|
| 1 | +// Copyright 2021 Datafuse Labs |
| 2 | +// |
| 3 | +// Licensed under the Apache License, Version 2.0 (the "License"); |
| 4 | +// you may not use this file except in compliance with the License. |
| 5 | +// You may obtain a copy of the License at |
| 6 | +// |
| 7 | +// http://www.apache.org/licenses/LICENSE-2.0 |
| 8 | +// |
| 9 | +// Unless required by applicable law or agreed to in writing, software |
| 10 | +// distributed under the License is distributed on an "AS IS" BASIS, |
| 11 | +// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. |
| 12 | +// See the License for the specific language governing permissions and |
| 13 | +// limitations under the License. |
| 14 | +#[cfg(test)] |
| 15 | +mod tests { |
| 16 | + |
| 17 | + use databend_common_ast::ast::Identifier; |
| 18 | + use databend_common_ast::parser::parse_sql; |
| 19 | + use databend_common_ast::parser::tokenize_sql; |
| 20 | + use databend_common_ast::parser::Dialect; |
| 21 | + use databend_common_exception::Result; |
| 22 | + |
| 23 | + use crate::planner::semantic::name_resolution::normalize_identifier; |
| 24 | + use crate::planner::semantic::name_resolution::NameResolutionContext; |
| 25 | + |
| 26 | + #[test] |
| 27 | + fn test_sql_to_ast_to_sql_quoted_identifiers() -> Result<()> { |
| 28 | + // Test case 1: Simple WITH clause with quoted identifiers |
| 29 | + let sql = r#"WITH "QuotedCTE" AS (SELECT * FROM "QuotedTable" WHERE id = 1) SELECT * FROM "QuotedCTE""#; |
| 30 | + |
| 31 | + // Parse SQL to AST |
| 32 | + // let tokens = sql.to_string(); |
| 33 | + let tokens = tokenize_sql(sql)?; |
| 34 | + let (stmt, _) = parse_sql(&tokens, Dialect::PostgreSQL)?; |
| 35 | + |
| 36 | + // Convert AST back to SQL |
| 37 | + let regenerated_sql = stmt.to_string(); |
| 38 | + |
| 39 | + // The regenerated SQL should still contain the quoted identifiers |
| 40 | + assert!( |
| 41 | + regenerated_sql.contains(r#""QuotedCTE""#), |
| 42 | + "Quoted CTE name lost in regenerated SQL" |
| 43 | + ); |
| 44 | + assert!( |
| 45 | + regenerated_sql.contains(r#""QuotedTable""#), |
| 46 | + "Quoted table name lost in regenerated SQL" |
| 47 | + ); |
| 48 | + |
| 49 | + // Test case 2: More complex WITH clause with mixed case identifiers |
| 50 | + let sql2 = r#"WITH "MixedCaseCTE" AS (SELECT id, "Value" FROM "MixedCaseTable" WHERE id = 1) SELECT * FROM "MixedCaseCTE""#; |
| 51 | + |
| 52 | + // Parse SQL to AST |
| 53 | + // let tokens2 = sql2.to_string(); |
| 54 | + let tokens2 = tokenize_sql(sql2)?; |
| 55 | + let (stmt2, _) = parse_sql(&tokens2, Dialect::PostgreSQL)?; |
| 56 | + |
| 57 | + // Convert AST back to SQL |
| 58 | + let regenerated_sql2 = stmt2.to_string(); |
| 59 | + |
| 60 | + // The regenerated SQL should still contain the quoted identifiers with correct case |
| 61 | + assert!( |
| 62 | + regenerated_sql2.contains(r#""MixedCaseCTE""#), |
| 63 | + "Quoted CTE name lost in regenerated SQL" |
| 64 | + ); |
| 65 | + assert!( |
| 66 | + regenerated_sql2.contains(r#""Value""#), |
| 67 | + "Quoted column name lost in regenerated SQL" |
| 68 | + ); |
| 69 | + assert!( |
| 70 | + regenerated_sql2.contains(r#""MixedCaseTable""#), |
| 71 | + "Quoted table name lost in regenerated SQL" |
| 72 | + ); |
| 73 | + |
| 74 | + Ok(()) |
| 75 | + } |
| 76 | + |
| 77 | + #[test] |
| 78 | + fn test_normalize_identifier_preserves_quotes() -> Result<()> { |
| 79 | + // Create a context where case sensitivity is off for both quoted and unquoted identifiers |
| 80 | + let context = NameResolutionContext { |
| 81 | + unquoted_ident_case_sensitive: false, |
| 82 | + quoted_ident_case_sensitive: false, |
| 83 | + deny_column_reference: false, |
| 84 | + }; |
| 85 | + |
| 86 | + // Test with a quoted identifier |
