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 +62,34 @@ 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
+
+TBD
+
+#### 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.
 #

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)
@@ -103,17 +110,25 @@ 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)

podcast_archiver/processor.py

@@ -19,6 +19,7 @@
     FutureEpisodeResult,
     ProcessingResult,
 )
+from podcast_archiver.urls import registry
 from podcast_archiver.utils import FilenameFormatter, handle_feed_request
 from podcast_archiver.utils.pretty_printing import PrettyPrintEpisodeRange
 
@@ -59,8 +60,9 @@ def process(self, url: str, dry_run: bool = False) -> ProcessingResult:
         return result
 
     def load_feed(self, url: str, known_feeds: dict[str, FeedInfo]) -> Feed | None:
-        with handle_feed_request(url):
-            feed = Feed(url=url, known_info=known_feeds.get(url))
+        resolved_url = registry.get_feed(url) or url
+        with handle_feed_request(resolved_url):
+            feed = Feed(url=resolved_url, known_info=known_feeds.get(url))
             known_feeds[url] = feed.info
             return feed
 
@@ -116,7 +118,7 @@ def process_feed(self, feed: Feed, dry_run: bool) -> ProcessingResult:
                 exists = isinstance(enqueued, EpisodeResult) and enqueued.result == DownloadResult.ALREADY_EXISTS
                 pretty_range.update(exists, episode)
 
-                if not dry_run or self.settings.verbose > 0:
+                if not dry_run:
                     results.append(enqueued)
 
                 if (max_count := self.settings.maximum_episode_count) and idx == max_count:

podcast_archiver/session.py

@@ -40,6 +40,22 @@ def get_and_raise(
         response.raise_for_status()
         return response
 
+    def request(self, method: str, url: str, **kwargs: Any) -> Response:  # type: ignore[override]
+        if url.startswith("https://"):
+            return super().request(method, url, **kwargs)
+
+        if url.startswith("http://"):
+            securl = "https" + url[4:]
+        else:
+            securl = "https://" + url
+
+        try:
+            return super().request(method, securl, **kwargs)
+        except Exception:
+            pass
+
+        return super().request(method, url, **kwargs)
+
 
 session = ArchiverSession()
 session.mount("http://", _adapter)

podcast_archiver/urls/__init__.py

@@ -0,0 +1,23 @@
+from podcast_archiver.urls.base import UrlSourceRegistry
+from podcast_archiver.urls.base64 import Base64EncodedUrlSource
+from podcast_archiver.urls.fireside import FiresideSource
+from podcast_archiver.urls.prefixed import UrlPrefixSource
+from podcast_archiver.urls.soundcloud import SoundCloudSource
+from podcast_archiver.urls.via_apple import (
+    ApplePodcastsByIdSource,
+    ApplePodcastsSource,
+    ContainingApplePodcastsUrlSource,
+)
+
+registry = UrlSourceRegistry()
+
+registry.register(ApplePodcastsSource)
+registry.register(ApplePodcastsByIdSource)
+registry.register(ContainingApplePodcastsUrlSource)
+registry.register(UrlPrefixSource)
+registry.register(Base64EncodedUrlSource)
+
+# Known website sources that define feeds as alternate+application/rss+xml
+# or use a deterministic URL pattern to find the feed URL from the website URL.
+registry.register(FiresideSource)
+registry.register(SoundCloudSource)