Postgres, Redshift: Support SIMILAR TO pattern matching expressions (#…

…2732) * Added first LIKE segment * See if this fixes recursive references * Added more segments * Committing PatternMatchingExpressionSegment so far * Added first LIKE segment * See if this fixes recursive references * Added more segments * Committing PatternMatchingExpressionSegment so far * Rolled back dev changes for the ANSI dialect * Copied Expression_A_Grammar to postgres * Added first LIKE segment * See if this fixes recursive references * Added more segments * Committing PatternMatchingExpressionSegment so far * Added first LIKE segment * See if this fixes recursive references * Added more segments * Committing PatternMatchingExpressionSegment so far * Rolled back dev changes for the ANSI dialect * Copied Expression_A_Grammar to postgres * Added Like operators * Updated postgres dialect * Updated implementation * Implemented SIMILAR TO * Added SIMILAR TO tests * Added POSIX Expression and Operator * Simplified implementation + added redshift tests * Linted * Added TODO * Added links to issues addressing TODOs * Simplified to LikeGrammar overwrite Co-authored-by: Barry Pollard <barry_pollard@hotmail.com> * Deleted bloat not needed due to LikeGrammar overwrite Co-authored-by: Barry Pollard <barry_pollard@hotmail.com> * Reverted LikeOperator changes in Postgres * Deleted unnecessary TODO * Updated yml hashes * Linted * Update src/sqlfluff/dialects/dialect_ansi.py Co-authored-by: Barry Pollard <barry_pollard@hotmail.com>
sqlfluff · Feb 28, 2022 · a1571ac · a1571ac
1 parent 413cd9e
commit a1571ac
Show file tree

Hide file tree

Showing 5 changed files with 1,027 additions and 0 deletions.
diff --git a/src/sqlfluff/dialects/dialect_postgres.py b/src/sqlfluff/dialects/dialect_postgres.py
@@ -202,6 +202,7 @@
 )
 
 postgres_dialect.replace(
+    LikeGrammar=OneOf("LIKE", "ILIKE", Sequence("SIMILAR", "TO")),
     ComparisonOperatorGrammar=OneOf(
         Ref("EqualsSegment"),
         Ref("GreaterThanSegment"),

diff --git a/test/fixtures/dialects/postgres/postgres_pattern_match_expressions.sql b/test/fixtures/dialects/postgres/postgres_pattern_match_expressions.sql
@@ -0,0 +1,72 @@
+-- postgres_pattern_match_expressions.sql
+/* examples of pattern match expressions
+( https://www.postgresql.org/docs/14/functions-matching.html ) that are
+supported in postgres. */
+
+-- LIKE/ILIKE expressions supported
+SELECT *
+FROM animals
+WHERE family LIKE '%ursidae%';
+
+SELECT *
+FROM animals
+WHERE family NOT LIKE '%ursidae%';
+
+SELECT *
+FROM animals
+WHERE genus ILIKE '%ursus%';
+
+SELECT *
+FROM animals
+WHERE genus NOT ILIKE '%ursus%';
+
+SELECT *
+FROM animals
+WHERE family LIKE '%ursidae%' ESCAPE '\\';
+
+SELECT *
+FROM animals
+WHERE genus NOT ILIKE '%ursus%' ESCAPE '\\';
+
+SELECT COALESCE(family LIKE '%ursidae%' ESCAPE '\\', FALSE) AS is_bear
+FROM animals;
+
+-- SIMILAR TO expressions supported
+SELECT *
+FROM animals
+WHERE family SIMILAR TO '%ursidae%';
+
+SELECT *
+FROM animals
+WHERE family NOT SIMILAR TO '%ursidae%';
+
+SELECT *
+FROM animals
+WHERE genus SIMILAR TO '%ursus%';
+
+SELECT *
+FROM animals
+WHERE genus NOT SIMILAR TO '%ursus%';
+
+SELECT *
+FROM animals
+WHERE family SIMILAR TO '%ursidae%' ESCAPE '\\';
+
+SELECT *
+FROM animals
+WHERE genus NOT SIMILAR TO '%ursus%' ESCAPE '\\';
+
+SELECT COALESCE(family SIMILAR TO '%ursidae%' ESCAPE '\\', FALSE) AS is_bear
+FROM animals;
+
+-- From https://github.com/sqlfluff/sqlfluff/issues/2722
+WITH cleaned_bear_financial_branch AS (
+    SELECT
+        branch_id,
+        TO_NUMBER(CASE WHEN honey_numerical_code SIMILAR TO '[0-9]{0,7}.?[0-9]{0,2}' THEN honey_numerical_code ELSE NULL END, '24601') AS honey_numerical_code
+    FROM bear_financial_branch
+)
+
+SELECT branch_id
+FROM cleaned_bear_financial_branch
+LIMIT 10;