Unscape HTML Entities from RSS feeds (#130915)

* Unscape HTML Entities from RSS feeds

* Improve tests

homeassistant/components/feedreader/config_flow.py

@@ -2,6 +2,7 @@
 
 from __future__ import annotations
 
+import html
 import logging
 from typing import Any
 import urllib.error
@@ -107,7 +108,7 @@ class FeedReaderConfigFlow(ConfigFlow, domain=DOMAIN):
                     return self.abort_on_import_error(user_input[CONF_URL], "url_error")
                 return self.show_user_form(user_input, {"base": "url_error"})
 
-        feed_title = feed["feed"]["title"]
+        feed_title = html.unescape(feed["feed"]["title"])
 
         return self.async_create_entry(
             title=feed_title,

homeassistant/components/feedreader/coordinator.py

@@ -4,6 +4,7 @@ from __future__ import annotations
 
 from calendar import timegm
 from datetime import datetime
+import html
 from logging import getLogger
 from time import gmtime, struct_time
 from typing import TYPE_CHECKING
@@ -102,7 +103,8 @@ class FeedReaderCoordinator(
         """Set up the feed manager."""
         feed = await self._async_fetch_feed()
         self.logger.debug("Feed data fetched from %s : %s", self.url, feed["feed"])
-        self.feed_author = feed["feed"].get("author")
+        if feed_author := feed["feed"].get("author"):
+            self.feed_author = html.unescape(feed_author)
         self.feed_version = feedparser.api.SUPPORTED_VERSIONS.get(feed["version"])
         self._feed = feed
 

homeassistant/components/feedreader/event.py

@@ -2,6 +2,7 @@
 
 from __future__ import annotations
 
+import html
 import logging
 
 from feedparser import FeedParserDict
@@ -76,15 +77,22 @@ class FeedReaderEvent(CoordinatorEntity[FeedReaderCoordinator], EventEntity):
         # so we always take the first entry in list, since we only care about the latest entry
         feed_data: FeedParserDict = data[0]
 
+        if description := feed_data.get("description"):
+            description = html.unescape(description)
+
+        if title := feed_data.get("title"):
+            title = html.unescape(title)
+
         if content := feed_data.get("content"):
             if isinstance(content, list) and isinstance(content[0], dict):
                 content = content[0].get("value")
+            content = html.unescape(content)
 
         self._trigger_event(
             EVENT_FEEDREADER,
             {
-                ATTR_DESCRIPTION: feed_data.get("description"),
-                ATTR_TITLE: feed_data.get("title"),
+                ATTR_DESCRIPTION: description,
+                ATTR_TITLE: title,
                 ATTR_LINK: feed_data.get("link"),
                 ATTR_CONTENT: content,
             },

tests/components/feedreader/conftest.py

@@ -64,6 +64,18 @@ def fixture_feed_only_summary(hass: HomeAssistant) -> bytes:
     return load_fixture_bytes("feedreader8.xml")
 
 
+@pytest.fixture(name="feed_htmlentities")
+def fixture_feed_htmlentities(hass: HomeAssistant) -> bytes:
+    """Load test feed data with HTML Entities."""
+    return load_fixture_bytes("feedreader9.xml")
+
+
+@pytest.fixture(name="feed_atom_htmlentities")
+def fixture_feed_atom_htmlentities(hass: HomeAssistant) -> bytes:
+    """Load test ATOM feed data with HTML Entities."""
+    return load_fixture_bytes("feedreader10.xml")
+
+
 @pytest.fixture(name="events")
 async def fixture_events(hass: HomeAssistant) -> list[Event]:
     """Fixture that catches alexa events."""

tests/components/feedreader/fixtures/feedreader10.xml

@@ -0,0 +1,19 @@
+<?xml version="1.0" encoding="utf-8"?>
+<feed
+	xmlns="http://www.w3.org/2005/Atom">
+	<title><![CDATA[ATOM RSS en espa&ntilde;ol]]></title>
+	<link href="http://example.org/"/>
+	<updated>2024-11-18T14:00:00Z</updated>
+	<author>
+		<name><![CDATA[Juan P&eacute;rez]]></name>
+	</author>
+	<id>urn:uuid:60a76c80-d399-11d9-b93C-0003939e0af6</id>
+	<entry>
+		<title><![CDATA[T&iacute;tulo]]></title>
+		<link href="http://example.org/2003/12/13/atom03"/>
+		<id>urn:uuid:1225c695-cfb8-4ebb-aaaa-80da344efa6a</id>
+		<updated>2024-11-18T14:00:00Z</updated>
+		<summary><![CDATA[Resumen en espa&ntilde;ol]]></summary>
+		<content><![CDATA[Contenido en espa&ntilde;ol]]></content>
+	</entry>
+</feed>

tests/components/feedreader/fixtures/feedreader9.xml

@@ -0,0 +1,21 @@
+<?xml version="1.0" encoding="UTF-8" ?>
+<rss version="2.0">
+    <channel>
+        <title><![CDATA[RSS en espa&ntilde;ol]]></title>
+        <description><![CDATA[Esto es un ejemplo de un feed RSS en espa&ntilde;ol]]></description>
+        <link>http://www.example.com/main.html</link>
+        <lastBuildDate>Mon, 18 Nov 2024 15:00:00 +1000</lastBuildDate>
+        <pubDate>Mon, 18 Nov 2024 15:00:00 +1000</pubDate>
+        <ttl>1800</ttl>
+
+        <item>
+            <title><![CDATA[T&iacute;tulo 1]]></title>
+            <description><![CDATA[Descripci&oacute;n 1]]></description>
+            <link>http://www.example.com/link/1</link>
+            <guid isPermaLink="false">GUID 1</guid>
+            <pubDate>Mon, 18 Nov 2024 15:00:00 +1000</pubDate>
+            <content><![CDATA[Contenido 1 en espa&ntilde;ol]]></content>
+        </item>
+
+    </channel>
+</rss>

