Parse bbfile dataset in Blackboard Ultra bodies

Some files can contain link metadata embedded in dataset HTML attributes. This metadata is needed to provide local files with accurate file names. Moreover, inner text of links inside body descriptions are now manually assigned when they are empty.

blackboard_sync/content/webdav.py

@@ -20,6 +20,7 @@
 # MA  02110-1301, USA.
 
 
+import json
 import mimetypes
 from pathlib import Path
 from requests import Response
@@ -55,21 +56,41 @@ def __init__(self, body: str, base_url: str,
     def _find_replace(self, soup: BeautifulSoup,
                       tag: str, attr: str, base_url: str) -> list[Link]:
         links = []
+        is_link = tag == "a"
 
         for el in soup.find_all(tag):
             # Add link for later download
             uri = el.get(attr)
 
             if uri:
+                # Parse JSON dataset
+                filename = _parse_bbfile(el) if is_link else None
                 # Handle url-encoding
-                filename = unquote(uri.split('/')[-1])
+                filename = filename or unquote(uri.split('/')[-1])
                 links.append(Link(href=uri, text=filename))
 
+                # Assign inner text if missing
+                if not el.string:
+                    el.append(filename)
+
                 # Replace for local instance
                 if uri.startswith(base_url):
                     el[attr] = filename
         return links
 
+    def _parse_bbfile(self, el: str) -> str | None:
+        json_data = el.get("data-bbfile")
+        filename = None
+
+        if json_data:
+            try:
+                bbfile = json.loads(json_data)
+            except json.JSONDecodeError:
+                pass
+            else:
+                filename = bbfile.get("linkName")
+        return filename
+
     @property
     def links(self) -> List[Link]:
         return self._links