Fix URL encoding issue in HttpDownloader to handle special characters

closes pulp#5686

CHANGES/5686.bugfix

@@ -0,0 +1 @@
+Implemented a more robust URL encoding mechanism that only encodes the path portion of the URL, ensuring that special characters are processed correctly without double encoding.

pulp_file/pytest_plugin.py

@@ -101,6 +101,16 @@ def basic_manifest_path(write_3_iso_file_fixture_data_factory):
     return write_3_iso_file_fixture_data_factory("basic")
 
 
+@pytest.fixture
+def encoded_manifest_path(file_fixtures_root):
+    file_fixtures_root.joinpath("encoded").mkdir()
+    file1 = generate_iso(file_fixtures_root.joinpath("encoded/long-name-%253a-encoded.iso"))
+    file2 = generate_iso(file_fixtures_root.joinpath("encoded/another-%25-encoded.iso"))
+    file3 = generate_iso(file_fixtures_root.joinpath("encoded/more-%3C-encoded.iso"))
+    generate_manifest(file_fixtures_root.joinpath("encoded/PULP_MANIFEST"), [file1, file2, file3])
+    return "/encoded/PULP_MANIFEST"
+
+
 @pytest.fixture
 def copy_manifest_only_factory(file_fixtures_root):
     def _copy_manifest_only(name):

pulp_file/tests/functional/api/test_sync.py

@@ -123,6 +123,19 @@ def test_duplicate_file_sync(
     assert file_repo.latest_version_href.endswith("/2/")
 
 
+@pytest.mark.parallel
+def test_encoded_file_name(file_repo, file_bindings, encoded_manifest_path, file_remote_factory, monitor_task):
+    remote = file_remote_factory(manifest_path=encoded_manifest_path, policy="immediate")
+    body = RepositorySyncURL(remote=remote.pulp_href)
+    monitor_task(file_bindings.RepositoriesFileApi.sync(file_repo.pulp_href, body).task)
+    file_repo = file_bindings.RepositoriesFileApi.read(file_repo.pulp_href)
+
+    version = file_bindings.RepositoriesFileVersionsApi.read(file_repo.latest_version_href)
+    assert version.content_summary.present["file.file"]["count"] == 3
+    assert version.content_summary.added["file.file"]["count"] == 3
+    assert file_repo.latest_version_href.endswith("/1/")
+
+
 @pytest.mark.parallel
 def test_filepath_includes_commas(
     file_bindings,

pulpcore/download/http.py

@@ -3,6 +3,7 @@
 import aiohttp
 import asyncio
 import backoff
+import urllib.parse
 
 from .base import BaseDownloader, DownloadResult
 from pulpcore.exceptions import (
@@ -49,6 +50,14 @@ def http_giveup_handler(exc):
     return False
 
 
+def encode_url(url):
+    """Helper function to encode only the path part of the URL."""
+    parsed_url = urllib.parse.urlparse(url)
+    encoded_path = urllib.parse.quote(parsed_url.path)
+    encoded_url = parsed_url._replace(path=encoded_path).geturl()
+    return encoded_url
+
+
 class HttpDownloader(BaseDownloader):
     """
     An HTTP/HTTPS Downloader built on `aiohttp`.
@@ -284,8 +293,9 @@ async def _run(self, extra_data=None):
         """
         if self.download_throttler:
             await self.download_throttler.acquire()
+        encoded_url = encode_url(self.url)
         async with self.session.get(
-            self.url, proxy=self.proxy, proxy_auth=self.proxy_auth, auth=self.auth
+            encoded_url, proxy=self.proxy, proxy_auth=self.proxy_auth, auth=self.auth
         ) as response:
             self.raise_for_status(response)
             to_return = await self._handle_response(response)