pk as default sort key

CHANGES.txt

@@ -4,7 +4,9 @@ Changes
 0.10.0 (2024-02-19)
 ------------------
 - store known primary keys in repository (data/primary_keys.json)
-- add new function bcdata.get_primary_keys(), use remote url as default primary key lookup for bc2pg (#163)
+- add new function bcdata.get_primary_keys(), using remote url as default primary key lookup for bc2pg (#163)
+- when available, use known primary key as default sortkey (#165)
+- enable bcdata.get_sorkey() function
 - --refresh option cleanup, now available only via CLI
 
 0.9.2 (2024-02-18)

bcdata/__init__.py

@@ -6,6 +6,7 @@
     get_count,
     get_data,
     get_features,
+    get_sortkey,
     list_tables,
     validate_name,
 )

bcdata/wfs.py

@@ -222,17 +222,22 @@ def get_schema(self, table):
     def get_sortkey(self, table):
         """Check data for unique columns available for sorting paged requests"""
         columns = list(self.get_schema(table)["properties"].keys())
-        # use OBJECTID as default sort key, if present
-        if "OBJECTID" in columns:
+        # use known primary key if it is present in the bcdata repository
+        known_primary_keys = bcdata.get_primary_keys()
+        if table.lower() in known_primary_keys:
+            return known_primary_keys[table.lower()].upper()
+        # if pk not known, use OBJECTID as default sort key when present
+        elif "OBJECTID" in columns:
             return "OBJECTID"
         # if OBJECTID is not present (several GSR tables), use SEQUENCE_ID
         elif "SEQUENCE_ID" in columns:
             return "SEQUENCE_ID"
-        # otherwise, it should be safe to presume first column is the primary key
-        # (WHSE_FOREST_VEGETATION.VEG_COMP_LYR_R1_POLY's FEATURE_ID appears to be
-        # the only public case, and very large veg downloads are likely better
-        # accessed via some other channel)
+        # otherwise, presume first column is best value to sort by
+        # (in some cases this will be incorrect)
         else:
+            log.warning(
+                f"Reliable sort key for {table} cannot be determined, defaulting to first column {columns[0]}"
+            )
             return columns[0]
 
     def list_tables(self):
@@ -525,6 +530,12 @@ def get_features(
     )
 
 
+def get_sortkey(dataset):
+    WFS = BCWFS()
+    table = WFS.validate_name(dataset)
+    return WFS.get_sortkey(table)
+
+
 def list_tables(refresh=False):
     WFS = BCWFS(refresh)
     return WFS.list_tables()

tests/test_wfs.py

@@ -63,6 +63,14 @@ def test_get_count_bounds():
     )
 
 
+def test_get_sortkey_known():
+    assert bcdata.get_sortkey(ASSESSMENTS_TABLE) == "STREAM_CROSSING_ID"
+
+
+def test_get_sortkey_unknown():
+    assert bcdata.get_sortkey(AIRPORTS_TABLE) == "SEQUENCE_ID"
+
+
 def test_get_data_asgdf():
     gdf = bcdata.get_data(UTMZONES_KEY, query="UTM_ZONE=10", as_gdf=True)
     assert type(gdf) is GeoDataFrame