[#137] Support the new blocking_columns format in potential_matches.sql

ipums · Jun 5, 2024 · f637500 · f637500
1 parent 83f92c4
commit f637500
Show file tree

Hide file tree

Showing 2 changed files with 3 additions and 3 deletions.
diff --git a/hlink/linking/matching/link_step_match.py b/hlink/linking/matching/link_step_match.py
@@ -46,7 +46,7 @@ def _run(self):
                     config["id_column"],
                 )
 
-        t_ctx["blocking_columns"] = [bc["column_name"] for bc in blocking]
+        t_ctx["blocking_columns"] = [[bc["column_name"]] for bc in blocking]
 
         blocking_exploded_columns = [
             bc["column_name"] for bc in blocking if "explode" in bc and bc["explode"]

diff --git a/hlink/linking/matching/templates/potential_matches.sql b/hlink/linking/matching/templates/potential_matches.sql
@@ -15,8 +15,8 @@ SELECT DISTINCT
 {% endif %}
 FROM exploded_df_a a
 JOIN exploded_df_b b ON 
-{% for col in blocking_columns %}
-a.{{ col }} = b.{{ col }} {{ "AND" if not loop.last }}
+{% for or_group in blocking_columns %}
+    ({% for col in or_group %}a.{{ col }} = b.{{ col }}{{ " OR " if not loop.last }}{% endfor %}) {{ "AND" if not loop.last }}
 {% endfor %}
 {% if distance_table %}
   {% for d in distance_table %}