Merge branch 'master' of ssh://git.house:/usr/local/git/base/kiosk

[kiosk.git] / bellevue_reporter_rss_renderer.py

diff --git a/bellevue_reporter_rss_renderer.py b/bellevue_reporter_rss_renderer.py
index 104147dfc170facc6008f4ecef4e5c3ad98be125..fec70e00d83cb62022f526c90515e35ad8b44165 100644 (file)
--- a/bellevue_reporter_rss_renderer.py
+++ b/bellevue_reporter_rss_renderer.py
@@ -1,12 +1,17 @@
 #!/usr/bin/env python3
 
+import logging
 import re
 from typing import List, Dict
 import xml
+import xml.etree.ElementTree as ET
 
 import generic_news_rss_renderer as gnrss
 
 
+logger = logging.getLogger(__name__)
+
+
 class bellevue_reporter_rss_renderer(gnrss.generic_news_rss_renderer):
     """Read the Bellevue Reporter's RSS feed."""
 
@@ -17,13 +22,7 @@ class bellevue_reporter_rss_renderer(gnrss.generic_news_rss_renderer):
         feed_uris: List[str],
         page_title: str,
     ):
-        super(bellevue_reporter_rss_renderer, self).__init__(
-            name_to_timeout_dict, feed_site, feed_uris, page_title
-        )
-        self.debug = True
-
-    def debug_prefix(self) -> str:
-        return "bellevue_reporter(%s)" % (self.page_title)
+        super().__init__(name_to_timeout_dict, feed_site, feed_uris, page_title)
 
     def get_headlines_page_prefix(self) -> str:
         return "bellevue-reporter"
@@ -34,7 +33,7 @@ class bellevue_reporter_rss_renderer(gnrss.generic_news_rss_renderer):
     def should_use_https(self) -> bool:
         return True
 
-    def munge_description(self, description: str) -> str:
+    def munge_description(self, description: str, item: ET.Element) -> str:
         description = re.sub("<[^>]+>", "", description)
         description = re.sub(
             "Bellevue\s+Reporter\s+Bellevue\s+Reporter", "", description
@@ -56,31 +55,68 @@ class bellevue_reporter_rss_renderer(gnrss.generic_news_rss_renderer):
     def looks_like_review(title: str, description: str) -> bool:
         return "review" in title or "Review" in title
 
+    @staticmethod
+    def looks_like_spam(title: str, description: str) -> bool:
+        return (
+            description is not None
+            and title is not None
+            and (
+                "marketplace" in description
+                or "national-marketplace" in description
+                or re.search("[Ww]eed", title) is not None
+                or re.search("[Cc]annabis", title) is not None
+                or re.search("[Cc]annabis", description) is not None
+                or "THC" in title
+                or re.search("[Tt]op.[Rr]ated", title) is not None
+                or re.search("[Ll]ose [Ww]eight", title) is not None
+                or re.search("[Ll]ose [Ww]eight", description) is not None
+            )
+        )
+
+    @staticmethod
+    def looks_very_boring(title: str, description: str) -> bool:
+        return description.lower() in title.lower()
+
     def item_is_interesting_for_headlines(
         self, title: str, description: str, item: xml.etree.ElementTree.Element
     ) -> bool:
+        unfiltered_description = item.findtext("description")
         if self.is_item_older_than_n_days(item, 10):
-            self.debug_print("%s: is too old!" % title)
+            logger.info(f"{title}: is too old!")
+            return False
+        if bellevue_reporter_rss_renderer.looks_like_spam(
+            title, unfiltered_description
+        ):
+            logger.debug(f"{title}: looks like spam")
             return False
         if bellevue_reporter_rss_renderer.looks_like_football(title, description):
-            self.debug_print("%s: looks like it's about football." % title)
+            logger.debug(f"{title}: looks like it's about football.")
             return False
         if bellevue_reporter_rss_renderer.looks_like_review(title, description):
-            self.debug_print("%s: looks like bullshit." % title)
+            logger.debug(f"{title}: looks like a review.")
+            return False
+        if bellevue_reporter_rss_renderer.looks_very_boring(title, description):
+            logger.debug(f"{title}: looks very boring.")
             return False
         return True
 
     def item_is_interesting_for_article(
         self, title: str, description: str, item: xml.etree.ElementTree.Element
     ) -> bool:
+        unfiltered_description = item.findtext("description")
         if self.is_item_older_than_n_days(item, 10):
-            self.debug_print("%s: is too old!" % title)
+            logger.debug(f"{title}: is too old!")
+            return False
+        if bellevue_reporter_rss_renderer.looks_like_spam(
+            title, unfiltered_description
+        ):
+            logger.debug(f"{title}: looks like spam")
             return False
         if bellevue_reporter_rss_renderer.looks_like_football(title, description):
-            self.debug_print("%s: looks like it's about football." % title)
+            logger.debug(f"{title}: looks like it's about football.")
             return False
         if bellevue_reporter_rss_renderer.looks_like_review(title, description):
-            self.debug_print("%s: looks like bullshit." % title)
+            logger.debug(f"{title}: looks like a review.")
             return False
         return True