SqlParserCS 0.3.0

Extensible SQL Lexer and Parser for .NET

This project is .NET port from the original Rust sqlparser-rs project.

It's a .NET library capable of parsing SQL statements into an abstract syntax tree (AST). It contains a lexer and parser for SQL that conforms with the ANSI/ISO SQL standard and other dialects.

Getting Started

Installation Install SqlParser from the .NET CLI as:

dotnet add package SqlParserCS

Or from the package manager console:

PM> Install-Package SqlParserCS

Parsing SQL Queries

Parse a simple SELECT statement:

using SqlParser;

var sql = """
	select top 10 *
        from my_table
        where abc > 99
        order by xyz desc
		""";

var ast =  new Parser().ParseSql(sql);

SqlParser reads the output into a token stream and parses it into a SQL syntax tree.

Demo Project

Need more than just a code snippet? The solution includes a demo console app where you can test out parsing SQL as well as look at few simple examples of how to use the AST to query an arbitrary data source like a CSV file. Download or clone the project, set the demo project as startup, and run it. The console app will prompt you through the examples.

AST Object Model

Once the syntax tree has been parsed, you can query and traverse the object model. The example below is (highly) manual way of parsing the output. The library contains numerous convenience methods to help traverse the tree gracefully.

// Get the query statement since parsed results can have multiple
// statements (separated by a semicolon).
var select = statements.First() as Statement.Select;
// Set a variable to the select statement's body
var body = select.Query.Body as SetExpression.SelectExpression;
// Find the table contained within the select
var table = body.Select.From.First().Relation as TableFactor.Table;
// Write out the table's name
Console.WriteLine(table.Name); // <-- writes "my_table"

C# Records

While C# and Rust programming languages have high feature parity, there are some things that (at present), C# cannot duplicate. In particular, Rust uses discriminated unions. Simply put, Rust allows enums to carry class-like (or record-like) implementations. C# can come close with creative use of record types, particularly due to their ByValue equality. However they still fall short of the Rust counterpart.

In this project, every effort has been made to duplicate the object hierarchy by using a combination of abstract base records with enclosed sub-class implementations. While somewhat unconventional, it allows each major syntax tree type to have a pseudo namespace within the containing record type.

For example, a Statement is a high level type that describes SELECT, INSERT, DELETE statements, etc. Each of those specific implementations is a child of the Statement abstract base record.

var statement = new Statement.Select(...)

<br/>

Text Output

The library supports multiple SQL dialects (Postgres, MsSql, etc.) as well as multiple output modes. The code and can be extended to display the syntax tree in custom formats.

This example calls ToString() and redners the AST's default format showing the object hierarchy

var sql = "select top 10 * from mytable where abc > 99 order by xyz desc";
var ast =  new Parser().ParseSql(sql);

Console.Write!($"{ast}");
// Elided and multi-lines for readability
Select { 
  Query = Query { 
    Body = SelectExpression { 
      Select = Select { 
        Projection = [ Wildcard { }],
          Distinct = False, 
          ...
        }, 
        From = [
          TableWithJoins { 
            Relation = Table { 
              Name = mytable
            }
          }
        ]  
    ...
  }
}

The same query can be output the built-in ToSql() method which produces a SQL string representing the original query after going through a process of normalizing the text (e.g. upper case keywords).

Console.WriteLine($"SQL: {ast.ToSql()}");

SQL: SELECT TOP 10 * FROM mytable WHERE abc > 10 ORDER BY xyz DESC

The syntax tree can also be formatted as JSON, for example. This example uses Newtonsoft to render the AST hierarchy.

JsonConvert.SerializeObject(statements.First(), Formatting.Indented)
// Elided for readability
{
   "Query": {
      "Body": {
         "Select": {
            "Projection": [
               {
                  "Expression": {
                     "Ident": {
                        "Value": "a",
                        "QuoteStyle": null
                     }
                  }
               }
	...

<br/>

Features

SQL Dialects

This library supports the following SQL dialects. You can also create your own dialect with support for parsing SQL at every level of the syntax tree hierarchy.

