stages/org.osbuild.tar
@@ -33,6 +33,10 @@ def main(inputs, output_dir, options):
     if options.get("sparse", False):
         extra_args += ["--sparse"]
 
+    transform = options.get("transform")
+    if transform:
+        extra_args += ["--transform", transform]
+
     # Set up the tar command.
     tar_cmd = [
         "tar",
diff --git a/stages/org.osbuild.tar.meta.json b/stages/org.osbuild.tar.meta.json
index 0e0eaea53..bec34f472 100644
--- a/stages/org.osbuild.tar.meta.json
+++ b/stages/org.osbuild.tar.meta.json
@@ -98,6 +98,10 @@
           "description": "Make archive files sparse",
           "type": "boolean",
           "default": false
+        },
+        "transform": {
+          "type": "string",
+          "description": "Used to transform filenames and directly passed to --transform"
         }
       }
     },
diff --git a/stages/test/test_tar.py b/stages/test/test_tar.py
index 10db07abe..75fb0bd1b 100644
--- a/stages/test/test_tar.py
+++ b/stages/test/test_tar.py
@@ -19,11 +19,16 @@
         "root-node": "include",
         "paths": ["file1"]
     }, "{'filename': 'out.tar', 'root-node': 'include', 'paths': ['file1']} is not valid under any of the given schemas"),
+    ({
+        "filename": "foo",
+        "transform": ["transform-cannot-be-passed-multiple-times"],
+    }, " is not of type 'string'"),
     # good
     ({"filename": "out.tar", "root-node": "include"}, ""),
     ({"filename": "out.tar", "paths": ["file1"]}, ""),
     ({"filename": "out.tar", "sparse": True}, ""),
     ({"filename": "out.tar"}, ""),
+    ({"filename": "out.tar", "transform": "s/foo/bar"}, ""),
 ])
 def test_schema_validation_tar(stage_schema, test_data, expected_err):
     test_input = {
@@ -99,3 +104,44 @@ def test_tar_paths(tmp_path, stage_module, fake_inputs):
     assert os.path.exists(tar_path)
     output = subprocess.check_output(["tar", "-tf", tar_path], encoding="utf-8").split("\n")
     assert ["file2", "file1", "file1", "file2", ""] == output
+
+
+@pytest.mark.skipif(not has_executable("tar"), reason="no tar executable")
+@pytest.mark.parametrize("transform,expected_tar_output", [
+    # unrelated transform
+    ("s/foo/bar/", ["file1", "file2", ""]),
+    # one file
+    ("s/^file1$/foo99/", ["foo99", "file2", ""]),
+    # one file
+    ("s/file1/foo99/;s/file2/bar11/", ["foo99", "bar11", ""]),
+])
+def test_tar_transform(tmp_path, stage_module, fake_inputs, transform, expected_tar_output):
+    options = {
+        "filename": "out.tar",
+        "paths": [
+            "file1",
+            "file2",
+        ],
+        "transform": transform,
+    }
+    stage_module.main(fake_inputs, tmp_path, options)
+
+    tar_path = os.path.join(tmp_path, "out.tar")
+    assert os.path.exists(tar_path)
+    output = subprocess.check_output(["tar", "-tf", tar_path], encoding="utf-8").split("\n")
+    assert output == expected_tar_output
+
+
+def test_tar_transform_no_sh(tmp_path, stage_module, fake_inputs):
+    options = {
+        "filename": "out.tar",
+        "paths": [
+            "file1",
+        ],
+        # GNU sed allows to run shell commands with "/e"
+        # ensure here we donot allow this
+        "transform": "s/file1/date/e",
+    }
+    with pytest.raises(subprocess.CalledProcessError) as ex:
+        stage_module.main(fake_inputs, tmp_path, options)
+    assert "exit status 2" in str(ex.value)