Refactor post_embed function and simplify first_image extraction logic

discord_rss_bot/feeds.py

@@ -1,6 +1,7 @@
 from __future__ import annotations
 
 import datetime
+import json
 import logging
 import os
 import pprint
@@ -36,6 +37,7 @@ from discord_rss_bot.hoyolab_api import extract_post_id_from_hoyolab_url
 from discord_rss_bot.hoyolab_api import fetch_hoyolab_post
 from discord_rss_bot.hoyolab_api import is_c3kay_feed
 from discord_rss_bot.is_url_valid import is_url_valid
+from discord_rss_bot.settings import default_custom_embed
 from discord_rss_bot.settings import default_custom_message
 from discord_rss_bot.settings import get_reader
 
@@ -570,5 +572,8 @@ def create_feed(reader: Reader, feed_url: str, webhook_dropdown: str) -> None:
     # This is the default message that will be sent to Discord.
     reader.set_tag(clean_feed_url, "custom_message", default_custom_message)  # pyright: ignore[reportArgumentType]
 
+    # Set the default embed tag when creating the feed
+    reader.set_tag(clean_feed_url, "embed", json.dumps(default_custom_embed))
+
     # Update the full-text search index so our new feed is searchable.
     reader.update_search()

discord_rss_bot/main.py

@@ -620,7 +620,7 @@ async def get_embed_page(
 
 
 @app.post("/embed", response_class=HTMLResponse)
-async def post_embed(
+async def post_embed(  # noqa: C901
     feed_url: Annotated[str, Form()],
     reader: Annotated[Reader, Depends(get_reader_dependency)],
     title: Annotated[str, Form()] = "",
@@ -657,15 +657,28 @@ async def post_embed(
     feed: Feed = reader.get_feed(urllib.parse.unquote(clean_feed_url))
 
     custom_embed: CustomEmbed = get_embed(reader, feed)
+    # Only overwrite fields that the user provided. This prevents accidental
+    # clearing of previously saved embed data when the form submits empty
+    # values for fields the user did not change.
+    if title:
         custom_embed.title = title
+    if description:
         custom_embed.description = description
+    if color:
         custom_embed.color = color
+    if image_url:
         custom_embed.image_url = image_url
+    if thumbnail_url:
         custom_embed.thumbnail_url = thumbnail_url
+    if author_name:
         custom_embed.author_name = author_name
+    if author_url:
         custom_embed.author_url = author_url
+    if author_icon_url:
         custom_embed.author_icon_url = author_icon_url
+    if footer_text:
         custom_embed.footer_text = footer_text
+    if footer_icon_url:
         custom_embed.footer_icon_url = footer_icon_url
 
     # Save the data.

tests/conftest.py

@@ -4,11 +4,14 @@ import os
 import shutil
 import sys
 import tempfile
+import warnings
 from contextlib import suppress
 from pathlib import Path
 from typing import TYPE_CHECKING
 from typing import Any
 
+from bs4 import MarkupResemblesLocatorWarning
+
 if TYPE_CHECKING:
     import pytest
 
@@ -34,6 +37,11 @@ def pytest_sessionstart(session: pytest.Session) -> None:
 
     os.environ["DISCORD_RSS_BOT_DATA_DIR"] = str(worker_data_dir)
 
+    # Tests call markdownify which may invoke BeautifulSoup on strings that look
+    # like URLs; that triggers MarkupResemblesLocatorWarning from bs4. Silence
+    # that warning during tests to avoid noisy output.
+    warnings.filterwarnings("ignore", category=MarkupResemblesLocatorWarning)
+
     # If modules were imported before this hook (unlikely), force them to use
     # the worker-specific location.
     settings_module: Any = sys.modules.get("discord_rss_bot.settings")