Add initial Redshift support

docs/source/dialects.rst

@@ -94,12 +94,20 @@ PostgreSQL
 ----------
 
 This is based around the `PostgreSQL spec`_. Many other SQL instances are often
-based on PostreSQL syntax. If you're running `AWS Redshift`_ or `Greenplum`_
-this is the dialect to use (until someone makes a specific dialect).
+based on PostreSQL syntax. If you're running an unsupported dialect, then
+this is often the dialect to use (until someone makes a specific dialect).
 
 .. _`PostgreSQL spec`: https://www.postgresql.org/docs/9.6/reference.html
-.. _`AWS Redshift`: https://aws.amazon.com/redshift/
-.. _`Greenplum`: https://greenplum.org/
+
+.. _redshift_dialect_ref:
+
+Redshift
+----------
+
+
+The dialect for `Amazon Redshift`_.
+
+.. _`Amazon Redshift`: https://aws.amazon.com/redshift/
 
 .. _snowflake_dialect_ref:
 

setup.py

@@ -51,6 +51,7 @@ def read(*names, **kwargs):
         "hive",
         "mysql",
         "postgres",
+        "redshift",
         "snowflake",
         "sqlite",
         "teradata",

src/sqlfluff/core/dialects/__init__.py

@@ -23,14 +23,15 @@
 _dialect_lookup = {
     "ansi": ("dialect_ansi", "ansi_dialect"),
     "bigquery": ("dialect_bigquery", "bigquery_dialect"),
+    "exasol": ("dialect_exasol", "exasol_dialect"),
+    "hive": ("dialect_hive", "hive_dialect"),
     "mysql": ("dialect_mysql", "mysql_dialect"),
-    "teradata": ("dialect_teradata", "teradata_dialect"),
     "postgres": ("dialect_postgres", "postgres_dialect"),
+    "redshift": ("dialect_redshift", "redshift_dialect"),
     "snowflake": ("dialect_snowflake", "snowflake_dialect"),
-    "exasol": ("dialect_exasol", "exasol_dialect"),
-    "tsql": ("dialect_tsql", "tsql_dialect"),
-    "hive": ("dialect_hive", "hive_dialect"),
     "sqlite": ("dialect_sqlite", "sqlite_dialect"),
+    "teradata": ("dialect_teradata", "teradata_dialect"),
+    "tsql": ("dialect_tsql", "tsql_dialect"),
 }
 
 _legacy_dialects = {

src/sqlfluff/dialects/dialect_redshift.py

@@ -0,0 +1,56 @@
+"""The Amazon Redshift dialect.
+
+This is based on postgres dialect, since it was initially based off of Postgres 8.
+We should monitor in future and see if it should be rebased off of ANSI
+"""
+
+from sqlfluff.core.parser import (
+    OneOf,
+    BaseSegment,
+)
+
+from sqlfluff.core.dialects import load_raw_dialect
+
+from sqlfluff.dialects.dialect_redshift_keywords import (
+    redshift_reserved_keywords,
+    redshift_unreserved_keywords,
+)
+
+
+postgres_dialect = load_raw_dialect("postgres")
+ansi_dialect = load_raw_dialect("ansi")
+
+redshift_dialect = postgres_dialect.copy_as("redshift")
+
+# Set Keywords
+redshift_dialect.sets("unreserved_keywords").clear()
+redshift_dialect.sets("unreserved_keywords").update(
+    [n.strip().upper() for n in redshift_unreserved_keywords.split("\n")]
+)
+
+redshift_dialect.sets("reserved_keywords").clear()
+redshift_dialect.sets("reserved_keywords").update(
+    [n.strip().upper() for n in redshift_reserved_keywords.split("\n")]
+)
+
+
+@redshift_dialect.segment(replace=True)
+class DateAddFunctionNameSegment(BaseSegment):
+    """DATEADD function name segment.
+
+    Override to support DATEDIFF as well
+    """
+
+    type = "function_name"
+    match_grammar = OneOf("DATEADD", "DATEDIFF")
+
+
+@redshift_dialect.segment(replace=True)
+class FunctionSegment(BaseSegment):
+    """A scalar or aggregate function.
+
+    Revert back to the ANSI definition to support ignore nulls
+    """
+
+    type = "function"
+    match_grammar = ansi_dialect.get_segment("FunctionSegment").match_grammar.copy()