commit 9049214d1c4b96524e251d65e06da6759c110f12
from: Antoine Lambert <anlambert@softwareheritage.org>
date: Wed Jul 05 11:31:04 2023 UTC

client: Fix content type check in AbstractHttpGitClient._smart_request

Some git servers can send a Content-Type header containing a charset
directive. Official git client can successfully clone repositories
hosted on such server but dulwich was raising a GitProtocolError as
the code to check the header value was not expecting a directive to
be present in it.

commit - c1c52f895822136ee952226e2ec1bc04d9f794f5
commit + 9049214d1c4b96524e251d65e06da6759c110f12
blob - 9c88a839a5f5c2a7b3f05de27ee04569f1ba59c0
blob + 857de2cde4eff29c68a2e9525413c97a06051ffd
--- dulwich/client.py
+++ dulwich/client.py
@@ -2007,7 +2007,7 @@ class AbstractHttpGitClient(GitClient):
         if isinstance(data, bytes):
             headers["Content-Length"] = str(len(data))
         resp, read = self._http_request(url, headers, data)
-        if resp.content_type != result_content_type:
+        if not resp.content_type.startswith(result_content_type):
             raise GitProtocolError(
                 "Invalid content-type from server: %s" % resp.content_type
             )
blob - b6df9ff105208b53328469be5a0300b71f73c885
blob + a50e46458e3f145a0786fbe269a83ff49620f67e
--- dulwich/tests/test_client.py
+++ dulwich/tests/test_client.py
@@ -1144,7 +1144,31 @@ class HttpGitClientTests(TestCase):
             else:
                 # check also the no redirection case
                 self.assertEqual(processed_url, base_url)
+
+    def test_smart_request_content_type_with_directive_check(self):
+        from urllib3.response import HTTPResponse
+
+        # we need to mock urllib3.PoolManager as this test will fail
+        # otherwise without an active internet connection
+        class PoolManagerMock:
+            def __init__(self) -> None:
+                self.headers: Dict[str, str] = {}
+
+            def request(self, method, url, fields=None, headers=None, redirect=True, preload_content=True):
+                return HTTPResponse(
+                    headers={
+                        "Content-Type": "application/x-git-upload-pack-result; charset=utf-8"
+                    },
+                    request_method=method,
+                    request_url=url,
+                    preload_content=preload_content,
+                    status=200,
+                )
 
+        clone_url = "https://hacktivis.me/git/blog.git/"
+        client = HttpGitClient(clone_url, pool_manager=PoolManagerMock(), config=None)
+        self.assertTrue(client._smart_request("git-upload-pack", clone_url, data=None))
+
 
 class TCPGitClientTests(TestCase):
     def test_get_url(self):