RomanNum3ral
/
scrapy
mirror of https://github.com/scrapy/scrapy.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

Fix the substitution pattern in open_in_browser(). (#7459) 

* Fix the substitution pattern in open_in_browser().

* Formatting.
This commit is contained in:
Andrey Rakhmatullin 2026-04-24 13:54:37 +05:00 committed by GitHub
parent 4e25686b20
commit fc4c57e795
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
2 changed files with 39 additions and 45 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
scrapy/utils/response.py
View File

@ -99,7 +99,7 @@ def open_in_browser(
if isinstance(response, HtmlResponse): if isinstance(response, HtmlResponse):
if b"<base" not in body: if b"<base" not in body:
_remove_html_comments(body) _remove_html_comments(body)
repl = rf'\0<base href="{response.url}">' repl = rf'\g<0><base href="{response.url}">'
body = re.sub(rb"<head(?:[^<>]*?>)", to_bytes(repl), body, count=1) body = re.sub(rb"<head(?:[^<>]*?>)", to_bytes(repl), body, count=1)
ext = ".html" ext = ".html"
elif isinstance(response, TextResponse): elif isinstance(response, TextResponse):

82
tests/test_utils_response.py
View File

@ -111,36 +111,27 @@ def test_response_status_message():
assert response_status_message(573) == "573 Unknown Status" assert response_status_message(573) == "573 Unknown Status"
def test_inject_base_url(): @pytest.mark.parametrize(
url = "http://www.example.com" "body",
[
def check_base_url(burl): pytest.param(
path = urlparse(burl).path b"""
if not path or not Path(path).exists():
path = burl.replace("file://", "")
bbody = Path(path).read_bytes()
assert bbody.count(b'<base href="' + to_bytes(url) + b'">') == 1
return True
r1 = HtmlResponse(
url,
body=b"""
<html> <html>
<head><title>Dummy</title></head> <head><title>Dummy</title></head>
<body><p>Hello world.</p></body> <body><p>Hello world.</p></body>
</html>""", </html>""",
) id="Simple",
r2 = HtmlResponse( ),
url, pytest.param(
body=b""" b"""
<html> <html>
<head id="foo"><title>Dummy</title></head> <head id="foo"><title>Dummy</title></head>
<body>Hello world.</body> <body>Hello world.</body>
</html>""", </html>""",
) id="<head> with attrs",
r3 = HtmlResponse( ),
url, pytest.param(
body=b""" b"""
<html> <html>
<head><title>Dummy</title></head> <head><title>Dummy</title></head>
<body> <body>
@ -148,19 +139,19 @@ def test_inject_base_url():
<p>Hello world.</p> <p>Hello world.</p>
</body> </body>
</html>""", </html>""",
) id="Misleading tag",
r4 = HtmlResponse( ),
url, pytest.param(
body=b""" b"""
<html> <html>
<!-- <head>Dummy comment</head> --> <!-- <head>Dummy comment</head> -->
<head><title>Dummy</title></head> <head><title>Dummy</title></head>
<body><p>Hello world.</p></body> <body><p>Hello world.</p></body>
</html>""", </html>""",
) id="Misleading comment",
r5 = HtmlResponse( ),
url, pytest.param(
body=b""" b"""
<html> <html>
<!--[if IE]> <!--[if IE]>
<head><title>IE head</title></head> <head><title>IE head</title></head>
@ -170,21 +161,24 @@ def test_inject_base_url():
<!--<![endif]--> <!--<![endif]-->
<body><p>Hello world.</p></body> <body><p>Hello world.</p></body>
</html>""", </html>""",
) id="Conditional comment",
),
],
)
def test_inject_base_url(body: bytes) -> None:
url = "http://www.example.com"
assert open_in_browser(r1, _openfunc=check_base_url), "Inject base url" def check_base_url(burl):
assert open_in_browser(r2, _openfunc=check_base_url), ( path = urlparse(burl).path
"Inject base url with argumented head" if not path or not Path(path).exists():
) path = burl.replace("file://", "")
assert open_in_browser(r3, _openfunc=check_base_url), ( bbody = Path(path).read_bytes()
"Inject unique base url with misleading tag" assert bbody.count(b'><base href="' + to_bytes(url) + b'">') == 1
) assert b"<head" in bbody
assert open_in_browser(r4, _openfunc=check_base_url), ( return True
"Inject unique base url with misleading comment"
) resp = HtmlResponse(url, body=body)
assert open_in_browser(r5, _openfunc=check_base_url), ( assert open_in_browser(resp, _openfunc=check_base_url)
"Inject unique base url with conditional comment"
)
def test_open_in_browser_redos_comment(): def test_open_in_browser_redos_comment():