Postgres: CREATE COLLATION (#3571)

* collate in table column

* Create collation

* delimited implies optional I guess

* Add note

* remove more optionals

* better tests

* Smush into a single test using UNIQUE instead of PK

* More tests

Co-authored-by: Barry Hart <barrywhart@yahoo.com>
Co-authored-by: Barry Pollard <barry_pollard@hotmail.com>

src/sqlfluff/dialects/dialect_postgres.py

@@ -1420,13 +1420,16 @@ class CreateTableStatementSegment(ansi.CreateTableStatementSegment):
                             Sequence(
                                 Ref("ColumnReferenceSegment"),
                                 Ref("DatatypeSegment"),
-                                Sequence(
-                                    "COLLATE",
-                                    Ref("QuotedLiteralSegment"),
-                                    optional=True,
-                                ),
                                 AnyNumberOf(
-                                    Ref("ColumnConstraintSegment", optional=True)
+                                    # A single COLLATE segment can come before or after
+                                    # constraint segments
+                                    OneOf(
+                                        Ref("ColumnConstraintSegment"),
+                                        Sequence(
+                                            "COLLATE",
+                                            Ref("ObjectReferenceSegment"),
+                                        ),
+                                    ),
                                 ),
                             ),
                             Ref("TableConstraintSegment"),
@@ -3318,6 +3321,7 @@ class StatementSegment(ansi.StatementSegment):
             Ref("AlterTypeStatementSegment"),
             Ref("AlterSchemaStatementSegment"),
             Ref("LockTableStatementSegment"),
+            Ref("CreateCollationStatementSegment"),
         ],
     )
 
@@ -4365,6 +4369,61 @@ class AlterTypeStatementSegment(BaseSegment):
     )
 
 
+class CreateCollationStatementSegment(BaseSegment):
+    """A `CREATE COLLATION` statement.
+
+    https://www.postgresql.org/docs/current/sql-createcollation.html
+    """
+
+    type = "create_collation_statement"
+    match_grammar: Matchable = Sequence(
+        "CREATE",
+        "COLLATION",
+        Ref("IfNotExistsGrammar", optional=True),
+        Ref("ObjectReferenceSegment"),
+        OneOf(
+            Bracketed(
+                Delimited(
+                    Sequence(
+                        "LOCALE",
+                        Ref("EqualsSegment"),
+                        Ref("QuotedLiteralSegment"),
+                    ),
+                    Sequence(
+                        "LC_COLLATE",
+                        Ref("EqualsSegment"),
+                        Ref("QuotedLiteralSegment"),
+                    ),
+                    Sequence(
+                        "LC_CTYPE",
+                        Ref("EqualsSegment"),
+                        Ref("QuotedLiteralSegment"),
+                    ),
+                    Sequence(
+                        "PROVIDER",
+                        Ref("EqualsSegment"),
+                        OneOf("ICU", "LIBC"),
+                    ),
+                    Sequence(
+                        "DETERMINISTIC",
+                        Ref("EqualsSegment"),
+                        Ref("BooleanLiteralGrammar"),
+                    ),
+                    Sequence(
+                        "VERSION",
+                        Ref("EqualsSegment"),
+                        Ref("QuotedLiteralSegment"),
+                    ),
+                )
+            ),
+            Sequence(
+                "FROM",
+                Ref("ObjectReferenceSegment"),
+            ),
+        ),
+    )
+
+
 class AlterSchemaStatementSegment(BaseSegment):
     """An `ALTER SCHEMA` statement.
 

src/sqlfluff/dialects/dialect_postgres_keywords.py

@@ -925,18 +925,21 @@ def get_keywords(keyword_list, keyword_type):
     ("DATE", "non-reserved"),
     ("DEPTH", "non-reserved"),
     ("DESCRIBE", "non-reserved"),
+    ("DETERMINISTIC", "non-reserved"),
     ("EXTENDED", "non-reserved"),
     ("FILE", "non-reserved"),
     ("FORCE_NOT_NULL", "non-reserved"),
     ("FORCE_NULL", "non-reserved"),
     ("FORCE_QUOTE", "non-reserved"),
     ("FORMAT", "non-reserved"),
     ("HASH", "non-reserved"),
+    ("ICU", "non-reserved"),
     ("IGNORE", "non-reserved"),
     ("IS_TEMPLATE", "non-reserved"),
     ("JSON", "non-reserved"),
     ("LC_COLLATE", "non-reserved"),
     ("LC_CTYPE", "non-reserved"),
+    ("LIBC", "non-reserved"),
     ("LIST", "non-reserved"),
     ("LOGIN", "non-reserved"),
     ("LOCALE", "non-reserved"),
@@ -950,6 +953,7 @@ def get_keywords(keyword_list, keyword_type):
     ("NOREPLICATION", "non-reserved"),
     ("NOSUPERUSER", "non-reserved"),
     ("PLAIN", "non-reserved"),
+    ("PROVIDER", "non-reserved"),
     ("PUBLIC", "non-reserved"),
     ("REMAINDER", "non-reserved"),
     ("REPLICATION", "non-reserved"),

test/fixtures/dialects/postgres/postgres_create_collation.sql

@@ -0,0 +1,7 @@
+CREATE COLLATION numeric (provider = icu, locale = 'en@colNumeric=yes');
+
+CREATE COLLATION french (locale = 'fr_FR.utf8');
+
+CREATE COLLATION german_phonebook (provider = icu, locale = 'de-u-co-phonebk');
+
+CREATE COLLATION german FROM "de_DE";

