Adds the option to fetch metadata to imports

bookmarks/services/bookmarks.py

@@ -69,15 +69,16 @@ def update_bookmark(bookmark: Bookmark, tag_string, current_user: User):
     return bookmark
 
 
-def enhance_with_website_metadata(bookmark: Bookmark):
+def enhance_with_website_metadata(bookmark: Bookmark, save: bool = True):
     metadata = website_loader.load_website_metadata(bookmark.url)
     if not bookmark.title:
         bookmark.title = metadata.title or ""
 
     if not bookmark.description:
         bookmark.description = metadata.description or ""
 
-    bookmark.save()
+    if save:
+        bookmark.save()
 
 
 def archive_bookmark(bookmark: Bookmark):

bookmarks/services/importer.py

@@ -1,6 +1,7 @@
 import logging
 from dataclasses import dataclass
 from typing import List
+from concurrent.futures import ThreadPoolExecutor
 
 from django.contrib.auth.models import User
 from django.utils import timezone
@@ -9,6 +10,7 @@
 from bookmarks.services import tasks
 from bookmarks.services.parser import parse, NetscapeBookmark
 from bookmarks.utils import parse_timestamp
+from bookmarks.services.bookmarks import enhance_with_website_metadata
 
 logger = logging.getLogger(__name__)
 
@@ -23,6 +25,7 @@ class ImportResult:
 @dataclass
 class ImportOptions:
     map_private_flag: bool = False
+    get_metadata_flag: bool = False
 
 
 class TagCache:
@@ -132,8 +135,8 @@ def _import_batch(
     existing_bookmarks = Bookmark.objects.filter(owner=user, url__in=batch_urls)
 
     # Create or update bookmarks from parsed Netscape bookmarks
-    bookmarks_to_create = []
-    bookmarks_to_update = []
+    bookmarks_to_create: List[Bookmark] = []
+    bookmarks_to_update: List[Bookmark]  = []
 
     for netscape_bookmark in netscape_bookmarks:
         result.total = result.total + 1
@@ -169,6 +172,18 @@ def _import_batch(
             logging.exception("Error importing bookmark: " + shortened_bookmark_tag_str)
             result.failed = result.failed + 1
 
+    if options.get_metadata_flag:
+        bookmarks_missing_metadata = filter(
+            lambda b: not b.title or not b.description,
+            bookmarks_to_create + bookmarks_to_update,
+        )
+
+        with ThreadPoolExecutor(5) as executor:
+            executor.map(
+                lambda b: enhance_with_website_metadata(b, save=False),
+                bookmarks_missing_metadata,
+            )
+
     # Bulk update bookmarks in DB
     Bookmark.objects.bulk_update(
         bookmarks_to_update,

bookmarks/templates/settings/general.html

@@ -325,6 +325,13 @@ <h2>Import</h2>
             private as shared bookmarks.
             Otherwise, all bookmarks will be imported as private bookmarks.
           </div>
+          <label for="import_get_metadata_flag" class="form-checkbox">
+            <input type="checkbox" id="import_get_metadata_flag" name="get_metadata_flag">
+            <i class="form-icon"></i> Fetch missing metadata
+          </label>
+          <div class="form-input-hint">
+            Enabling this option will fetch metadata for missing title or description fields for both updated and newly added bookmarks.
+          </div>
         </div>
         <div class="form-group">
           <div class="input-group width-75 width-md-100">

bookmarks/views/settings.py

@@ -198,7 +198,8 @@ def integrations(request):
 def bookmark_import(request):
     import_file = request.FILES.get("import_file")
     import_options = importer.ImportOptions(
-        map_private_flag=request.POST.get("map_private_flag") == "on"
+        map_private_flag=request.POST.get("map_private_flag") == "on",
+        get_metadata_flag=request.POST.get("get_metadata_flag") == "on",
     )
 
     if import_file is None: