Fix Clickhouse identifiers format (#5890)

Co-authored-by: Pavel Strybuk <[email protected]>
Co-authored-by: Greg Finley <[email protected]>

src/sqlfluff/dialects/dialect_clickhouse.py

@@ -205,7 +205,7 @@
  NakedIdentifierSegment=SegmentGenerator(
  # Generate the anti template from the set of reserved keywords
  lambda dialect: RegexParser(
- r"[A-Z0-9_]*[A-Z][A-Z0-9_]*",
+ r"[a-zA-Z_][0-9a-zA-Z_]*",
  IdentifierSegment,
  type="naked_identifier",
  anti_template=r"^(" + r"|".join(dialect.sets("reserved_keywords")) + r")$",

test/fixtures/dialects/clickhouse/create_database.sql

@@ -1,10 +1,12 @@
 CREATE DATABASE db_name;
+CREATE DATABASE _123;
 CREATE DATABASE db_name COMMENT 'SingleQuoted';
 CREATE DATABASE db_name COMMENT "DoubleQuoted";
 CREATE DATABASE db_name COMMENT 'SingleQuoted three words';
 CREATE DATABASE db_name COMMENT "DoubleQuoted three words";
 CREATE DATABASE db_name COMMENT 'Weird characters: !@#$%^&*()_+{}|:"<>?';
 CREATE DATABASE db_name ON CLUSTER cluster;
+CREATE DATABASE db_name ON CLUSTER _123;
 CREATE DATABASE db_name ON CLUSTER "cluster";
 CREATE DATABASE db_name ON CLUSTER "underscore_cluster";
 CREATE DATABASE db_name ON CLUSTER 'cluster';

test/fixtures/dialects/clickhouse/create_database.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: 9e78f284bcd9f6853ca4d74d9fba0c6e2ad5e76d795a87763c4cf732346c8b8b
+_hash: 6f61dcb5b9c9926b99cc4981bf90c36d88637f220c3ec6a5db6f2b47c66262f5
 file:
 - statement:
  create_database_statement:
@@ -12,6 +12,13 @@ file:
  - database_reference:
  naked_identifier: db_name
 - statement_terminator: ;
+- statement:
+ create_database_statement:
+ - keyword: CREATE
+ - keyword: DATABASE
+ - database_reference:
+ naked_identifier: _123
+- statement_terminator: ;
 - statement:
  create_database_statement:
  - keyword: CREATE
@@ -68,6 +75,17 @@ file:
  - keyword: CLUSTER
  - naked_identifier: cluster
 - statement_terminator: ;
+- statement:
+ create_database_statement:
+ - keyword: CREATE
+ - keyword: DATABASE
+ - database_reference:
+ naked_identifier: db_name
+ - on_cluster_clause:
+ - keyword: 'ON'
+ - keyword: CLUSTER
+ - naked_identifier: _123
+- statement_terminator: ;
 - statement:
  create_database_statement:
  - keyword: CREATE

test/fixtures/dialects/clickhouse/create_table.sql

@@ -25,6 +25,8 @@ CREATE TABLE WatchLog_new(date Date, UserId Int64, EventType String, Cnt UInt64)
 
 CREATE TABLE WatchLog as WatchLog_old ENGINE=Merge(currentDatabase(), '^WatchLog');
 
+CREATE TABLE _2 as _1 ENGINE=Merge(currentDatabase(), '^WatchLog');
+
 CREATE TABLE hits_all AS hits
 ENGINE = Distributed(logs, default, hits)
 SETTINGS

test/fixtures/dialects/clickhouse/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: 612f5a6269c6f3b3edfb7afeee15e095dce9d4e4d6adfaeaab1ccab45eb525ce
+_hash: d1ae6fd707a9460859df0a7f42811f1bf45148a3ca3dd4c540fe7cffbc8b8100
 file:
 - statement:
  create_table_statement:
@@ -246,6 +246,36 @@ file:
  quoted_literal: "'^WatchLog'"
  - end_bracket: )
 - statement_terminator: ;
+- statement:
+ create_table_statement:
+ - keyword: CREATE
+ - keyword: TABLE
+ - table_reference:
+ naked_identifier: _2
+ - keyword: as
+ - table_reference:
+ naked_identifier: _1
+ - engine:
+ keyword: ENGINE
+ comparison_operator:
+ raw_comparison_operator: '='
+ table_engine_function:
+ function_name:
+ function_name_identifier: Merge
+ bracketed:
+ - start_bracket: (
+ - expression:
+ function:
+ function_name:
+ function_name_identifier: currentDatabase
+ bracketed:
+ start_bracket: (
+ end_bracket: )
+ - comma: ','
+ - expression:
+ quoted_literal: "'^WatchLog'"
+ - end_bracket: )
+- statement_terminator: ;
 - statement:
  create_table_statement:
  - keyword: CREATE

test/fixtures/dialects/clickhouse/identifier.sql

@@ -0,0 +1,9 @@
+SELECT * FROM _1.Table;
+
+SELECT * FROM _1._2;
+
+SELECT bar AS _1 FROM foo;
+
+SELECT * FROM foo.bar _1;
+
+SELECT * FROM foo.bar AS _1;

test/fixtures/dialects/clickhouse/identifier.yml

@@ -0,0 +1,102 @@
+# 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: 6e05286924da8f64b190fb0d8d1ceec52aa5437535002e0aefd97cb12452b851
+file:
+- statement:
+ select_statement:
+ select_clause:
+ keyword: SELECT
+ select_clause_element:
+ wildcard_expression:
+ wildcard_identifier:
+ star: '*'
+ from_clause:
+ keyword: FROM
+ from_expression:
+ from_expression_element:
+ table_expression:
+ table_reference:
+ - naked_identifier: _1
+ - dot: .
+ - naked_identifier: Table
+- statement_terminator: ;
+- statement:
+ select_statement:
+ select_clause:
+ keyword: SELECT
+ select_clause_element:
+ wildcard_expression:
+ wildcard_identifier:
+ star: '*'
+ from_clause:
+ keyword: FROM
+ from_expression:
+ from_expression_element:
+ table_expression:
+ table_reference:
+ - naked_identifier: _1
+ - dot: .
+ - naked_identifier: _2
+- statement_terminator: ;
+- statement:
+ select_statement:
+ select_clause:
+ keyword: SELECT
+ select_clause_element:
+ column_reference:
+ naked_identifier: bar
+ alias_expression:
+ keyword: AS
+ naked_identifier: _1
+ from_clause:
+ keyword: FROM
+ from_expression:
+ from_expression_element:
+ table_expression:
+ table_reference:
+ naked_identifier: foo
+- statement_terminator: ;
+- statement:
+ select_statement:
+ select_clause:
+ keyword: SELECT
+ select_clause_element:
+ wildcard_expression:
+ wildcard_identifier:
+ star: '*'
+ from_clause:
+ keyword: FROM
+ from_expression:
+ from_expression_element:
+ table_expression:
+ table_reference:
+ - naked_identifier: foo
+ - dot: .
+ - naked_identifier: bar
+ alias_expression:
+ naked_identifier: _1
+- statement_terminator: ;
+- statement:
+ select_statement:
+ select_clause:
+ keyword: SELECT
+ select_clause_element:
+ wildcard_expression:
+ wildcard_identifier:
+ star: '*'
+ from_clause:
+ keyword: FROM
+ from_expression:
+ from_expression_element:
+ table_expression:
+ table_reference:
+ - naked_identifier: foo
+ - dot: .
+ - naked_identifier: bar
+ alias_expression:
+ keyword: AS
+ naked_identifier: _1
+- statement_terminator: ;