tests/components/feedreader/snapshots/test_event.ambr

@@ -0,0 +1,27 @@
+# serializer version: 1
+# name: test_event_htmlentities[feed_atom_htmlentities]
+  ReadOnlyDict({
+    'content': 'Contenido en español',
+    'description': 'Resumen en español',
+    'event_type': 'feedreader',
+    'event_types': list([
+      'feedreader',
+    ]),
+    'friendly_name': 'Mock Title',
+    'link': 'http://example.org/2003/12/13/atom03',
+    'title': 'Título',
+  })
+# ---
+# name: test_event_htmlentities[feed_htmlentities]
+  ReadOnlyDict({
+    'content': 'Contenido 1 en español',
+    'description': 'Descripción 1',
+    'event_type': 'feedreader',
+    'event_types': list([
+      'feedreader',
+    ]),
+    'friendly_name': 'Mock Title',
+    'link': 'http://www.example.com/link/1',
+    'title': 'Título 1',
+  })
+# ---

tests/components/feedreader/test_config_flow.py

@@ -246,3 +246,38 @@ async def test_options_flow(hass: HomeAssistant) -> None:
     assert result["data"] == {
         CONF_MAX_ENTRIES: 10,
     }
+
+
+@pytest.mark.parametrize(
+    ("fixture_name", "expected_title"),
+    [
+        ("feed_htmlentities", "RSS en español"),
+        ("feed_atom_htmlentities", "ATOM RSS en español"),
+    ],
+)
+async def test_feed_htmlentities(
+    hass: HomeAssistant,
+    feedparser,
+    setup_entry,
+    fixture_name,
+    expected_title,
+    request: pytest.FixtureRequest,
+) -> None:
+    """Test starting a flow by user from a feed with HTML Entities in the title."""
+    with patch(
+        "homeassistant.components.feedreader.config_flow.feedparser.http.get",
+        side_effect=[request.getfixturevalue(fixture_name)],
+    ):
+        # init user flow
+        result = await hass.config_entries.flow.async_init(
+            DOMAIN, context={"source": SOURCE_USER}
+        )
+        assert result["type"] is FlowResultType.FORM
+        assert result["step_id"] == "user"
+
+        # success
+        result = await hass.config_entries.flow.async_configure(
+            result["flow_id"], user_input={CONF_URL: URL}
+        )
+        assert result["type"] is FlowResultType.CREATE_ENTRY
+        assert result["title"] == expected_title

tests/components/feedreader/test_event.py

@@ -3,6 +3,9 @@
 from datetime import timedelta
 from unittest.mock import patch
 
+import pytest
+from syrupy.assertion import SnapshotAssertion
+
 from homeassistant.components.feedreader.event import (
     ATTR_CONTENT,
     ATTR_DESCRIPTION,
@@ -59,3 +62,31 @@ async def test_event_entity(
         assert state.attributes[ATTR_LINK] == "http://www.example.com/link/1"
         assert state.attributes[ATTR_CONTENT] == "This is a summary"
         assert state.attributes[ATTR_DESCRIPTION] == "Description 1"
+
+
+@pytest.mark.parametrize(
+    ("fixture_name"),
+    [
+        ("feed_htmlentities"),
+        ("feed_atom_htmlentities"),
+    ],
+)
+async def test_event_htmlentities(
+    hass: HomeAssistant,
+    snapshot: SnapshotAssertion,
+    fixture_name,
+    request: pytest.FixtureRequest,
+) -> None:
+    """Test feed event entity with HTML Entities."""
+    entry = create_mock_entry(VALID_CONFIG_DEFAULT)
+    entry.add_to_hass(hass)
+    with patch(
+        "homeassistant.components.feedreader.coordinator.feedparser.http.get",
+        side_effect=[request.getfixturevalue(fixture_name)],
+    ):
+        assert await hass.config_entries.async_setup(entry.entry_id)
+        await hass.async_block_till_done()
+
+        state = hass.states.get("event.mock_title")
+        assert state
+        assert state.attributes == snapshot

tests/components/feedreader/test_init.py

@@ -12,6 +12,7 @@ import pytest
 
 from homeassistant.components.feedreader.const import DOMAIN
 from homeassistant.core import Event, HomeAssistant
+from homeassistant.helpers import device_registry as dr
 import homeassistant.util.dt as dt_util
 
 from . import async_setup_config_entry, create_mock_entry
@@ -357,3 +358,23 @@ async def test_feed_errors(
         freezer.tick(timedelta(hours=1, seconds=1))
         async_fire_time_changed(hass)
         await hass.async_block_till_done(wait_background_tasks=True)
+
+
+async def test_feed_atom_htmlentities(
+    hass: HomeAssistant, feed_atom_htmlentities, device_registry: dr.DeviceRegistry
+) -> None:
+    """Test ATOM feed author with HTML Entities."""
+
+    entry = create_mock_entry(VALID_CONFIG_DEFAULT)
+    entry.add_to_hass(hass)
+    with patch(
+        "homeassistant.components.feedreader.coordinator.feedparser.http.get",
+        side_effect=[feed_atom_htmlentities],
+    ):
+        assert await hass.config_entries.async_setup(entry.entry_id)
+        await hass.async_block_till_done()
+
+        device_entry = device_registry.async_get_device(
+            identifiers={(DOMAIN, entry.entry_id)}
+        )
+        assert device_entry.manufacturer == "Juan Pérez"