Refine support for GitHub /archive/refs/tags/ URLs in url2purl (#161)

Signed-off-by: tdruez <tdruez@nexb.com>

Author: tdruez

CHANGELOG.rst

@@ -4,6 +4,10 @@ Changelog
 0.16.0 (unreleased)
 -------------------
 
+- Refine support for GitHub /archive/refs/tags/ URLs in ``url2purl``.
+  The whole tag is now captured as the version.
+  This allows to properly reconstruct valid URLs in ``purl2url``.
+
 0.15.5 (2024-07-24)
 -------------------
 

src/packageurl/contrib/url2purl.py

@@ -502,6 +502,13 @@ def build_github_purl(url):
     Return a PackageURL object from GitHub `url`.
     """
 
+    # https://github.com/apache/nifi/archive/refs/tags/rel/nifi-2.0.0-M3.tar.gz
+    archive_tags_pattern = (
+        r"https?://github.com/(?P<namespace>.+)/(?P<name>.+)"
+        r"/archive/refs/tags/"
+        r"(?P<version>.+).(zip|tar.gz|tar.bz2|.tgz)"
+    )
+
     # https://github.com/nexB/scancode-toolkit/archive/v3.1.1.zip
     archive_pattern = (
         r"https?://github.com/(?P<namespace>.+)/(?P<name>.+)"
@@ -538,6 +545,7 @@ def build_github_purl(url):
     git_pattern = r"https?://github.com/(?P<namespace>.+)/(?P<name>.+).(git)"
 
     patterns = (
+        archive_tags_pattern,
         archive_pattern,
         raw_pattern,
         blob_pattern,

tests/contrib/data/url2purl.json

@@ -193,6 +193,7 @@
   "https://github.com/TG1999/fetchcode/fetchcode/src": "pkg:github/tg1999/fetchcode@fetchcode#src",
   "https://github.com/NEXB/SCANCODE-TOOLKIT/tree/develop/PLUGINS/scancode-ctags-macosx_10_9_intel": "pkg:github/nexb/scancode-toolkit@develop#PLUGINS/scancode-ctags-macosx_10_9_intel",
   "https://github.com/NEXB/SCANCODE-TOOLKIT/tree/DEVELOP/PLUGINS/scancode-ctags-macosx_10_9_intel": "pkg:github/nexb/scancode-toolkit@DEVELOP#PLUGINS/scancode-ctags-macosx_10_9_intel",
+  "https://github.com/apache/nifi/archive/refs/tags/rel/nifi-2.0.0-M3.tar.gz": "pkg:github/apache/nifi@rel/nifi-2.0.0-M3",
   "https://raw.githubusercontent.com/volatilityfoundation/dwarf2json/master/LICENSE.txt": "pkg:github/volatilityfoundation/dwarf2json@master#LICENSE.txt",
   "https://raw.githubusercontent.com/LeZuse/flex-sdk/master/frameworks/projects/mx/src/mx/containers/accordionClasses/AccordionHeader.as": "pkg:github/lezuse/flex-sdk@master#frameworks/projects/mx/src/mx/containers/accordionClasses/AccordionHeader.as",
   "https://raw.githubusercontent.com/NCIP/lexevs/master/lgSharedLibraries/jettison/jettison-1.1.jar": "pkg:github/ncip/lexevs@master#lgSharedLibraries/jettison/jettison-1.1.jar",

tests/contrib/test_url2purl.py

@@ -61,7 +61,7 @@ def python_safe(s):
 
 def get_url2purl_test_method(test_url, expected_purl):
     def test_method(self):
-        self.assertEqual(expected_purl, get_purl(test_url))
+        self.assertEqual(expected_purl, get_purl(test_url), msg=test_url)
 
     return test_method