Fix DT layout change and stabilize HE publish dates

quasarr/downloads/sources/dt.py

@@ -33,15 +33,7 @@ def get_download_links(self, shared_state, url, mirrors, title, password):
             r.raise_for_status()
             soup = BeautifulSoup(r.text, "html.parser")
 
-            article = soup.find("article")
-            if not article:
-                info(f"Could not find article block for {title}")
-                mark_hostname_issue(
-                    Source.initials, "download", "Could not find article block"
-                )
-                return None
-
-            body = article.find("div", class_="full-story-body")
+            body = soup.find("div", class_="full-story-body")
             if not body:
                 info(f"Could not find download section for {title}")
                 mark_hostname_issue(

quasarr/providers/version.py

@@ -5,7 +5,7 @@
 import re
 import sys
 
-__version__ = "4.7.8"
+__version__ = "4.7.9"
 
 
 def get_version():

quasarr/search/sources/dt.py

@@ -80,7 +80,7 @@ def feed(
 
             for article in feed.find_all("article"):
                 try:
-                    link_tag = article.select_one("h1.dt-flexible-title a")
+                    link_tag = article.select_one(".dt-flexible-title a")
                     if not link_tag:
                         # ad/promo articles lack a release title link; skip quietly
                         debug(f"Link tag not found in article: {article}")
@@ -238,7 +238,7 @@ def search(
             for article in page.find_all("article"):
                 try:
                     # search-result articles use h1.font-weight-bold, while
-                    # the feed listing uses h1.dt-flexible-title
+                    # the feed listing uses .dt-flexible-title
                     link_tag = article.select_one("h1.font-weight-bold a")
                     if not link_tag:
                         debug(f"No title link in search-article: {article}")
@@ -340,24 +340,29 @@ def _extract_size(text):
 
 
 def _parse_published_datetime(article):
-    mon = article.select_one("small.dt-fs-11").text.strip()
-    day = article.find("h4", class_="m-0").text.strip()
-    year = article.find("h6", class_="m-0").text.strip()
-    month_num = datetime.datetime.strptime(mon, "%b").month
-
-    time_icon = article.select_one("i.bi-clock")
-    if time_icon:
-        # its parent <span> contains e.g. "19:12"
-        raw = time_icon.parent.get_text(strip=True)
-        m = re.search(r"(\d{1,2}:\d{2})", raw)
-        if m:
-            hh, mm = map(int, m.group(1).split(":"))
-        else:
-            hh, mm = 0, 0
+    month_tag = article.select_one("small.dt-fs-11")
+    if not month_tag:
+        # feed listings show "29.10.2026 07:47" instead of the calendar block
+        m = re.search(
+            r"\b(\d{2})\.(\d{2})\.(\d{4})\s+(\d{1,2}):(\d{2})\b",
+            article.get_text(" "),
+        )
+        day, month_num, year, hh, mm = map(int, m.groups())
     else:
+        day = int(article.find("h4", class_="m-0").text.strip())
+        year = int(article.find("h6", class_="m-0").text.strip())
+        month_num = datetime.datetime.strptime(month_tag.text.strip(), "%b").month
+
+        time_icon = article.select_one("i.bi-clock")
         hh, mm = 0, 0
+        if time_icon:
+            # its parent <span> contains e.g. "19:12"
+            raw = time_icon.parent.get_text(strip=True)
+            m = re.search(r"(\d{1,2}:\d{2})", raw)
+            if m:
+                hh, mm = map(int, m.group(1).split(":"))
 
     # this timezone is fixed to CET+1 and might be wrong
     cet = timezone(timedelta(hours=1))
-    dt = datetime.datetime(int(year), month_num, int(day), hh, mm, tzinfo=cet)
+    dt = datetime.datetime(year, month_num, day, hh, mm, tzinfo=cet)
     return dt.isoformat()

quasarr/search/sources/he.py

@@ -188,6 +188,8 @@ def search(
                         search_type,
                         str(e) if "e" in dir() else "Error occurred",
                     )
+                # the relative "Posted N days ago" drifts with every search
+                published = _parse_published_meta(soup) or published
                 release_imdb_id = None
                 try:
                     imdb_link = soup.find(
@@ -245,6 +247,17 @@ def search(
         return releases
 
