From 67453ed17ce08002b8b76d39941d57a0f71e1d7c Mon Sep 17 00:00:00 2001
From: JB <jb-apps@outlook.com>
Date: Tue, 11 Nov 2025 07:29:13 -0800
Subject: [PATCH] feat: Implement filtering of unsupported download formats in
 CalibreOPDSClient

---
 abogen/integrations/calibre_opds.py | 32 ++++++++++++++++++------
 tests/test_calibre_opds.py          | 38 +++++++++++++++++++++++++++++
 2 files changed, 63 insertions(+), 7 deletions(-)

diff --git a/abogen/integrations/calibre_opds.py b/abogen/integrations/calibre_opds.py
index f2ea3f4..ff425ef 100644
--- a/abogen/integrations/calibre_opds.py
+++ b/abogen/integrations/calibre_opds.py
@@ -4,6 +4,7 @@ import dataclasses
 import html
 import re
 from dataclasses import dataclass, field
+from pathlib import PurePosixPath
 from typing import Any, Dict, Iterable, List, Mapping, Optional
 from urllib.parse import urljoin, urlparse
 from xml.etree import ElementTree as ET
@@ -38,6 +39,8 @@ _EPUB_MIME_TYPES = {
     "application/x-zip",
     "application/x-zip-compressed",
 }
+_SUPPORTED_DOWNLOAD_MIME_TYPES = set(_EPUB_MIME_TYPES) | {"application/pdf"}
+_SUPPORTED_DOWNLOAD_EXTENSIONS = {".epub", ".pdf"}
 
 
 class CalibreOPDSError(RuntimeError):
@@ -270,7 +273,12 @@ class CalibreOPDSClient:
         feed_id = root.findtext("atom:id", default=None, namespaces=NS)
         feed_title = root.findtext("atom:title", default=None, namespaces=NS)
         links = self._parse_links(root.findall("atom:link", NS), base_url)
-        entries = [self._parse_entry(node, base_url) for node in root.findall("atom:entry", NS)]
+        parsed_entries = [self._parse_entry(node, base_url) for node in root.findall("atom:entry", NS)]
+        entries = [
+            entry
+            for entry in parsed_entries
+            if entry.download and self._is_supported_download(entry.download)
+        ]
         return OPDSFeed(id=feed_id, title=feed_title, entries=entries, links=links)
 
     def _parse_entry(self, node: ET.Element, base_url: str) -> OPDSEntry:
@@ -576,14 +584,27 @@ class CalibreOPDSClient:
             results[key] = entry
         return results
 
+    @staticmethod
+    def _is_supported_download(link: OPDSLink) -> bool:
+        mime = (link.type or "").split(";")[0].strip().lower()
+        if mime in _SUPPORTED_DOWNLOAD_MIME_TYPES:
+            return True
+        href = (link.href or "").strip()
+        if not href:
+            return False
+        parsed_path = urlparse(href).path or ""
+        extension = PurePosixPath(parsed_path).suffix.lower()
+        return extension in _SUPPORTED_DOWNLOAD_EXTENSIONS
+
     @staticmethod
     def _select_download_link(links: Mapping[str, OPDSLink] | Iterable[OPDSLink]) -> Optional[OPDSLink]:
         if isinstance(links, Mapping):
             iterable: List[OPDSLink] = list(links.values())
         else:
             iterable = list(links)
+        supported = [link for link in iterable if CalibreOPDSClient._is_supported_download(link)]
         best: Optional[OPDSLink] = None
-        for link in iterable:
+        for link in supported:
             rel = (link.rel or "").lower()
             if "acquisition" not in rel:
                 continue
@@ -594,11 +615,8 @@ class CalibreOPDSClient:
                 best = link
         if best:
             return best
-        # Fallback to first epub-like link even without explicit acquisition rel
-        for link in iterable:
-            mime = (link.type or "").lower()
-            if mime in _EPUB_MIME_TYPES:
-                return link
+        if supported:
+            return supported[0]
         # No valid acquisition-style link exposed
         return None
 
diff --git a/tests/test_calibre_opds.py b/tests/test_calibre_opds.py
index edd5975..22d4186 100644
--- a/tests/test_calibre_opds.py
+++ b/tests/test_calibre_opds.py
@@ -104,3 +104,41 @@ def test_calibre_opds_relative_urls_keep_catalog_prefix() -> None:
   assert client._make_url("books/sample.epub") == "http://example.com/opds/books/sample.epub"
   assert client._make_url("/cover/1") == "http://example.com/cover/1"
   assert client._make_url("?page=2") == "http://example.com/opds/?page=2"
+
+
+def test_calibre_opds_filters_out_unsupported_formats() -> None:
+    client = CalibreOPDSClient("http://example.com/catalog")
+    xml_payload = """<?xml version=\"1.0\" encoding=\"UTF-8\"?>
+    <feed xmlns=\"http://www.w3.org/2005/Atom\">
+      <id>catalog</id>
+      <title>Example Catalog</title>
+      <entry>
+        <id>audio-book</id>
+        <title>Unsupported Audio</title>
+        <link rel=\"http://opds-spec.org/acquisition\"
+              href=\"books/sample.mp3\"
+              type=\"audio/mpeg\" />
+      </entry>
+      <entry>
+        <id>pdf-book</id>
+        <title>Allowed PDF</title>
+        <link rel=\"http://opds-spec.org/acquisition\"
+              href=\"books/sample.pdf\"
+              type=\"application/pdf\" />
+      </entry>
+      <entry>
+        <id>epub-book</id>
+        <title>Allowed EPUB</title>
+        <link rel=\"http://opds-spec.org/acquisition\"
+              href=\"books/sample.epub\" />
+      </entry>
+    </feed>
+    """
+
+    feed = client._parse_feed(xml_payload, base_url="http://example.com/catalog")
+
+    identifiers = {entry.id for entry in feed.entries}
+    assert identifiers == {"pdf-book", "epub-book"}
+    for entry in feed.entries:
+        assert entry.download is not None
+        assert entry.download.href.endswith((".pdf", ".epub"))