Merge pull request #663 from codders/feat/fix-immobiliare-20241217

Handle flats with more than 5 rooms on immobiliare
flathunters · Dec 17, 2024 · 44a8037 · 44a8037
2 parents 511001b + 71266a7
commit 44a8037
Show file tree

Hide file tree

Showing 2 changed files with 4 additions and 2 deletions.
diff --git a/flathunter/crawler/immobiliare.py b/flathunter/crawler/immobiliare.py
@@ -55,7 +55,7 @@ def extract_data(self, soup):
 
             detail_texts = [ item.find("span").text.strip() for item in details_list ]
             room_counts = [ match.group(1) for text in detail_texts
-                if (match := re.match(r"(\d+) local[ie]", text)) is not None ]
+                if (match := re.match(r"(\d+)\+? local[ie]", text)) is not None ]
             if len(room_counts) > 0:
                 rooms = room_counts[0]
             else:

diff --git a/test/crawler/test_crawl_immobiliare.py b/test/crawler/test_crawl_immobiliare.py
@@ -24,4 +24,6 @@ def test(self):
         self.assertTrue(entries[0]['url'].startswith(
             "https://www.immobiliare.it/annunci/"), u"URL should be an apartment link")
         for attr in ['title', 'price', 'size', 'rooms', 'address', 'image']:
-            self.assertIsNotNone(entries[0][attr], attr + " should be set")
+            self.assertIsNotNone(
+                entries[0][attr], attr + " should be set (" + entries[0]['url'] + ")"
+            )