feat: Add auto-discovery from popular services

.pre-commit-config.yaml

@@ -62,5 +62,6 @@ repos:
         language: system
         require_serial: true
         pass_filenames: false
+        always_run: true
         files: ^podcast_archiver/config\.py$
         types: [python]

README.md

@@ -62,12 +62,40 @@ podcast-archiver --dir ~/Podcasts --feed https://feeds.feedburner.com/TheAnthrop
 
 Podcast Archiver expects values to its `--feed/-f` argument to be URLs pointing to an [RSS feed of a podcast](https://archive.is/jYk3E).
 
+If you are not certain if the link you have for a show that you like, you can try and pass it to Podcast Archiver directly. The archiver supports a variety of links from popular podcast players and platforms, including [Apple Podcasts](https://podcasts.apple.com/us/browse), [Overcast.fm](https://overcast.fm/), [Castro](https://castro.fm/), and [Pocket Casts](https://pocketcasts.com/):
+
+```sh
+# Archive from Apple Podcasts URL
+podcast-archiver -f https://podcasts.apple.com/us/podcast/black-girl-gone-a-true-crime-podcast/id1556267741
+# ... or just the ID
+podcast-archiver -f 1556267741
+
+# From Overcast podcast URL
+podcast-archiver -f https://overcast.fm/itunes394775318/99-invisible
+# ... or episode sharing links (will resolve to all episodes)
+podcast-archiver -f https://overcast.fm/+AAyIOzrEy1g
+```
+
+#### Supported services
+
+The table below lists most of the supported services and URLs. If you think that some service you like is missing here, [please let me know](/~https://github.com/janw/podcast-archiver/issues/new)!
+
+| Service                          | Example URL                                                                            | Notes |
+| -------------------------------- | -------------------------------------------------------------------------------------- | ----- |
+| Apple Podcasts                   | <https://podcasts.apple.com/us/podcast/the-anthropocene-reviewed/id1342003491>         |       |
+| [Overcast](https://overcast.fm/) | <https://overcast.fm/itunes394775318/99-invisible>, <https://overcast.fm/+AAyIOzrEy1g> |       |
+| [Castro](https://castro.fm/)     | <https://castro.fm/podcast/f996ae94-70a2-4d9c-afbc-c70b5bacd120>                       |       |
+
+#### Local files
+
 Feeds can also be "fetched" from a local file:
 
 ```bash
 podcast-archiver -f file:/Users/janw/downloaded_feed.xml
 ```
 
