Add breaking news feature

mas-4 · Aug 24, 2024 · d440bb4 · d440bb4
1 parent 43c0041
commit d440bb4
Show file tree

Hide file tree

Showing 2 changed files with 13 additions and 1 deletion.
diff --git a/app/site/data.py b/app/site/data.py
@@ -186,6 +186,7 @@ def get_main_headline_df():
             'title': Headline.processed,
             'agency': Agency.name,
             'bias': Agency._bias,  # noqa prot attr
+            'appearance': Article.first_accessed,
             'first_accessed': Headline.first_accessed,
             'last_accessed': Headline.last_accessed,
             'position': Headline.position,
@@ -213,6 +214,9 @@ def get_main_headline_df():
         if len(df) == 0:
             sys.exit("No headlines found. Exiting.")
 
+        df['appearance'] = df['appearance'].dt.tz_localize('utc').dt.tz_convert('US/Eastern')
+        now = pd.Timestamp.now(tz='US/Eastern')
+        df['howlong'] = (now - df['appearance']).dt.total_seconds()
         df['first_accessed'] = df['first_accessed'].dt.tz_localize('utc').dt.tz_convert('US/Eastern')
         df['last_accessed'] = df['last_accessed'].dt.tz_localize('utc').dt.tz_convert('US/Eastern')
         roundcol = ['vader_compound', 'afinn', 'topic_score']

diff --git a/app/site/page_headlines.py b/app/site/page_headlines.py
@@ -104,8 +104,10 @@ def cluster_and_summarize(self, df):
         clusters_list = [{'cluster': key, 'data': group.to_dict(orient='records')} for key, group in grouped]
         for cluster in clusters_list:
             cluster['coverage'] = round(len(cluster['data']) / len(Scrapers) * 100, 2)
+            cluster['first'] = max(cluster['data'], key=lambda x: x['howlong'])['howlong']
 
-        clusters_list.sort(key=lambda x: len(x['data']), reverse=True)
+        # clusters_list.sort(key=lambda x: len(x['data']), reverse=True)
+        clusters_list.sort(key=lambda x: x['first'])
         self.make_agency_lists(clusters_list)
         self.context['clusters'] = clusters_list
 
@@ -114,6 +116,12 @@ def make_agency_lists(self, clusters_list):
         for cluster in clusters_list:
             cluster['data'].sort(key=lambda x: x['agency'])
             hrefs = [f"<p>{len(cluster['data'])} headlines / {cluster['coverage']}% coverage</p>"]
+            # if cluster['first'] / 60 < 4 * 60:
+            minutes = cluster['first'] // 60
+            if minutes < 90:
+                hrefs[-1] = hrefs[-1] + f'<h3>🚨🚨🚨BREAKING! {int(minutes)}m ago! </h3>'
+            else:
+                hrefs[-1] = hrefs[-1] + f'<p>First broken {int(cluster["first"] // 60 // 60)} hours ago</p>'
             last_bias = -3
             for a in sorted(cluster['data'], key=lambda x: x['bias']):
                 if a['bias'] != last_bias: