Fixes #1063: Use github purl, repo and version for a github release archive in SBOM

CHANGELOG.rst

@@ -1,3 +1,8 @@
+Release 0.14.0 (unreleased)
+===========================
+
+* Use github purl, repo and version for a github release archive in SBOM (#1063)
+
 Release 0.13.0 (released 2026-03-30)
 ====================================
 

dfetch/reporting/sbom_reporter.py

@@ -206,6 +206,7 @@ def add_project(
         else:
             purl = vcs_url_to_purl(project.remote_url, version=version, subpath=subpath)
         name = project.name if purl.type == "generic" else purl.name
+        version = purl.version or version
         location = self.manifest.find_name_in_manifest(project.name)
         component = Component(
             name=name,

dfetch/vcs/archive.py

@@ -27,6 +27,7 @@
 import http.client
 import os
 import pathlib
+import re
 import shutil
 import stat
 import sys
@@ -47,6 +48,7 @@
     copy_src_subset,
     prune_files_by_pattern,
 )
+from dfetch.util.versions import coerce
 
 logger = get_logger(__name__)
 
@@ -82,7 +84,21 @@ def archive_url_to_purl(
     version: str | None = None,
     subpath: str | None = None,
 ) -> PackageURL:
-    """Build a generic PackageURL for an archive download URL."""
+    """Build a github or generic PackageURL for an archive download URL."""
+    if match := re.search(
+        r"https://github\.com/(?P<org>[^/]+)/(?P<repo>[^/]+)/releases/download/(?P<version>[^/]+)/",
+        download_url,
+    ):
+        prefix, current_version, _ = coerce(
+            match["version"],
+        )
+        return PackageURL(
+            type="github",
+            namespace=match["org"].lower(),
+            name=match["repo"].lower(),
+            version=str(current_version) if current_version else prefix,
+        )
+
     parsed = urllib.parse.urlparse(download_url)
     basename = os.path.basename(parsed.path)
     name = strip_archive_extension(basename) or "unknown"

features/report-sbom-archive.feature

@@ -116,3 +116,39 @@ Feature: Create a CycloneDX SBOM for archive dependencies
                 ]
             }
             """
+
+    Scenario: An github release archive uses github purl and repo
+        Given the manifest 'dfetch.yaml'
+            """
+            manifest:
+                version: '0.1'
+
+                remotes:
+                - name: github
+                  url-base: https://github.com/
+
+                projects:
+                - name: TF-PSA-Crypto
+                  vcs: archive
+                  remote: github
+                  dst: ext/TF-PSA-Crypto
+                  repo-path: Mbed-TLS/TF-PSA-Crypto/releases/download/tf-psa-crypto-1.0.0/tf-psa-crypto-1.0.0.tar.bz2
+                  integrity:
+                      hash: sha256:31f0df2ca17897b5db2757cb0307dcde267292ba21ade831663d972a7a5b7d40
+            """
+        And all projects are updated
+        When I run "dfetch report -t sbom"
+        Then the 'report.json' json file includes
+            """
+            {
+                "components": [
+                    {
+                        "group": "mbed-tls",
+                        "name": "tf-psa-crypto",
+                        "purl": "pkg:github/mbed-tls/tf-psa-crypto@1.0.0",
+                        "type": "library",
+                        "version": "1.0.0"
+                    }
+                ]
+            }
+            """