- data cleaning & CONDOP bug fixes

homeharvest/__init__.py

@@ -53,7 +53,9 @@ def scrape_property(
     if not properties_dfs:
         return pd.DataFrame()
 
+    properties_dfs = [df for df in properties_dfs if not df.empty]
+
     with warnings.catch_warnings():
         warnings.simplefilter("ignore", category=FutureWarning)
 
-        return pd.concat(properties_dfs, ignore_index=True, axis=0)[ordered_properties].replace({"None": "", None: ""})
+        return pd.concat(properties_dfs, ignore_index=True, axis=0)[ordered_properties].replace({"None": pd.NA, None: pd.NA, "": pd.NA})

homeharvest/core/scrapers/models.py

@@ -36,6 +36,7 @@ class PropertyType(Enum):
     CONDO_TOWNHOME = "CONDO_TOWNHOME"
     CONDO_TOWNHOME_ROWHOME_COOP = "CONDO_TOWNHOME_ROWHOME_COOP"
     CONDO = "CONDO"
+    CONDOP = "CONDOP"
     CONDOS = "CONDOS"
     COOP = "COOP"
     DUPLEX_TRIPLEX = "DUPLEX_TRIPLEX"
@@ -52,6 +53,7 @@ class PropertyType(Enum):
 
 @dataclass
 class Address:
+    full_line: str | None = None
     street: str | None = None
     unit: str | None = None
     city: str | None = None

homeharvest/core/scrapers/realtor/__init__.py

@@ -52,6 +52,7 @@ class RealtorScraper(Scraper):
                             listing_id
                         }
                         address {
+                            line
                             street_direction
                             street_number
                             street_name
@@ -236,6 +237,7 @@ class RealtorScraper(Scraper):
                             stories
                         }
                         address {
+                            line
                             street_direction
                             street_number
                             street_name
@@ -352,6 +354,7 @@ class RealtorScraper(Scraper):
                                         street_number
                                         street_name
                                         street_suffix
+                                        line
                                         unit
                                         city
                                         state_code
@@ -657,6 +660,8 @@ class RealtorScraper(Scraper):
         if not self.extra_property_data:
             return {}
 
+        #: TODO: migrate "advertisers" and "estimates" to general query
+
         query = """query GetHome($property_id: ID!) {
                     home(property_id: $property_id) {
                         __typename
@@ -765,6 +770,7 @@ class RealtorScraper(Scraper):
             address = result["address"]
 
         return Address(
+            full_line=address.get("line"),
             street=" ".join(
                 part for part in [
                     address.get("street_number"),
@@ -800,7 +806,7 @@ class RealtorScraper(Scraper):
         return Description(
             primary_photo=primary_photo,
             alt_photos=RealtorScraper.process_alt_photos(result.get("photos")),
-            style=PropertyType(style) if style else None,
+            style=PropertyType.__getitem__(style) if style and style in PropertyType.__members__ else None,
             beds=description_data.get("beds"),
             baths_full=description_data.get("baths_full"),
             baths_half=description_data.get("baths_half"),

homeharvest/utils.py

@@ -10,6 +10,7 @@ ordered_properties = [
     "status",
     "text",
     "style",
+    "full_street_line",
     "street",
     "unit",
     "city",
@@ -55,6 +56,7 @@ def process_result(result: Property) -> pd.DataFrame:
 
     if "address" in prop_data:
         address_data = prop_data["address"]
+        prop_data["full_street_line"] = address_data.full_line
         prop_data["street"] = address_data.street
         prop_data["unit"] = address_data.unit
         prop_data["city"] = address_data.city

pyproject.toml

@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "homeharvest"
-version = "0.3.22"
+version = "0.3.24"
 description = "Real estate scraping library"
 authors = ["Zachary Hampton <zachary@bunsly.com>", "Cullen Watson <cullen@bunsly.com>"]
 homepage = "https://github.com/Bunsly/HomeHarvest"

tests/test_realtor.py

@@ -156,3 +156,12 @@ def test_realtor_without_extra_details():
     ]
 
     assert results[0] != results[1]
+
+
+def test_pr_zip_code():
+    results = scrape_property(
+        location="00741",
+        listing_type="for_sale",
+    )
+
+    assert results is not None and len(results) > 0