+#### Testing without downloading
+
 To find out if you have to the right feed, you may want to use the `--dry-run` option to output the discovered feed information and found episodes. It will prevent all downloads.
 
 ```sh

config.yaml.example

@@ -1,5 +1,5 @@
 ## Podcast-Archiver configuration
-## Generated using podcast-archiver v2.0.2
+## Generated using podcast-archiver v2.1.0
 
 # Field 'feeds': Feed URLs to archive.
 #

cspell.config.yaml

@@ -53,3 +53,8 @@ words:
   - venv
   - virtualenv
   - willhaus
+  - rsps
+  - soundcloud
+  - logbuch
+  - netzpolitik
+  - wochendaemmerung

hack/rich-codex.sh

@@ -3,6 +3,7 @@
 TMPDIR=$(mktemp -d 2>/dev/null || mktemp -d -t 'tmpdir')
 
 export FORCE_COLOR="1"
+export TERM="xterm-16color"
 export COLUMNS="120"
 export CREATED_FILES="created.txt"
 export DELETED_FILES="deleted.txt"
@@ -17,4 +18,4 @@ export PODCAST_ARCHIVER_IGNORE_DATABASE=true
 # shellcheck disable=SC2064
 trap "rm -rf '$TMPDIR'" EXIT
 
-exec poetry run rich-codex --terminal-width $COLUMNS --notrim
+poetry run rich-codex --terminal-width $COLUMNS --notrim --terminal-theme DIMMED_MONOKAI

podcast_archiver/console.py

@@ -5,7 +5,8 @@
 
 _theme = Theme(
     {
-        "error": "bold dark_red",
+        "error": "dark_red bold",
+        "errorhint": "dark_red dim",
         "warning": "orange1 bold",
         "warning_hint": "orange1 dim",
         "completed": "dark_cyan bold",

podcast_archiver/constants.py

@@ -7,7 +7,7 @@
 USER_AGENT = f"{PROG_NAME}/{__version__} (/~https://github.com/janw/podcast-archiver)"
 ENVVAR_PREFIX = "PODCAST_ARCHIVER"
 
-REQUESTS_TIMEOUT = 30
+REQUESTS_TIMEOUT = (5, 30)
 
 SUPPORTED_LINK_TYPES_RE = re.compile(r"^(audio|video)/")
 DOWNLOAD_CHUNK_SIZE = 256 * 1024

podcast_archiver/exceptions.py

@@ -45,3 +45,7 @@ class NotModified(PodcastArchiverException):
     def __init__(self, info: FeedInfo, *args: object) -> None:
         super().__init__(*args)
         self.info = info
+
+
+class NotSupported(PodcastArchiverException):
+    pass

podcast_archiver/models/feed.py

@@ -10,7 +10,7 @@
 from pydantic import AliasChoices, BaseModel, ConfigDict, Field, field_validator
 
 from podcast_archiver.constants import MAX_TITLE_LENGTH
-from podcast_archiver.exceptions import NotModified
+from podcast_archiver.exceptions import NotModified, NotSupported
 from podcast_archiver.logging import logger, rprint
 from podcast_archiver.models.episode import EpisodeOrFallback
 from podcast_archiver.models.field_types import LenientDatetime
@@ -90,6 +90,13 @@ def truncate_title(cls, value: str) -> str:
     def field_titles(cls) -> list[str]:
         return [field.title for field in cls.model_fields.values() if field.title]
 
+    @property
+    def alternate_rss(self) -> str | None:
+        for link in self.links:
+            if link.rel == "alternate" and link.link_type == "application/rss+xml":
+                return link.href
+        return None
+
 
 class FeedPage(BaseModel):
     model_config = ConfigDict(arbitrary_types_allowed=True)
@@ -103,38 +110,46 @@ class FeedPage(BaseModel):
     episodes: list[EpisodeOrFallback] = Field(default_factory=list, validation_alias=AliasChoices("entries", "items"))
 
     @classmethod
-    def parse_feed(cls, source: str | bytes, alt_url: str | None) -> FeedPage:
+    def parse_feed(cls, source: str | bytes, alt_url: str | None, retry: bool = False) -> FeedPage:
         feedobj = feedparser.parse(source)
         obj = cls.model_validate(feedobj)
-        if obj.bozo and (exc := obj.bozo_exception) and isinstance(exc, SAXParseException):
-            url = source if isinstance(source, str) and not alt_url else alt_url
+        if not obj.bozo:
+            return obj
+
+        if (fallback_url := obj.feed.alternate_rss) and not retry:
+            logger.info("Attempting to fetch alternate feed at %s", fallback_url)
+            return cls.from_url(fallback_url, retry=True)
+
+        url = source if isinstance(source, str) and not alt_url else alt_url
+        if (exc := obj.bozo_exception) and isinstance(exc, SAXParseException):
             rprint(f"Feed content is not well-formed for {url}", style="warning")
-            rprint(f"Continuing processing but here be dragons ({exc.getMessage()})", style="warning_hint")
-        return obj
+            rprint(f"Attemping processing but here be dragons ({exc.getMessage()})", style="warninghint")
+
+        raise NotSupported(f"Content at {url} is not supported")
 
     @classmethod
-    def from_url(cls, url: str, *, known_info: FeedInfo | None = None) -> FeedPage:
+    def from_url(cls, url: str, *, known_info: FeedInfo | None = None, retry: bool = False) -> FeedPage:
         parsed = urlparse(url)
         if parsed.scheme == "file":
             return cls.parse_feed(parsed.path, None)
 
         if not known_info:
-            return cls.from_response(session.get_and_raise(url), alt_url=url)
+            return cls.from_response(session.get_and_raise(url), alt_url=url, retry=retry)
 
         response = session.get_and_raise(url, last_modified=known_info.last_modified)
         if response.status_code == HTTPStatus.NOT_MODIFIED:
             logger.debug("Server reported 'not modified' from %s, skipping fetch.", known_info.last_modified)
             raise NotModified(known_info)
 
-        instance = cls.from_response(response, alt_url=url)
+        instance = cls.from_response(response, alt_url=url, retry=retry)
         if instance.feed.updated_time == known_info.updated_time:
             logger.debug("Feed's updated time %s did not change, skipping fetch.", known_info.updated_time)
             raise NotModified(known_info)
 
         return instance
 
     @classmethod
-    def from_response(cls, response: Response, alt_url: str | None) -> FeedPage:
-        instance = cls.parse_feed(response.content, alt_url=alt_url)
+    def from_response(cls, response: Response, alt_url: str | None, retry: bool) -> FeedPage:
+        instance = cls.parse_feed(response.content, alt_url=alt_url, retry=retry)
         instance.feed.last_modified = response.headers.get("Last-Modified")
         return instance