lava-oushudb-dt-sql-parser/src/parser/trino/index.ts
Hayden a99721162b
feat: collect entity (#265)
* feat: add text and word utils

* feat: add entity collector class

* refactor: rename SyntaxContextType to EntityContextType

* refactor: improve EntityCollector

* feat: improve mysql parser grammar

* feat: add mysql entity collector

* test: mysql entity collector tests

* feat: remove useless method

* feat: improve spark grammar file

* feat: add spark entity collector

* test: spark entity collector unit tests

* feat: remove useless code

* feat: add queryStatement label

* feat: add crateDatabaseStmt

* feat: add trino entity collector

* feat: rename trinosql to trino

* test: trino collect entity unit tests

* test: fix spark test

* feat(impala): support impale entity collector (#256)

* Feat/collect entity hive (#263)

* feat(hive): support hive collect entity

* feat(hive): update tableAllColumns

* feat: replace antlr4ts with antlr4ng

* feat(pgsql): pgsql collect entity (#268)

* feat(pgsql): pgsql collect entity

* feat(pgsql): optimize some name

---------

Co-authored-by: zhaoge <>

* feat: get word text by token.text

* feat: supprt collect db/function and add splitListener (#270)

* feat: supprt collect db/function and add splitListner

* feat: remove SplitListener interface in baseParser to use SplitListener in root

* fix(mysql): fix show create xxx not celloct as createXXXEntity type

* test: fix pgsql unit tests

* Feat/error recover predicate (#274)

* feat: optimize pgsql grammar

* feat: add sql parser base

* feat: apply SQLParserBase

* feat: add geAllEntities method

* test: test collect table when missing column

* feat: compose collect and suggestion (#276)

* feat: mark stmt which contain caret

* test: correct name of getAllEntities

* test: remove misscolumn unit tests

* test: add suggestionWithEntity tests

* feat: flink collect entity (#277)

* feat: improve flink sql parser

* feat: support flink entity collector

* test: flink entity collect unit test

* feat: move combine entities to parent class

---------

Co-authored-by: 霜序 <976060700@qq.com>
Co-authored-by: XCynthia <942884029@qq.com>
2024-03-26 14:28:27 +08:00

134 lines
5.1 KiB
TypeScript

import { Token } from 'antlr4ng';
import { CandidatesCollection } from 'antlr4-c3';
import { TrinoSqlLexer } from '../../lib/trinosql/TrinoSqlLexer';
import { TrinoSqlParser, ProgramContext } from '../../lib/trinosql/TrinoSqlParser';
import BasicParser from '../common/basicParser';
import { Suggestions, EntityContextType, SyntaxSuggestion } from '../common/basic-parser-types';
import { StmtContextType } from '../common/entityCollector';
import TrinoSqlSplitListener from './trinoSplitListener';
import TrinoEntityCollector from './trinoEntityCollector';
export { TrinoSqlSplitListener, TrinoEntityCollector };
export default class TrinoSQL extends BasicParser<TrinoSqlLexer, ProgramContext, TrinoSqlParser> {
protected createLexerFromCharStream(charStreams) {
const lexer = new TrinoSqlLexer(charStreams);
return lexer;
}
protected createParserFromTokenStream(tokenStream) {
const parser = new TrinoSqlParser(tokenStream);
return parser;
}
protected get splitListener() {
return new TrinoSqlSplitListener();
}
protected createEntityCollector(input: string, caretTokenIndex?: number) {
return new TrinoEntityCollector(input, caretTokenIndex);
}
protected preferredRules: Set<number> = new Set([
TrinoSqlParser.RULE_catalogName,
TrinoSqlParser.RULE_catalogNameCreate,
TrinoSqlParser.RULE_schemaName,
TrinoSqlParser.RULE_schemaNameCreate,
TrinoSqlParser.RULE_tableName,
TrinoSqlParser.RULE_tableNameCreate,
TrinoSqlParser.RULE_viewName,
TrinoSqlParser.RULE_viewNameCreate,
TrinoSqlParser.RULE_functionName,
TrinoSqlParser.RULE_columnName,
TrinoSqlParser.RULE_columnNameCreate,
]);
protected processCandidates(
candidates: CandidatesCollection,
allTokens: Token[],
caretTokenIndex: number,
tokenIndexOffset: number
): Suggestions<Token> {
const originalSyntaxSuggestions: SyntaxSuggestion<Token>[] = [];
const keywords: string[] = [];
for (let candidate of candidates.rules) {
const [ruleType, candidateRule] = candidate;
const startTokenIndex = candidateRule.startTokenIndex + tokenIndexOffset;
const tokenRanges = allTokens.slice(
startTokenIndex,
caretTokenIndex + tokenIndexOffset + 1
);
let syntaxContextType: EntityContextType | StmtContextType;
switch (ruleType) {
case TrinoSqlParser.RULE_catalogName: {
syntaxContextType = EntityContextType.CATALOG;
break;
}
case TrinoSqlParser.RULE_schemaName: {
syntaxContextType = EntityContextType.DATABASE;
break;
}
case TrinoSqlParser.RULE_schemaNameCreate: {
syntaxContextType = EntityContextType.DATABASE_CREATE;
break;
}
case TrinoSqlParser.RULE_tableName: {
syntaxContextType = EntityContextType.TABLE;
break;
}
case TrinoSqlParser.RULE_tableNameCreate: {
syntaxContextType = EntityContextType.TABLE_CREATE;
break;
}
case TrinoSqlParser.RULE_viewName: {
syntaxContextType = EntityContextType.VIEW;
break;
}
case TrinoSqlParser.RULE_viewNameCreate: {
syntaxContextType = EntityContextType.VIEW_CREATE;
break;
}
case TrinoSqlParser.RULE_functionName: {
syntaxContextType = EntityContextType.FUNCTION;
break;
}
case TrinoSqlParser.RULE_columnNameCreate: {
syntaxContextType = EntityContextType.COLUMN_CREATE;
break;
}
case TrinoSqlParser.RULE_columnName: {
syntaxContextType = EntityContextType.COLUMN;
break;
}
default:
break;
}
if (syntaxContextType) {
originalSyntaxSuggestions.push({
syntaxContextType,
wordRanges: tokenRanges,
});
}
}
for (let candidate of candidates.tokens) {
const symbolicName = this._parser.vocabulary.getSymbolicName(candidate[0]);
const displayName = this._parser.vocabulary.getDisplayName(candidate[0]);
if (symbolicName && symbolicName.startsWith('KW_')) {
const keyword =
displayName.startsWith("'") && displayName.endsWith("'")
? displayName.slice(1, -1)
: displayName;
keywords.push(keyword);
}
}
return {
syntax: originalSyntaxSuggestions,
keywords,
};
}
}