Merge pull request #25 from CartoDB/new_normalize_census_name

New normalize census name

src/pg/sql/40_observatory_utility.sql

@@ -173,3 +173,23 @@ BEGIN
   return result;
 END;
 $$ LANGUAGE plpgsql;
+
+-- Function that replaces all non digits or letters with _ trims and lowercases the
+-- passed measure name
+
+CREATE OR REPLACE FUNCTION cdb_observatory._OBS_StandardizeMeasureName(measure_name text)
+RETURNS text
+AS $$
+DECLARE
+  result text;
+BEGIN
+  -- Turn non letter or digits to _
+  result = regexp_replace(measure_name, '[^\dA-Za-z]+','_', 'g');
+  -- Remove duplicate _'s
+  result = regexp_replace(result,'_{2,}','_', 'g');
+  -- Trim _'s from beginning and end
+  result = trim(both  '_' from result);
+  result = lower(result);
+  RETURN result;
+END;
+$$ LANGUAGE plpgsql;

src/pg/test/expected/40_observatory_utility_test.out

@@ -140,6 +140,13 @@ SELECT cdb_observatory._OBS_GetRelatedColumn(
  {"\"es.ine\".total_pop",NULL,"\"us.census.acs\".B01001001"}
 (1 row)
 
+-- should give back a standardized measure name
+SELECT cdb_observatory._OBS_StandardizeMeasureName('test 343 %% 2 qqq }}{{}}');
+ _obs_standardizemeasurename 
+-----------------------------
+ test_343_2_qqq
+(1 row)
+
 \i test/sql/drop_fixtures.sql
 SET client_min_messages TO NOTICE;
 \set ECHO none

src/pg/test/sql/40_observatory_utility_test.sql

@@ -84,4 +84,7 @@ SELECT cdb_observatory._OBS_GetRelatedColumn(
      'denominator'
  );
 
+-- should give back a standardized measure name
+SELECT cdb_observatory._OBS_StandardizeMeasureName('test 343 %% 2 qqq }}{{}}');
+
 \i test/sql/drop_fixtures.sql