Merge pull request #59 from ni-ze/newWorld

[ISSUE #60]Add parser unit tests
alibaba · Jan 16, 2023 · 8f8488e · 8f8488e
2 parents 31d92ba + d99e310
commit 8f8488e
Show file tree

Hide file tree

Showing 42 changed files with 1,960 additions and 249 deletions.
diff --git a/.gitignore b/.gitignore
@@ -16,7 +16,7 @@ target/
 *.iml
 *.ipr
 *.gz
-*/rsqldb.cs
+
 
 ### NetBeans ###
 nbproject/private/
@@ -29,7 +29,7 @@ nbdist/
 
 node_modules/
 _book/
-
+rsqldb-parser/src/main/antlr4/com.alibaba.rsqldb.parser/gen/
+SqlLexer.tokens
 */dependency-reduced-pom.xml
-/rsqldb-disk/log/
-/rsqldb-disk/server/rsqldb.cs
+
diff --git a/rsqldb-parser/src/main/antlr4/com.alibaba.rsqldb.parser/SqlLexer.g4 b/rsqldb-parser/src/main/antlr4/com.alibaba.rsqldb.parser/SqlLexer.g4
@@ -39,6 +39,7 @@ LESS_SYMBOL:                         '<';
 NOT_EQUAL_SYMBOL:                    '<>'       |   '!=';
 GREATER_EQUAL_SYMBOL:                '>=';
 LESS_EQUAL_SYMBOL:                   '<=';
+//support like
 
 // Operators. Bit
 
@@ -100,6 +101,7 @@ AND:                                    'AND'               |       'and';
 IN:                                     'IN'                |       'in';
 INTO:                                   'INTO'              |       'into';
 LEFT:                                   'LEFT'              |       'left';
+INNER:                                  'INNER'             |       'inner';
 ON:                                     'ON'                |       'on';
 YEAR:                                   'YEAR'              |       'year';
 MONTH:                                  'MONTH'             |       'month';
@@ -129,6 +131,7 @@ INTEGER
     : NUM+
     ;
 
+// '123', "1234"
 QUOTED_NUMBER
     : '\'' NUMBER '\''
     | '"' NUMBER '"'

diff --git a/rsqldb-parser/src/main/antlr4/com.alibaba.rsqldb.parser/SqlParser.g4 b/rsqldb-parser/src/main/antlr4/com.alibaba.rsqldb.parser/SqlParser.g4
@@ -38,7 +38,7 @@ tableProperties
     ;
 
 tableProperty
-    : identifier EQUAL_SYMBOL literal
+    : identifier EQUAL_SYMBOL value
     ;
 
 query
@@ -59,7 +59,7 @@ havingPhrase
     : HAVING booleanExpression
     ;
 joinPhrase
-    : (LEFT)? JOIN tableName (AS identifier)? ON joinCondition
+    : (LEFT | INNER)? JOIN tableName (AS identifier)? ON joinCondition
     ;
 
 selectField
@@ -83,21 +83,22 @@ oneJoinCondition
 
 booleanExpression
     : booleanExpression (AND | OR) booleanExpression                    #jointExpression
-    | fieldName operator literal                                        #operatorExpression
+    | fieldName operator value                                        #operatorExpression
     | fieldName IS NULL                                                 #isNullExpression
     | fieldName BETWEEN NUMBER AND NUMBER                               #betweenExpression
     | fieldName IN values                                               #inExpression
-    | function operator literal                                         #functionExpression
+    | function operator value                                         #functionExpression
     ;
 
-literal
-    : NULL                                  #nullLiteral
-    | (TRUE | FALSE)                        #booleanLiteral
-    | STRING                                #stringLiteral
-    | VARIABLE                              #variableLiteral
-    | NUMBER                                #numberLiteral
-    | QUOTED_NUMBER                         #quotedNumberLiteral
-    | QUOTED_STRING                         #quotedStringLiteral
+value
+    : NULL                                  #nullValue
+    | (TRUE | FALSE)                        #booleanValue
+    | STRING                                #stringValue
+    | VARIABLE                              #variableValue
+    | NUMBER                                #numberValue
+    | QUOTED_NUMBER                         #quotedNumberValue
+    | QUOTED_STRING                         #quotedStringValue
+    | BACKQUOTED_STRING                     #backQuotedStringValue
     ;
 
 function
@@ -142,7 +143,7 @@ timeunit
 
 
 values
-    : LR_BRACKET literal (COMMA literal)* RR_BRACKET
+    : LR_BRACKET value (COMMA value)* RR_BRACKET
     ;
 
 columnConstraint

diff --git a/rsqldb-parser/src/main/java/com/alibaba/rsqldb/parser/DefaultParser.java b/rsqldb-parser/src/main/java/com/alibaba/rsqldb/parser/DefaultParser.java
@@ -46,7 +46,10 @@ public List<Statement> parseStatement(String sql) throws SyntaxErrorException {
         CommonTokenStream tokens = new CommonTokenStream(sqlLexer);
         com.alibaba.rsqldb.parser.SqlParser parser = new com.alibaba.rsqldb.parser.SqlParser(tokens);
 
+        sqlLexer.removeErrorListeners();
         sqlLexer.addErrorListener(new DefaultErrorListener());
+
+        parser.removeErrorListeners();
         parser.addErrorListener(new DefaultErrorListener());
 
         com.alibaba.rsqldb.parser.SqlParser.SqlStatementsContext statements = parser.sqlStatements();

diff --git a/rsqldb-parser/src/main/java/com/alibaba/rsqldb/parser/impl/DefaultErrorListener.java b/rsqldb-parser/src/main/java/com/alibaba/rsqldb/parser/impl/DefaultErrorListener.java
@@ -15,16 +15,46 @@
  */
 package com.alibaba.rsqldb.parser.impl;
 
+import com.alibaba.rsqldb.common.exception.SyntaxErrorException;
+import com.alibaba.rsqldb.parser.SqlParser;
+import com.alibaba.rsqldb.parser.util.ParserUtil;
 import org.antlr.v4.runtime.BaseErrorListener;
+import org.antlr.v4.runtime.CharStream;
+import org.antlr.v4.runtime.CommonToken;
 import org.antlr.v4.runtime.RecognitionException;
 import org.antlr.v4.runtime.Recognizer;
+import org.antlr.v4.runtime.RuleContext;
+import org.antlr.v4.runtime.misc.Interval;
+import org.slf4j.Logger;
+import org.slf4j.LoggerFactory;
 
 public class DefaultErrorListener extends BaseErrorListener {
+    private static final Logger logger = LoggerFactory.getLogger(DefaultErrorListener.class);
+    private final KeyWordPredictor predictor = new KeyWordPredictor();
+
     @Override
-    public void syntaxError(Recognizer<?, ?> recognizer, Object offendingSymbol, int line, int charPositionInLine, String msg, RecognitionException e) {
-        System.err.println("line " + line + ":" + charPositionInLine + " " + msg);
+    public void syntaxError(Recognizer<?, ?> recognizer, Object offendingSymbol,
+                            int line, int charPositionInLine, String msg, RecognitionException e) {
+        if (predictor.apply(msg, e)) {
+            return;
+        }
+
+        CharStream inputStream = ((CommonToken) offendingSymbol).getInputStream();
+        String sql = inputStream.getText(Interval.of(0, 1000));
+
+        logger.error("error when parse sql. position=(line:{}, char index:{}), error msg:{}, \n {}", line, charPositionInLine, msg, sql);
         if (e != null) {
-            throw new RuntimeException(e);
+            if (e.getOffendingToken() != null) {
+                String targetIdentifier = e.getOffendingToken().getText();
+                boolean keyWord = ParserUtil.isKeyWord(targetIdentifier);
+                if (keyWord) {
+                    logger.error("{} is a keyWord, change it to `{}`", targetIdentifier, targetIdentifier);
+                }
+            }
+            throw new SyntaxErrorException(msg, e);
+        } else {
+            throw new SyntaxErrorException(msg);
         }
     }
+
 }