Files
email-unsubscribe-links-scr…/tests/test_app.py
T
Rsclub2_2 734e4f0320 Abmelde-Link auch finden, wenn nur "hier" verlinkt ist
Hat ein Link einen kurzen Text ohne Stichwort, entscheidet der Satz, in
dem er steht (bis Punkt, Zeilenumbruch oder Rand des Absatzes). Text
anderer Links und ausgeschriebene Adressen zählen nicht zum Satz.
2026-11-16 09:25:56 +01:00

161 lines
7.3 KiB
Python
Raw Blame History

This file contains ambiguous Unicode characters
This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.
"""Tests for link extraction, link safety and the trash actions (fake IMAP server)."""
from email.message import EmailMessage
import pytest
import app as unsub
def _mail(html: str, subject: str = "Newsletter") -> bytes:
msg = EmailMessage()
msg["Subject"] = subject
msg["From"] = "news@example.org"
msg.set_content("plain")
msg.add_alternative(html, subtype="html")
return msg.as_bytes()
def test_finds_anchor_and_plain_links_without_duplicates():
html = (
'<a href="https://ex.org/u?id=1">Abmelden</a> <a href="https://ex.org/home">Home</a>'
" https://ex.org/unsubscribe/2 https://ex.org/unsubscribe/2"
)
assert unsub.find_unsubscribe_links(html) == ["https://ex.org/u?id=1", "https://ex.org/unsubscribe/2"]
def test_escaped_ampersand_in_an_href_does_not_give_a_second_mangled_link():
html = ('<a href="https://ex.org/unsubscribe?u=1&amp;t=abc">Abmelden</a>'
' <a href="https://ex.org/unsubscribe-all?u=1">Von allen abmelden</a>')
assert unsub.find_unsubscribe_links(html) == ["https://ex.org/unsubscribe?u=1&t=abc",
"https://ex.org/unsubscribe-all?u=1"]
def test_plain_text_link_keeps_its_place_behind_an_escaped_duplicate():
html = ('<a href="https://ex.org/unsubscribe?u=1&amp;t=abc">Abmelden</a>'
' oder hier: https://ex.org/unsubscribe/2.')
assert unsub.find_unsubscribe_links(html) == ["https://ex.org/unsubscribe?u=1&t=abc",
"https://ex.org/unsubscribe/2"]
@pytest.mark.parametrize("tail", [".", ",", ";", ":", "!", "?", ")", "”", "’", "»", ").", "!”"])
def test_punctuation_after_a_plain_text_link_is_not_part_of_it(tail):
text = "Abmelden (siehe https://ex.org/unsubscribe/2" + tail + " danach)"
assert unsub.find_unsubscribe_links(text) == ["https://ex.org/unsubscribe/2"]
@pytest.mark.parametrize("href", ["javascript:alert(1)", "data:text/html,<script>x</script>", "JavaScript:x"])
def test_dangerous_links_are_dropped(href):
assert unsub.find_unsubscribe_links(f'<a href="{href}">Unsubscribe</a>') == []
def test_encoded_subject_is_decoded():
msg = EmailMessage()
msg["Subject"] = "Grüße vom Newsletter"
parsed = unsub.email.message_from_bytes(msg.as_bytes())
assert unsub.get_subject(parsed) == "Grüße vom Newsletter"
class FakeIMAP:
def __init__(self, mails: dict[bytes, bytes]):
self.mails = mails
self.calls = []
def uid(self, command, *args):
self.calls.append((command, args))
if command == "search":
return "OK", [b" ".join(self.mails)]
if command == "fetch":
return "OK", [(b"1 (BODY[] {n}", self.mails[args[0]]), b")"]
return "OK", [b""]
def expunge(self):
self.calls.append(("expunge", ()))
def logout(self):
pass
def test_collect_links_keys_by_uid():
fake = FakeIMAP({b"41": _mail('<a href="https://ex.org/u">unsubscribe</a>'), b"42": _mail("<p>no links</p>")})
result = unsub.collect_links(fake)
assert list(result) == ["41"] and result["41"]["links"] == ["https://ex.org/u"]
def test_move_to_trash_uses_uids_and_ignores_garbage(monkeypatch):
fake = FakeIMAP({})
monkeypatch.setattr(unsub, "connect", lambda: fake)
assert unsub.move_emails_to_trash(["7", "1:*", "8 OR ALL"]) == 1
assert ("copy", ("7", "Trash")) in fake.calls
assert ("store", ("7", "+FLAGS", "(\\Deleted)")) in fake.calls
assert not any(c[0] in ("copy", "store") and c[1][0] != "7" for c in fake.calls)