| 87 | + let quoted_ident = Identifier { |
| 88 | + span: Default::default(), |
| 89 | + name: "MixedCase".to_string(), |
| 90 | + quote: Some('"'), |
| 91 | + ident_type: Default::default(), |
| 92 | + }; |
| 93 | + |
| 94 | + // Normalize the identifier |
| 95 | + let normalized = normalize_identifier("ed_ident, &context); |
| 96 | + |
| 97 | + // Check that the quote information is preserved |
| 98 | + assert_eq!( |
| 99 | + normalized.quote, |
| 100 | + Some('"'), |
| 101 | + "Quote information was lost during normalization" |
| 102 | + ); |
| 103 | + assert_eq!( |
| 104 | + normalized.name, "mixedcase", |
| 105 | + "Name should be lowercase after normalization" |
| 106 | + ); |
| 107 | + |
| 108 | + // Test with an unquoted identifier |
| 109 | + let unquoted_ident = Identifier { |
| 110 | + span: Default::default(), |
| 111 | + name: "MixedCase".to_string(), |
| 112 | + quote: None, |
| 113 | + ident_type: Default::default(), |
| 114 | + }; |
| 115 | + |
| 116 | + // Normalize the identifier |
| 117 | + let normalized = normalize_identifier(&unquoted_ident, &context); |
| 118 | + |
| 119 | + // Check that it remains unquoted |
| 120 | + assert_eq!( |
| 121 | + normalized.quote, None, |
| 122 | + "Unquoted identifier should remain unquoted" |
| 123 | + ); |
| 124 | + assert_eq!( |
| 125 | + normalized.name, "mixedcase", |
| 126 | + "Name should be lowercase after normalization" |
| 127 | + ); |
| 128 | + |
| 129 | + Ok(()) |
| 130 | + } |
| 131 | + |
| 132 | + #[test] |
| 133 | + fn test_multiple_normalizations() -> Result<()> { |
| 134 | + // This test simulates the scenario where an identifier is normalized multiple times |
| 135 | + // which could happen during multiple parsing/planning phases |
| 136 | + |
| 137 | + // Create contexts with different case sensitivity settings |
| 138 | + let case_insensitive = NameResolutionContext { |
| 139 | + unquoted_ident_case_sensitive: false, |
| 140 | + quoted_ident_case_sensitive: false, |
| 141 | + deny_column_reference: false, |
| 142 | + }; |
| 143 | + |
| 144 | + let mixed_sensitivity = NameResolutionContext { |
| 145 | + unquoted_ident_case_sensitive: false, |
| 146 | + quoted_ident_case_sensitive: true, |
| 147 | + deny_column_reference: false, |
| 148 | + }; |
| 149 | + |
| 150 | + // Start with a quoted identifier |
| 151 | + let original = Identifier { |
| 152 | + span: Default::default(), |
| 153 | + name: "MixedCaseIdentifier".to_string(), |
| 154 | + quote: Some('"'), |
| 155 | + ident_type: Default::default(), |
| 156 | + }; |
| 157 | + |
| 158 | + // First normalization (case insensitive for all) |
| 159 | + let first_norm = normalize_identifier(&original, &case_insensitive); |
| 160 | + |
| 161 | + // Second normalization (case sensitive for quoted) |
| 162 | + let second_norm = normalize_identifier(&first_norm, &mixed_sensitivity); |
| 163 | + |
| 164 | + // Third normalization (back to case insensitive) |
| 165 | + let third_norm = normalize_identifier(&second_norm, &case_insensitive); |
| 166 | + |
| 167 | + // Check that the quote information is preserved through all normalizations |
| 168 | + assert_eq!( |
| 169 | + first_norm.quote, |
| 170 | + Some('"'), |
| 171 | + "Quote lost after first normalization" |
| 172 | + ); |
| 173 | + assert_eq!( |
| 174 | + second_norm.quote, |
| 175 | + Some('"'), |
| 176 | + "Quote lost after second normalization" |
| 177 | + ); |
| 178 | + assert_eq!( |
| 179 | + third_norm.quote, |
| 180 | + Some('"'), |
| 181 | + "Quote lost after third normalization" |
| 182 | + ); |
| 183 | + |
| 184 | + // The name should be lowercase due to the case insensitive context |
| 185 | + assert_eq!(third_norm.name, "mixedcaseidentifier"); |
| 186 | + |
| 187 | + Ok(()) |
| 188 | + } |
| 189 | +} |
0 commit comments