golang: Ensure to case-encode URLs for retrieving package info

When a go package name contains uppercase characters in it, associated
goproxy URLs need to be case-encoded by replacing every uppercase letter
with an exclamation mark followed by the corresponding lower-case letter.

This fixes the loading of such packages.

See https://go.dev/ref/mod#goproxy-protocol.

Related to T4124

swh/loader/package/golang/loader.py

@@ -5,6 +5,7 @@
 
 import json
 import logging
+import re
 from typing import Iterator, Optional, Sequence, Tuple
 
 import attr
@@ -17,6 +18,10 @@ from swh.storage.interface import StorageInterface
 logger = logging.getLogger(__name__)
 
 
+def _uppercase_encode(url: str) -> str:
+    return re.sub("([A-Z]{1})", r"!\1", url).lower()
+
+
 @attr.s
 class GolangPackageInfo(BasePackageInfo):
     name = attr.ib(type=str)
@@ -46,6 +51,7 @@ class GolangLoader(PackageLoader[GolangPackageInfo]):
         ), "Go package URL (%s) not from %s" % (url, self.GOLANG_PKG_DEV_URL)
         self.name = url[len(self.GOLANG_PKG_DEV_URL) + 1 :]
         self.url = url.replace(self.GOLANG_PKG_DEV_URL, self.GOLANG_PROXY_URL)
+        self.url = _uppercase_encode(self.url)
 
     def get_versions(self) -> Sequence[str]:
         return api_info(f"{self.url}/@v/list").decode().splitlines()
@@ -55,7 +61,7 @@ class GolangLoader(PackageLoader[GolangPackageInfo]):
         return json.loads(latest)["Version"]
 
     def _raw_info(self, version: str) -> dict:
-        url = f"{self.url}/@v/{version}.info"
+        url = f"{self.url}/@v/{_uppercase_encode(version)}.info"
         return json.loads(api_info(url))
 
     def get_package_info(self, version: str) -> Iterator[Tuple[str, GolangPackageInfo]]:

swh/loader/package/golang/tests/data/https_proxy.golang.org/github.com_adam-hanna_array!operations_@latest

+{"Version":"v1.0.1","Time":"2022-03-23T18:02:43Z"}
\ No newline at end of file

swh/loader/package/golang/tests/data/https_proxy.golang.org/github.com_adam-hanna_array!operations_@v_list

+v1.0.1

swh/loader/package/golang/tests/data/https_proxy.golang.org/github.com_adam-hanna_array!operations_@v_v1.0.1.info

+{"Version":"v1.0.1","Time":"2022-03-23T18:02:43Z"}
\ No newline at end of file

swh/loader/package/golang/tests/test_golang.py

@@ -11,3 +11,12 @@ def test_golang_loader_first_visit(swh_storage, requests_mock_datadir):
     loader = GolangLoader(swh_storage, url)
 
     assert loader.load()["status"] == "eventful"
+
+
+def test_golang_loader_package_name_with_uppercase_characters(
+    swh_storage, requests_mock_datadir
+):
+    url = "https://pkg.go.dev/github.com/adam-hanna/arrayOperations"
+    loader = GolangLoader(swh_storage, url)
+
+    assert loader.load()["status"] == "eventful"

[Phabricator Migration user]

Some references in the commit message have been migrated:

• T4124 is now swh/meta#4124 (closed)