Use PEP 503 rules to validate upload filename

Author: uranusjr

tests/unit/forklift/test_legacy.py

@@ -20,7 +20,6 @@
 from cgi import FieldStorage
 from unittest import mock
 
-import pkg_resources
 import pretend
 import pytest
 import requests
@@ -2388,7 +2387,16 @@ def test_upload_fails_with_diff_filename_same_blake2(
             "400 File already exists. See /the/help/url/ for more information."
         )
 
-    def test_upload_fails_with_wrong_filename(self, pyramid_config, db_request):
+    @pytest.mark.parametrize(
+        "filename",
+        [
+            "nope-{}.tar.gz",
+            "nope-{}-py3-none-any.whl",
+        ],
+    )
+    def test_upload_fails_with_wrong_filename(
+        self, pyramid_config, db_request, filename
+    ):
         pyramid_config.testing_securitypolicy(userid=1)
 
         user = UserFactory.create()
@@ -2398,7 +2406,7 @@ def test_upload_fails_with_wrong_filename(self, pyramid_config, db_request):
         release = ReleaseFactory.create(project=project, version="1.0")
         RoleFactory.create(user=user, project=project)
 
-        filename = "nope-{}.tar.gz".format(release.version)
+        filename = filename.format(release.version)
 
         db_request.POST = MultiDict(
             {
@@ -2422,8 +2430,8 @@ def test_upload_fails_with_wrong_filename(self, pyramid_config, db_request):
 
         assert resp.status_code == 400
         assert resp.status == (
-            "400 Start filename for {!r} with {!r}.".format(
-                project.name, pkg_resources.safe_name(project.name).lower()
+            "400 Filename {!r} must match project {!r}.".format(
+                filename, project.normalized_name
             )
         )
 

warehouse/forklift/legacy.py

@@ -26,7 +26,6 @@
 import packaging.specifiers
 import packaging.utils
 import packaging.version
-import pkg_resources
 import requests
 import stdlib_list
 import wtforms
@@ -626,6 +625,14 @@ def full_validate(self):
             )
 
 
+def _is_valid_filename(filename, specified_canonical_name):
+    if filename.endswith(".whl"):
+        parsed_canonical_name = packaging.utils.parse_wheel_filename(filename)[0]
+    else:
+        parsed_canonical_name = packaging.utils.parse_sdist_filename(filename)[0]
+    return parsed_canonical_name == specified_canonical_name
+
+
 _safe_zipnames = re.compile(r"(purelib|platlib|headers|scripts|data).+", re.I)
 # .tar uncompressed, .tar.gz .tgz, .tar.bz2 .tbz2
 _tar_filenames_re = re.compile(r"\.(?:tar$|t(?:ar\.)?(?P<z_type>gz|bz2)$)")
@@ -1192,11 +1199,11 @@ def file_upload(request):
 
     # Make sure that our filename matches the project that it is being uploaded
     # to.
-    prefix = pkg_resources.safe_name(project.name).lower()
-    if not pkg_resources.safe_name(filename).lower().startswith(prefix):
+    canonical_name = project.normalized_name
+    if not _is_valid_filename(filename, canonical_name):
         raise _exc_with_message(
             HTTPBadRequest,
-            "Start filename for {!r} with {!r}.".format(project.name, prefix),
+            "Filename {!r} must match project {!r}.".format(filename, canonical_name),
         )
 
     # Check the content type of what is being uploaded