Include dialect name and help text to lexer error message

sql-formatter-org · Nov 15, 2023 · c3efb63 · c3efb63
1 parent 49f611a
commit c3efb63
Show file tree

Hide file tree

Showing 6 changed files with 66 additions and 31 deletions.
diff --git a/src/dialect.ts b/src/dialect.ts
@@ -33,7 +33,7 @@ export const createDialect = (options: DialectOptions): Dialect => {
 };
 
 const dialectFromOptions = (dialectOptions: DialectOptions): Dialect => ({
-  tokenizer: new Tokenizer(dialectOptions.tokenizerOptions),
+  tokenizer: new Tokenizer(dialectOptions.tokenizerOptions, dialectOptions.name),
   formatOptions: processDialectFormatOptions(dialectOptions.formatOptions),
 });
 

diff --git a/src/lexer/Tokenizer.ts b/src/lexer/Tokenizer.ts
@@ -12,7 +12,7 @@ export default class Tokenizer {
   private rulesBeforeParams: TokenRule[];
   private rulesAfterParams: TokenRule[];
 
-  constructor(private cfg: TokenizerOptions) {
+  constructor(private cfg: TokenizerOptions, private dialectName: string) {
     this.rulesBeforeParams = this.buildRulesBeforeParams(cfg);
     this.rulesAfterParams = this.buildRulesAfterParams(cfg);
   }
@@ -23,7 +23,7 @@ export default class Tokenizer {
       ...this.buildParamRules(this.cfg, paramTypesOverrides),
       ...this.rulesAfterParams,
     ];
-    const tokens = new TokenizerEngine(rules).tokenize(input);
+    const tokens = new TokenizerEngine(rules, this.dialectName).tokenize(input);
     return this.cfg.postProcess ? this.cfg.postProcess(tokens) : tokens;
   }
 

diff --git a/src/lexer/TokenizerEngine.ts b/src/lexer/TokenizerEngine.ts
@@ -22,7 +22,7 @@ export default class TokenizerEngine {
   private input = ''; // The input SQL string to process
   private index = 0; // Current position in string
 
-  constructor(private rules: TokenRule[]) {}
+  constructor(private rules: TokenRule[], private dialectName: string) {}
 
   /**
    * Takes a SQL string and breaks it into tokens.
@@ -58,7 +58,20 @@ export default class TokenizerEngine {
   private createParseError(): Error {
     const text = this.input.slice(this.index, this.index + 10);
     const { line, col } = lineColFromIndex(this.input, this.index);
-    return new Error(`Parse error: Unexpected "${text}" at line ${line} column ${col}`);
+    return new Error(
+      `Parse error: Unexpected "${text}" at line ${line} column ${col}.\n${this.dialectInfo()}`
+    );
+  }
+
+  private dialectInfo(): string {
+    if (this.dialectName === 'sql') {
+      return (
+        `This likely happens because you're using the default "sql" dialect.\n` +
+        `If possible, please select a more specific dialect (like sqlite, postgresql, etc).`
+      );
+    } else {
+      return `SQL dialect used: "${this.dialectName}".`;
+    }
   }
 
   private getWhitespace(): string | undefined {

diff --git a/test/sqlFormatter.test.ts b/test/sqlFormatter.test.ts
@@ -9,10 +9,26 @@ describe('sqlFormatter', () => {
     }).toThrow('Unsupported SQL dialect: blah');
   });
 