test/fixtures/dialects/postgres/postgres_create_collation.yml

@@ -0,0 +1,69 @@
+# YML test files are auto-generated from SQL files and should not be edited by
+# hand. To help enforce this, the "hash" field in the file must match a hash
+# computed by SQLFluff when running the tests. Please run
+# `python test/generate_parse_fixture_yml.py`  to generate them after adding or
+# altering SQL files.
+_hash: 309835f67eb282ebe597717e51871c763cbc54e4501ca3bda650a877633f1023
+file:
+- statement:
+    create_collation_statement:
+    - keyword: CREATE
+    - keyword: COLLATION
+    - object_reference:
+        identifier: numeric
+    - bracketed:
+      - start_bracket: (
+      - keyword: provider
+      - comparison_operator:
+          raw_comparison_operator: '='
+      - keyword: icu
+      - comma: ','
+      - keyword: locale
+      - comparison_operator:
+          raw_comparison_operator: '='
+      - literal: "'en@colNumeric=yes'"
+      - end_bracket: )
+- statement_terminator: ;
+- statement:
+    create_collation_statement:
+    - keyword: CREATE
+    - keyword: COLLATION
+    - object_reference:
+        identifier: french
+    - bracketed:
+        start_bracket: (
+        keyword: locale
+        comparison_operator:
+          raw_comparison_operator: '='
+        literal: "'fr_FR.utf8'"
+        end_bracket: )
+- statement_terminator: ;
+- statement:
+    create_collation_statement:
+    - keyword: CREATE
+    - keyword: COLLATION
+    - object_reference:
+        identifier: german_phonebook
+    - bracketed:
+      - start_bracket: (
+      - keyword: provider
+      - comparison_operator:
+          raw_comparison_operator: '='
+      - keyword: icu
+      - comma: ','
+      - keyword: locale
+      - comparison_operator:
+          raw_comparison_operator: '='
+      - literal: "'de-u-co-phonebk'"
+      - end_bracket: )
+- statement_terminator: ;
+- statement:
+    create_collation_statement:
+    - keyword: CREATE
+    - keyword: COLLATION
+    - object_reference:
+        identifier: german
+    - keyword: FROM
+    - object_reference:
+        identifier: '"de_DE"'
+- statement_terminator: ;

test/fixtures/dialects/postgres/postgres_create_table.sql

@@ -224,3 +224,11 @@ CREATE TABLE users (
     domain_id INTEGER REFERENCES groups (group_id) ON UPDATE RESTRICT,
     other_id INTEGER REFERENCES groups (group_id) MATCH SIMPLE
 );
+
+CREATE TABLE orders
+(
+id bigint NOT NULL DEFAULT NEXTVAL('orders_id_seq'::regclass),
+constraint_collate_constraints text UNIQUE COLLATE numeric NOT NULL PRIMARY KEY,
+constraints_collate text NOT NULL UNIQUE COLLATE numeric,
+collate_constraints text COLLATE numeric NOT NULL UNIQUE
+);

test/fixtures/dialects/postgres/postgres_create_table.yml

@@ -3,7 +3,7 @@
 # computed by SQLFluff when running the tests. Please run
 # `python test/generate_parse_fixture_yml.py`  to generate them after adding or
 # altering SQL files.
-_hash: 1ce86226be391d0503f1a45ce8367faeb5628f17a241abc85aa0af4e308ef3f9
+_hash: 15fd90958831ab1eeb765d759684c4f2521f9098f50e58befc6a29f31fcef957
 file:
 - statement:
     create_table_statement:
@@ -1336,3 +1336,76 @@ file:
         - keyword: SIMPLE
       - end_bracket: )
 - statement_terminator: ;
+- statement:
+    create_table_statement:
+    - keyword: CREATE
+    - keyword: TABLE
+    - table_reference:
+        identifier: orders
+    - bracketed:
+      - start_bracket: (
+      - column_reference:
+          identifier: id
+      - data_type:
+          keyword: bigint
+      - column_constraint_segment:
+        - keyword: NOT
+        - keyword: 'NULL'
+      - column_constraint_segment:
+          keyword: DEFAULT
+          function:
+            function_name:
+              function_name_identifier: NEXTVAL
+            bracketed:
+              start_bracket: (
+              expression:
+                literal: "'orders_id_seq'"
+                cast_expression:
+                  casting_operator: '::'
+                  data_type:
+                    data_type_identifier: regclass
+              end_bracket: )
+      - comma: ','
+      - column_reference:
+          identifier: constraint_collate_constraints
+      - data_type:
+          keyword: text
+      - column_constraint_segment:
+          keyword: UNIQUE
+      - keyword: COLLATE
+      - object_reference:
+          identifier: numeric
+      - column_constraint_segment:
+        - keyword: NOT
+        - keyword: 'NULL'
+      - column_constraint_segment:
+        - keyword: PRIMARY
+        - keyword: KEY
+      - comma: ','
+      - column_reference:
+          identifier: constraints_collate
+      - data_type:
+          keyword: text
+      - column_constraint_segment:
+        - keyword: NOT
+        - keyword: 'NULL'
+      - column_constraint_segment:
+          keyword: UNIQUE
+      - keyword: COLLATE
+      - object_reference:
+          identifier: numeric
+      - comma: ','
+      - column_reference:
+          identifier: collate_constraints
+      - data_type:
+          keyword: text
+      - keyword: COLLATE
+      - object_reference:
+          identifier: numeric
+      - column_constraint_segment:
+        - keyword: NOT
+        - keyword: 'NULL'
+      - column_constraint_segment:
+          keyword: UNIQUE
+      - end_bracket: )
+- statement_terminator: ;