def test_move_all_form_posts_every_uid(monkeypatch):
moved = []
monkeypatch.setattr(unsub, "move_emails_to_trash", lambda uids: moved.extend(uids) or len(uids))
client = unsub.app.test_client()
assert client.post("/move-all-to-trash", data={"email_id": ["41", "42"]}).status_code == 302
assert moved == ["41", "42"]
def test_result_page_renders_hidden_ids_for_move_all(monkeypatch):
monkeypatch.setattr(unsub, "connect", lambda: FakeIMAP({b"41": _mail('<a href="https://ex.org/u">Abmelden</a>')}))
html = unsub.app.test_client().get("/result").get_data(as_text=True)
assert html.count('name="email_id" value="41"') == 2 # single form + move-all form
assert 'rel="noopener noreferrer"' in html
# --- link named "hier": the word that says "unsubscribe" stands next to the link, in the same sentence ---
LIEFERANDO = (
'<table><tr><td>Wir schicken Dir diese E-Mail, weil Du mit der E-Mail-Adresse '
'<a href="mailto:philipp@example.org">philipp@example.org</a> unseren Newsletter abonniert hast.<br/>'
'Wenn Du keinen Newsletter von uns mehr bekommen möchtest, kannst Du ihn '
'<a href="https://click.example/u/1"><u>hier</u></a> abbestellen.<br/><br/>'
'Lieferando ist ein Angebot von Takeaway.com<br/>'
'E-Mail <a href="https://click.example/view/2"><u>online</u></a> ansehen.</td></tr></table>'
)
def test_link_named_hier_counts_when_its_sentence_says_unsubscribe():
assert unsub.find_unsubscribe_links(LIEFERANDO) == ["https://click.example/u/1"]
@pytest.mark.parametrize("html", [
'To unsubscribe, click <a href="https://click.example/u/1">here</a>.',
'<p>Du kannst dich <a href="https://click.example/u/1"><span>hier</span></a> <b>abmelden</b>.</p>',
'<div>Newsletter <a href="https://click.example/u/1">hier</a>\n abbestellen</div>',
])
def test_sentence_around_a_short_link_is_read_across_inline_tags(html):
assert unsub.find_unsubscribe_links(html) == ["https://click.example/u/1"]
@pytest.mark.parametrize("html", [
# another sentence of the same paragraph
'<p>Zum Abbestellen bitte antworten. Mehr dazu gibt es <a href="https://click.example/x">hier</a>.</p>',
'<p>Mehr dazu gibt es <a href="https://click.example/x">hier</a>. Abbestellen geht per Antwort.</p>',
# a line break ends the sentence
'<td>Newsletter abbestellen<br><a href="https://click.example/x">hier</a> geht es zum Shop</td>',
# another block
'<p>Newsletter abbestellen ist einfach</p><p><a href="https://click.example/x">hier</a> zum Shop</p>',
'<p><a href="https://click.example/x">hier</a> zum Shop</p><div>Newsletter abbestellen</div>',
# a link with a text of its own is judged by that text
'<p>Abbestellen kannst du jederzeit, lies dazu <a href="https://click.example/x">unsere ausführlichen '
'Hinweise zum Datenschutz</a>.</p>',
# the word belongs to another link (navigation row) or to a URL written out in the text
'<p><a href="https://ex.org/unsubscribe/1">Abmelden</a> | <a href="https://click.example/x">Home</a></p>',
'<p>https://ex.org/unsubscribe/1 oder <a href="https://click.example/x">Home</a></p>',
# not a web link
'<p>Zum Abbestellen <a href="mailto:x@example.org">hier</a> schreiben.</p>',
])
def test_short_link_outside_the_unsubscribe_sentence_is_not_taken(html):
assert "https://click.example/x" not in unsub.find_unsubscribe_links(html)
assert all("mailto" not in link for link in unsub.find_unsubscribe_links(html))
def test_link_found_by_its_own_text_is_not_listed_twice_and_comes_first():
html = ('<p>Du kannst dich <a href="https://click.example/a">hier</a> abmelden oder '
'<a href="https://click.example/b">Newsletter abbestellen</a>.</p>')
assert unsub.find_unsubscribe_links(html) == ["https://click.example/b", "https://click.example/a"]