Fix if-range handling and its unit tests

The value used for conditional range requests should be updated when the
download is restarted (not *resumed*). The unit tests testing
conditional range requests were broken, using code 206 for the
non-matching responses. This violates the HTTP spec and fools the
resumption logic into thinking the conditional range request was
fulfilled (when it wasn't!).

The test cases were fixed to use code 200 as expected and extended
to ensure the content key is updated across a chain of non-matching
repsonses (which is borderline impossible, but hey it did catch my
eye so why not fix it).

Author: ichard26

src/pip/_internal/network/download.py

@@ -216,7 +216,7 @@ def _process_response(self, download: _FileDownload, resp: Response) -> None:
             logger.warning("Connection timed out while downloading.")
 
     def _attempt_resumes_or_redownloads(
-        self, download: _FileDownload, resp: Response
+        self, download: _FileDownload, first_resp: Response
     ) -> None:
         """Attempt to resume/restart the download if connection was dropped."""
 
@@ -231,14 +231,15 @@ def _attempt_resumes_or_redownloads(
             )
 
             try:
-                resume_resp = self._http_get_resume(download, should_match=resp)
+                resume_resp = self._http_get_resume(download, should_match=first_resp)
                 # Fallback: if the server responded with 200 (i.e., the file has
                 # since been modified or range requests are unsupported) or any
                 # other unexpected status, restart the download from the beginning.
                 must_restart = resume_resp.status_code != HTTPStatus.PARTIAL_CONTENT
                 if must_restart:
                     download.reset_file()
                     download.size = _get_http_response_size(resume_resp)
+                    first_resp = resume_resp
 
                 self._process_response(download, resume_resp)
             except (ConnectionError, ReadTimeoutError, OSError):

tests/unit/test_network_download.py

@@ -229,9 +229,9 @@ def test_parse_content_disposition(
             [(24, None), (36, None)],
             b"new-0cfa7e9d-1868-4dd7-9fb3-f2561d5dfd89",
         ),
-        # The downloader should fail after n resume_retries attempts.
+        # The downloader should fail after N resume_retries attempts.
         # This prevents the downloader from getting stuck if the connection
-        # is unstable and the server doesn't not support range requests.
+        # is unstable and the server does NOT support range requests.
         (
             1,
             [
@@ -241,7 +241,7 @@ def test_parse_content_disposition(
             [(24, None)],
             None,
         ),
-        # The downloader should use If-Range header to make the range
+        # The downloader should use the If-Range header to make the range
         # request conditional if it is possible to check for modifications
         # (e.g. if we know the creation time of the initial response).
         (
@@ -255,17 +255,28 @@ def test_parse_content_disposition(
                     200,
                     b"0cfa7e9d-1868-4dd7-9fb3-",
                 ),
+                (
+                    {
+                        "content-length": "42",
+                        "last-modified": "Wed, 21 Oct 2015 07:30:00 GMT",
+                    },
+                    200,
+                    b"new-0cfa7e9d-1868-4dd7-9fb3-f2561d5dfd89",
+                ),
                 (
                     {
                         "content-length": "12",
                         "last-modified": "Wed, 21 Oct 2015 07:54:00 GMT",
                     },
-                    206,
+                    200,
                     b"f2561d5dfd89",
                 ),
             ],
-            [(24, "Wed, 21 Oct 2015 07:28:00 GMT")],
-            b"0cfa7e9d-1868-4dd7-9fb3-f2561d5dfd89",
+            [
+                (24, "Wed, 21 Oct 2015 07:28:00 GMT"),
+                (40, "Wed, 21 Oct 2015 07:30:00 GMT"),
+            ],
+            b"f2561d5dfd89",
         ),
         # ETag is preferred over Last-Modified for the If-Range condition.
         (
@@ -286,12 +297,12 @@ def test_parse_content_disposition(
                         "last-modified": "Wed, 21 Oct 2015 07:54:00 GMT",
                         "etag": '"33a64df551425fcc55e4d42a148795d9f25f89d4"',
                     },
-                    206,
+                    200,
                     b"f2561d5dfd89",
                 ),
             ],
             [(24, '"33a64df551425fcc55e4d42a148795d9f25f89d4"')],
-            b"0cfa7e9d-1868-4dd7-9fb3-f2561d5dfd89",
+            b"f2561d5dfd89",
         ),
     ],
 )