From f2afa2f054ba9e6c5d142e00233f0073925e7893 Mon Sep 17 00:00:00 2001
From: Sviatoslav Sydorenko <wk@sydorenko.org.ua>
Date: Thu, 25 Feb 2021 18:02:07 +0100
Subject: [PATCH] Fix how pure-Python HTTP parser interprets `//`

---
 CHANGES/5498.bugfix       |  6 ++++++
 aiohttp/http_parser.py    | 14 +++++++++++++-
 tests/test_http_parser.py |  1 +
 3 files changed, 20 insertions(+), 1 deletion(-)
 create mode 100644 CHANGES/5498.bugfix

diff --git a/CHANGES/5498.bugfix b/CHANGES/5498.bugfix
new file mode 100644
index 00000000000..c11630e8743
--- /dev/null
+++ b/CHANGES/5498.bugfix
@@ -0,0 +1,6 @@
+Fix interpretation difference of the pure-Python and the Cython-based
+HTTP parsers construct a ``yarl.URL`` object for HTTP request-target.
+
+Before this fix, the Python parser would turn the URI's absolute-path
+for ``//some-path`` into ``/`` while the Cython code preserved it as
+``//some-path``. Now, both do the latter.
diff --git a/aiohttp/http_parser.py b/aiohttp/http_parser.py
index 9487378da05..1045b6c0926 100644
--- a/aiohttp/http_parser.py
+++ b/aiohttp/http_parser.py
@@ -522,6 +522,9 @@ def parse_message(self, lines: List[bytes]) -> RawRequestMessage:
                 "Status line is too long", str(self.max_line_size), str(len(path))
             )
 
+        path_part, _hash_separator, url_fragment = path.partition("#")
+        path_part, _question_mark_separator, qs_part = path_part.partition("?")
+
         # method
         if not METHRE.match(method):
             raise BadStatusLine(method)
@@ -562,7 +565,16 @@ def parse_message(self, lines: List[bytes]) -> RawRequestMessage:
             compression,
             upgrade,
             chunked,
-            URL(path),
+            # NOTE: `yarl.URL.build()` is used to mimic what the Cython-based
+            # NOTE: parser does, otherwise it results into the same
+            # NOTE: HTTP Request-Line input producing different
+            # NOTE: `yarl.URL()` objects
+            URL.build(
+                path=path_part,
+                query_string=qs_part,
+                fragment=url_fragment,
+                encoded=True,
+            ),
         )
 
 
diff --git a/tests/test_http_parser.py b/tests/test_http_parser.py
index 77c62addf6b..172d7bc30cf 100644
--- a/tests/test_http_parser.py
+++ b/tests/test_http_parser.py
@@ -530,6 +530,7 @@ def test_http_request_parser_two_slashes(parser: Any) -> None:
 
     assert msg.method == "GET"
     assert msg.path == "//path"
+    assert msg.url.path == "//path"
     assert msg.version == (1, 1)
     assert not msg.should_close
     assert msg.compression is None