First pass at CONCAT function (dask-contrib#253)

Author: charlesbluca

dask_sql/physical/rex/core/call.py

@@ -740,6 +740,7 @@ class RexCallPlugin(BaseRexPlugin):
         "truncate": Operation(da.trunc),
         # string operations
         "||": ReduceOperation(operation=operator.add),
+        "concat": ReduceOperation(operation=operator.add),
         "char_length": TensorScalarOperation(lambda x: x.str.len(), lambda x: len(x)),
         "upper": TensorScalarOperation(lambda x: x.str.upper(), lambda x: x.upper()),
         "lower": TensorScalarOperation(lambda x: x.str.lower(), lambda x: x.lower()),

tests/integration/test_rex.py

@@ -411,25 +411,26 @@ def test_string_functions(c):
         """
         SELECT
             a || 'hello' || a AS a,
-            CHAR_LENGTH(a) AS b,
-            UPPER(a) AS c,
-            LOWER(a) AS d,
-            POSITION('a' IN a FROM 4) AS e,
-            POSITION('ZL' IN a) AS f,
-            TRIM('a' FROM a) AS g,
-            TRIM(BOTH 'a' FROM a) AS h,
-            TRIM(LEADING 'a' FROM a) AS i,
-            TRIM(TRAILING 'a' FROM a) AS j,
-            OVERLAY(a PLACING 'XXX' FROM -1) AS k,
-            OVERLAY(a PLACING 'XXX' FROM 2 FOR 4) AS l,
-            OVERLAY(a PLACING 'XXX' FROM 2 FOR 1) AS m,
-            SUBSTRING(a FROM -1) AS n,
-            SUBSTRING(a FROM 10) AS o,
-            SUBSTRING(a FROM 2) AS p,
-            SUBSTRING(a FROM 2 FOR 2) AS q,
-            INITCAP(a) AS r,
-            INITCAP(UPPER(a)) AS s,
-            INITCAP(LOWER(a)) AS t
+            CONCAT(a, 'hello', a) as b,
+            CHAR_LENGTH(a) AS c,
+            UPPER(a) AS d,
+            LOWER(a) AS e,
+            POSITION('a' IN a FROM 4) AS f,
+            POSITION('ZL' IN a) AS g,
+            TRIM('a' FROM a) AS h,
+            TRIM(BOTH 'a' FROM a) AS i,
+            TRIM(LEADING 'a' FROM a) AS j,
+            TRIM(TRAILING 'a' FROM a) AS k,
+            OVERLAY(a PLACING 'XXX' FROM -1) AS l,
+            OVERLAY(a PLACING 'XXX' FROM 2 FOR 4) AS m,
+            OVERLAY(a PLACING 'XXX' FROM 2 FOR 1) AS n,
+            SUBSTRING(a FROM -1) AS o,
+            SUBSTRING(a FROM 10) AS p,
+            SUBSTRING(a FROM 2) AS q,
+            SUBSTRING(a FROM 2 FOR 2) AS r,
+            INITCAP(a) AS s,
+            INITCAP(UPPER(a)) AS t,
+            INITCAP(LOWER(a)) AS u
         FROM
             string_table
         """
@@ -438,25 +439,26 @@ def test_string_functions(c):
     expected_df = pd.DataFrame(
         {
             "a": ["a normal stringhelloa normal string"],
-            "b": [15],
-            "c": ["A NORMAL STRING"],
-            "d": ["a normal string"],
-            "e": [7],
-            "f": [0],
-            "g": [" normal string"],
+            "b": ["a normal stringhelloa normal string"],
+            "c": [15],
+            "d": ["A NORMAL STRING"],
+            "e": ["a normal string"],
+            "f": [7],
+            "g": [0],
             "h": [" normal string"],
             "i": [" normal string"],
-            "j": ["a normal string"],
-            "k": ["XXXormal string"],
-            "l": ["aXXXmal string"],
-            "m": ["aXXXnormal string"],
-            "n": ["a normal string"],
-            "o": ["string"],
-            "p": [" normal string"],
-            "q": [" n"],
-            "r": ["A Normal String"],
+            "j": [" normal string"],
+            "k": ["a normal string"],
+            "l": ["XXXormal string"],
+            "m": ["aXXXmal string"],
+            "n": ["aXXXnormal string"],
+            "o": ["a normal string"],
+            "p": ["string"],
+            "q": [" normal string"],
+            "r": [" n"],
             "s": ["A Normal String"],
             "t": ["A Normal String"],
+            "u": ["A Normal String"],
         }
     )