 
+def _parse_published_meta(soup):
+    meta = soup.find("meta", attrs={"itemprop": "datePublished"})
+    try:
+        published = datetime.fromisoformat(meta["content"])
+        return published.astimezone(timezone.utc).strftime(
+            "%a, %d %b %Y %H:%M:%S +0000"
+        )
+    except Exception:
+        return ""
+
+
 def _parse_posted_ago(txt):
     try:
         m = re.search(

tests/test_dt_source.py

@@ -0,0 +1,103 @@
+# -*- coding: utf-8 -*-
+
+import unittest
+from unittest.mock import MagicMock, patch
+
+from bs4 import BeautifulSoup
+
+from quasarr.downloads.sources import dt
+from quasarr.search.sources.dt import _parse_published_datetime
+
+
+def _response(html):
+    response = MagicMock()
+    response.text = html
+    return response
+
+
+class DtDownloadLinksTests(unittest.TestCase):
+    def _get_links(self, html):
+        shared_state = MagicMock()
+        shared_state.values = {"user_agent": "test-agent"}
+        session = MagicMock()
+        session.get.return_value = _response(html)
+        with (
+            patch.object(dt.requests, "Session", return_value=session),
+            patch.object(dt, "clear_hostname_issue"),
+            patch.object(dt, "mark_hostname_issue"),
+        ):
+            return dt.Source().get_download_links(
+                shared_state,
+                "https://source.invalid/post.html",
+                None,
+                "Example.Show.S01E01.1080p.WEB.h264-GROUP",
+                "source.invalid",
+            )
+
+    def test_finds_story_body_outside_article(self):
+        result = self._get_links(
+            """
+            <article><div class="h1 dt-flexible-title">Example</div></article>
+            <section><div class="card-body"><div class="full-story-body">
+                <a href="https://rapidgator.invalid/file/one">one</a>
+                <a href="https://www.imdb.com/title/tt0000001/">imdb</a>
+            </div></div></section>
+            """
+        )
+
+        self.assertEqual(
+            {"links": [["https://rapidgator.invalid/file/one", "rapidgator"]]},
+            result,
+        )
+
+    def test_finds_story_body_inside_article(self):
+        result = self._get_links(
+            """
+            <article><div class="full-story-body">
+                <a href="https://ddownload.invalid/file/two">two</a>
+            </div></article>
+            """
+        )
+
+        self.assertEqual(
+            {"links": [["https://ddownload.invalid/file/two", "ddownload"]]},
+            result,
+        )
+
+
+class DtPublishedDateTests(unittest.TestCase):
+    def _parse(self, html):
+        article = BeautifulSoup(html, "html.parser").find("article")
+        return _parse_published_datetime(article)
+
+    def test_parses_feed_listing_date(self):
+        self.assertEqual(
+            "2026-10-29T07:47:00+01:00",
+            self._parse(
+                """
+                <article>
+                    <span>Today, 07:47</span>
+                    <span>29.10.2026 07:47</span>
+                </article>
+                """
+            ),
+        )
+
+    def test_parses_search_calendar_block(self):
+        self.assertEqual(
+            "2026-10-24T14:39:00+01:00",
+            self._parse(
+                """
+                <article>
+                    <small class="dt-fs-11">Oct</small>
+                    <h4 class="m-0">24</h4>
+                    <h6 class="m-0">2026</h6>
+                    <span><i class="bi bi-clock"></i>14:39</span>
+                </article>
+                """
+            ),
+        )
+
+
+if __name__ == "__main__":
+    unittest.main()

tests/test_he_published_date.py

@@ -0,0 +1,59 @@
+# -*- coding: utf-8 -*-
+
+import time
+import unittest
+from unittest.mock import MagicMock, patch
+
+from quasarr.search.sources import he
+
+POST_URL = "https://source.invalid/synthetic-show-s01e01/"
+
+SEARCH_HTML = f"""
+<div class="item"><div class="data">
+  <h5><a href="{POST_URL}">Synthetic.Show.S01E01.1080p.WEB.h264-GRP – 1.2 GB</a></h5>
+  <span class="post-time">Posted 2 days ago</span>
+</div></div>
+"""
+
+
+def _response(html):
+    response = MagicMock()
+    response.content = html.encode("utf-8")
+    return response
+
+
+class HePublishedDateTests(unittest.TestCase):
+    def _search(self, post_html):
+        shared_state = MagicMock()
+        shared_state.values = {
+            "user_agent": "test-agent",
+            "config": lambda section: {"he": "source.invalid"},
+        }
+
+        def get(url, **kwargs):
+            return _response(post_html if url == POST_URL else SEARCH_HTML)
+
+        with (
+            patch.object(he.requests, "get", side_effect=get),
+            patch.object(he, "generate_download_link", return_value="link"),
+            patch.object(he, "clear_hostname_issue"),
+            patch.object(he, "mark_hostname_issue"),
+        ):
+            releases = he.Source().search(shared_state, time.time(), 5000)
+        return [release["details"]["date"] for release in releases]
+
+    def test_uses_absolute_published_date_from_post(self):
+        self.assertEqual(
+            ["Thu, 23 Oct 2031 10:11:40 +0000"],
+            self._search(
+                '<meta itemprop="datePublished" content="2031-10-23T13:41:40+03:30"/>'
+            ),
+        )
+
+    def test_falls_back_to_relative_post_time(self):
+        with patch.object(he, "_parse_posted_ago", return_value="relative"):
+            self.assertEqual(["relative"], self._search("<div></div>"))
+
+
+if __name__ == "__main__":
+    unittest.main()