• Generic SQL dialect
• ANSI SQL dialect
• BigQuery SQL dialect
• ClickHouse SQL dialect
• DuckDb SQL dialect
• Hive SQL dialect
• MS SQL dialect
• My SQL dialect
• PostgreSQL dialect
• Redshift SQL dialect
• SQLite dialect

Supporting custom SQL dialects

The library supports a number of inbuilt dialects, but it is not comprehensive. As such, you can create your own dialect with hooks into the parser that allow you to customize the AST as it's being constructed.

Please see the included examples for reference.

Visitors

The library also contains a base class visitor implementation that allows you to walk the entire syntax tree. The visitor only traverses non-null elements in the tree and resports all Statements, Expressions, and Relations before and after each element is visited. The base class is a jumping-off point for traversing the AST, and is only one of several ways the AST can be navigated.

var visitor = new YourCustomVisitor();
ast.Visit(visitor);

Syntax vs Semantics

Like the Rust counterpart, this library provides only a syntax parser, and tries to avoid applying any SQL semantics, and accepts queries that specific databases would reject, even when using that Database's specific Dialect.

For example, CREATE TABLE(x int, x int) is accepted by this crate, even though most SQL engines will reject this statement due to the repeated column name x.

This library avoids semantic analysis because it varies drastically between dialects and implementations. If you want to do semantic analysis, feel free to use this project as a base.

Design

The core expression parser uses the [Pratt Parser] design, which is a top-down operator-precedence (TDOP) parser, while the surrounding SQL statement parser is a traditional, hand-written recursive descent parser. Eli Bendersky has a good tutorial on TDOP parsers, if you are interested in learning more about the technique.

Like the Rust project author's, we're are a fan of this design pattern over parser generators for the following due to simplicity, performancee, ease of debugging, and extensibility.

If you like Lexing & Parsing, I have a similar example of a hand-written CSS lexer/parser.

Contributions are welcome!

Contributions are highly encouraged! Please keep in mind that this project is a port of the Rust SQL parser, and every effort will be made to keep the projects functionally in sync. Therefore pull requests will generally follow the spirit of the oriignal project. Ideally, pull requests that affect the parser or a dialect specifically will be submitted back to the original Rust project as well.

Pull requests that add support for or fix a bug in a feature in the SQL standard, or a feature in a popular RDBMS, like Microsoft SQL Server or PostgreSQL, will likely be accepted after a brief review.

PRs without tests will not be reviewed or merged. Please run all tests prior to any pull request

If you are unable to submit a patch, feel free to file an issue instead. Please try to include:

• some representative examples of the syntax you wish to support or fix;
• the relevant bits of the SQL grammar, if the syntax is part of SQL:2016; and
• links to documentation for the feature for a few of the most popular databases that support it.

If you need support for a feature, you will likely need to implement it yourself. Our goal as maintainers is to facilitate the integration of various features from various contributors, but not to provide the implementations ourselves, as we simply don't have the resources.

SQL compliance

This project adops the same compliance position as the Rust project:

SQL was first standardized in 1987, and revisions of the standard have been published regularly since. Most revisions have added significant new features to the language, and as a result no database claims to support the full breadth of features. This parser currently supports most of the SQL-92 syntax, plus some syntax from newer versions that have been explicitly requested, plus some MSSQL, PostgreSQL, and other dialect-specific syntax. Whenever possible, the online SQL:2016 grammar is used to guide what syntax to accept.

Unfortunately, stating anything more specific about compliance is difficult. There is no publicly available test suite that can assess compliance automatically, and doing so manually would strain the project's limited resources. Still, we are interested in eventually supporting the full SQL dialect, and we are slowly building out our own test suite.

If you are assessing whether this project will be suitable for your needs, you'll likely need to experimentally verify whether it supports the subset of SQL that you need. Please file issues about any unsupported queries that you discover. Doing so helps us prioritize support for the portions of the standard that are actually used. Note that if you urgently need support for a feature, you will likely need to write the implementation yourself. See the Contributing section for details.

License, etc.

SqlParser is based on the work originally developed by Andy Grove and other contributors

SqlParser is Copyright © 2023 Tyler Brinks and other contributors under the MIT license.