Improve support for SourceForge URLs in url2purl #139

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

Author: tdruez

CHANGELOG.rst

@@ -1,6 +1,12 @@
 Changelog
 =========
 
+0.13.2 (2024-01-04)
+-------------------
+
+- Improve support for SourceForge URLs in `url2purl`.
+  https://github.com/package-url/packageurl-python/issues/139
+
 0.13.1 (2023-12-11)
 -------------------
 

setup.cfg

@@ -1,6 +1,6 @@
 [metadata]
 name = packageurl-python
-version = 0.13.1
+version = 0.13.2
 license = MIT
 description = A purl aka. Package URL parser and builder
 long_description = file:README.rst

src/packageurl/contrib/url2purl.py

@@ -344,7 +344,7 @@ def build_pypi_purl(uri):
 register_pattern("nuget", nuget_api_pattern)
 
 
-@purl_router.route("https?://.*sourceforge.net/project/.*")
+@purl_router.route("https?://.*sourceforge.net/projects?/.*")
 def build_sourceforge_purl(uri):
     # We use a more general route pattern instead of using `sourceforge_pattern`
     # below by itself because we want to capture all sourceforge download URLs,
@@ -353,9 +353,11 @@ def build_sourceforge_purl(uri):
     # URL that we can't handle.
 
     # http://master.dl.sourceforge.net/project/libpng/zlib/1.2.3/zlib-1.2.3.tar.bz2
+    # https://sourceforge.net/projects/scribus/files/scribus/1.6.0/scribus-1.6.0.tar.gz/download
     sourceforge_pattern = (
-        r"^https?://.*sourceforge.net/project/"
+        r"^https?://.*sourceforge.net/projects?/"
         r"(?P<namespace>([^/]+))/"  # do not allow more "/" segments
+        r"(files/)?"  # optional segment for "*/download" type URLs
         r"(?P<name>.+)/"
         r"(?P<version>[0-9\.]+)/"  # version restricted to digits and dots
         r"(?P=name)-(?P=version).*"  # {name}-{version} repeated in the filename

tests/contrib/data/url2purl.json

@@ -171,6 +171,7 @@
   "http://master.dl.sourceforge.net/project/urlchecker/lu/ng/urlchecker/urlchecker/1.7/urlchecker-1.7-javadoc.jar": "pkg:sourceforge/urlchecker?download_url=http://master.dl.sourceforge.net/project/urlchecker/lu/ng/urlchecker/urlchecker/1.7/urlchecker-1.7-javadoc.jar",
   "http://master.dl.sourceforge.net/project/zclasspath/maven2/org/zclasspath/zclasspath/1.5/zclasspath-1.5.jar": "pkg:sourceforge/zclasspath?download_url=http://master.dl.sourceforge.net/project/zclasspath/maven2/org/zclasspath/zclasspath/1.5/zclasspath-1.5.jar",
   "http://master.dl.sourceforge.net/project/googleimagedown/project/v1.1/GoogleImageDownloader-v1.1-src.tar.bz2": "pkg:sourceforge/googleimagedown?download_url=http://master.dl.sourceforge.net/project/googleimagedown/project/v1.1/GoogleImageDownloader-v1.1-src.tar.bz2",
+  "https://sourceforge.net/projects/scribus/files/scribus/1.6.0/scribus-1.6.0.tar.gz/download": "pkg:sourceforge/scribus/scribus@1.6.0",
   "https://crates.io/api/v1/crates/rand/0.7.2/download": "pkg:cargo/rand@0.7.2",
   "https://crates.io/api/v1/crates/clap/2.33.0/download": "pkg:cargo/clap@2.33.0",
   "https://crates.io/api/v1/crates/structopt/0.3.11/download": "pkg:cargo/structopt@0.3.11",