-  it('throws error when encountering unsupported characters', () => {
-    expect(() => {
-      format('SELECT «weird-stuff»');
-    }).toThrow('Parse error: Unexpected "«weird-stu" at line 1 column 8');
+  describe('when encountering unsupported characters with default dialect', () => {
+    it('throws error suggesting a use of a more specific dialect', () => {
+      expect(() => {
+        format('SELECT «weird-stuff»');
+      }).toThrow(
+        `Parse error: Unexpected "«weird-stu" at line 1 column 8.\n` +
+          `This likely happens because you're using the default "sql" dialect.\n` +
+          `If possible, please select a more specific dialect (like sqlite, postgresql, etc).`
+      );
+    });
+  });
+
+  describe('when encountering unsupported characters with sqlite dialect', () => {
+    it('throws error including the name of the used dialect', () => {
+      expect(() => {
+        format('SELECT «weird-stuff»', { language: 'sqlite' });
+      }).toThrow(
+        `Parse error: Unexpected "«weird-stu" at line 1 column 8.\nSQL dialect used: "sqlite".`
+      );
+    });
   });
 
   it('throws error when encountering incorrect SQL grammar', () => {

diff --git a/test/unit/Parser.test.ts b/test/unit/Parser.test.ts
@@ -3,18 +3,21 @@ import { createParser } from '../../src/parser/createParser.js';
 
 describe('Parser', () => {
   const parse = (sql: string) => {
-    const tokenizer = new Tokenizer({
-      reservedClauses: ['FROM', 'WHERE', 'LIMIT', 'CREATE TABLE'],
-      reservedSelect: ['SELECT'],
-      reservedSetOperations: ['UNION', 'UNION ALL'],
-      reservedJoins: ['JOIN'],
-      reservedFunctionNames: ['SQRT', 'CURRENT_TIME'],
-      reservedKeywords: ['BETWEEN', 'LIKE', 'ON', 'USING'],
-      operators: [':'],
-      extraParens: ['[]', '{}'],
-      stringTypes: ["''-qq"],
-      identTypes: ['""-qq'],
-    });
+    const tokenizer = new Tokenizer(
+      {
+        reservedClauses: ['FROM', 'WHERE', 'LIMIT', 'CREATE TABLE'],
+        reservedSelect: ['SELECT'],
+        reservedSetOperations: ['UNION', 'UNION ALL'],
+        reservedJoins: ['JOIN'],
+        reservedFunctionNames: ['SQRT', 'CURRENT_TIME'],
+        reservedKeywords: ['BETWEEN', 'LIKE', 'ON', 'USING'],
+        operators: [':'],
+        extraParens: ['[]', '{}'],
+        stringTypes: ["''-qq"],
+        identTypes: ['""-qq'],
+      },
+      'sql'
+    );
 
     return createParser(tokenizer).parse(sql, {});
   };

diff --git a/test/unit/Tokenizer.test.ts b/test/unit/Tokenizer.test.ts
@@ -2,16 +2,19 @@ import Tokenizer from '../../src/lexer/Tokenizer.js';
 
 describe('Tokenizer', () => {
   const tokenize = (sql: string) =>
-    new Tokenizer({
-      reservedClauses: ['FROM', 'WHERE', 'LIMIT', 'CREATE TABLE'],
-      reservedSelect: ['SELECT'],
-      reservedSetOperations: ['UNION', 'UNION ALL'],
-      reservedJoins: ['JOIN'],
-      reservedFunctionNames: ['SQRT', 'CURRENT_TIME'],
-      reservedKeywords: ['BETWEEN', 'LIKE', 'ON', 'USING'],
-      stringTypes: ["''-qq"],
-      identTypes: ['""-qq'],
-    }).tokenize(sql, {});
+    new Tokenizer(
+      {
+        reservedClauses: ['FROM', 'WHERE', 'LIMIT', 'CREATE TABLE'],
+        reservedSelect: ['SELECT'],
+        reservedSetOperations: ['UNION', 'UNION ALL'],
+        reservedJoins: ['JOIN'],
+        reservedFunctionNames: ['SQRT', 'CURRENT_TIME'],
+        reservedKeywords: ['BETWEEN', 'LIKE', 'ON', 'USING'],
+        stringTypes: ["''-qq"],
+        identTypes: ['""-qq'],
+      },
+      'sql'
+    ).tokenize(sql, {});
 
   it('tokenizes whitespace to empty array', () => {
     expect(tokenize(' \t\n \n\r ')).toEqual([]);