[ao3] extract 'date_completed' and 'date_updated' metadata (#6013)

mikf · Sep 17, 2024 · 1d7df9c · 1d7df9c
1 parent 064bc89
commit 1d7df9c
Show file tree

Hide file tree

Showing 2 changed files with 7 additions and 0 deletions.
diff --git a/gallery_dl/extractor/ao3.py b/gallery_dl/extractor/ao3.py
@@ -70,6 +70,7 @@ def items(self):
         extr = text.extract_from(self.request(url).text)
 
         fmts = {}
+        path = ""
         download = extr(' class="download"', "</ul>")
         for dl in text.extract_iter(download, ' href="', "</"):
             path, _, type = dl.rpartition('">')
@@ -95,6 +96,10 @@ def items(self):
             "series"       : extr('<dd class="series">', "</dd>"),
             "date"         : text.parse_datetime(
                 extr('<dd class="published">', "<"), "%Y-%m-%d"),
+            "date_completed": text.parse_datetime(
+                extr('>Completed:</dt><dd class="status">', "<"), "%Y-%m-%d"),
+            "date_updated" : text.parse_timestamp(
+                path.rpartition("updated_at=")[2]),
             "words"        : text.parse_int(
                 extr('<dd class="words">', "<").replace(",", "")),
             "chapters"     : text.parse_int(

diff --git a/test/results/ao3.py b/test/results/ao3.py
@@ -19,6 +19,8 @@
     "chapters" : 27,
     "comments" : range(800, 2000),
     "date"     : "dt:2023-06-11 00:00:00",
+    "date_completed": "dt:2024-05-10 00:00:00",
+    "date_updated"  : "dt:2024-07-08 00:27:04",
     "extension": "pdf",
     "filename" : "The_Wildcard",
     "id"       : 47802076,