Basic Query Operations

本文介绍了LINQ的基础查询语法,包括筛选、排序、联接及分组等操作,并提供了丰富的示例代码,帮助读者掌握LINQ的基本用法。

摘要生成于 C知道 ,由 DeepSeek-R1 满血版支持, 前往体验 >

LINQ 的查询语法和 SQL 非常相似,只不过为了智能提示和类型推断将 select 子句放在最后。

int[] nums = { 1, 2, 3, 4 };
var q = from n in nums select n;


"from n in nums" 中的两个变量用法类似 "foreach (var n in nums)","select n" 用于返回当前变量。

1. Filtering

所谓 Filtering,就是 where 子句,用于进行条件过滤。

Query Syntax

int[] nums = { 1, 2, 3, 4 };
var q = from n in nums where n % 2 == 0 select n;

foreach (var v in q)
  Console.WriteLine(v);


Method Syntax

var q = nums.Where(n => n % 2 == 0);


输出:
2
4

2. Ordering

使用 "orderby ... ascending" 或 "orderby ... descending" 子句可以实现对结果的排序操作。

Query Syntax

int[] nums = { 1, 2, 3, 4 };
var q = from n in nums orderby n descending select n;

foreach (var v in q)
  Console.WriteLine(v);


Method Syntax

var q = nums.OrderByDescending(n => n);


输出:
4
3
2
1

3. Joining

下面这个例子稍微有点复杂,我们的意图是将 User.Sex 从整数转换成对应的性别字符。 "join...in..." 的作用是将额外的数据源合并进来,而 "on...equals..." 则用于说明合并的条件,最后我们使用匿名类型来创建所需的新数据容器。

Query Syntax

var sex = new[] {
       new { Id = 0, Name = "女"},
       new { Id = 1, Name = "男"}
     };

var users = new[] {
 new { Name = "User1", Sex = 1 },
 new { Name = "User2", Sex = 0 },
 new { Name = "User3", Sex = 1 },
 new { Name = "User4", Sex = 0 },
 new { Name = "User5", Sex = 1 },
 new { Name = "User6", Sex = 0 },
 new { Name = "User7", Sex = 1 },
 new { Name = "User8", Sex = 0 },
 new { Name = "User9", Sex = 1 }
 };

var q = from u in users
    join s in sex
    on u.Sex equals s.Id
    select new { Name = u.Name, Sex = s.Name };

foreach (var u in q)
  Console.WriteLine("Name:{0} Sex:{1}", u.Name, u.Sex);


Method Syntax

var q = users.Join(sex, u => u.Sex, s => s.Id, (u, s) => new { Name = u.Name, Sex = s.Name });


输出:
Name:User1 Sex:男
Name:User2 Sex:女
Name:User3 Sex:男
Name:User4 Sex:女
Name:User5 Sex:男
Name:User6 Sex:女
Name:User7 Sex:男
Name:User8 Sex:女
Name:User9 Sex:男

4. Grouping

使用 "group...by..." 可以按照一定的条件进行分组返回,该语句可以省略 select 子句。

Query Syntax

var users = new[] {
 new { Name = "User1", Sex = 1 },
 new { Name = "User2", Sex = 0 },
 new { Name = "User3", Sex = 1 },
 new { Name = "User4", Sex = 0 },
 new { Name = "User5", Sex = 1 },
 new { Name = "User6", Sex = 0 },
 new { Name = "User7", Sex = 1 },
 new { Name = "User8", Sex = 0 },
 new { Name = "User9", Sex = 1 }
};

var q = from u in users group u by u.Sex;

foreach (var group in q)
{
  Console.WriteLine("---------------");

  foreach (var u in group)
    Console.WriteLine(u.Name);
}


Method Syntax

var q = users.GroupBy(u => u.Sex);


输出:
---------------
User1
User3
User5
User7
User9
---------------
User2
User4
User6
User8

""" title: SQL Server Access author: MENG author_urls: - https://github.com/mengvision description: A tool for reading database information and executing SQL queries, supporting multiple databases such as MySQL, PostgreSQL, SQLite, and Oracle. It provides functionalities for listing all tables, describing table schemas, and returning query results in CSV format. A versatile DB Agent for seamless database interactions. required_open_webui_version: 0.5.4 requirements: pymysql, sqlalchemy, cx_Oracle version: 0.1.6 licence: MIT # Changelog ## [0.1.6] - 2025-03-11 ### Added - Added `get_table_indexes` method to retrieve index information for a specific table, supporting MySQL, PostgreSQL, SQLite, and Oracle. - Enhanced metadata capabilities by providing detailed index descriptions (e.g., index name, columns, and type). - Improved documentation to include the new `get_table_indexes` method and its usage examples. - Updated error handling in `get_table_indexes` to provide more detailed feedback for unsupported database types. ## [0.1.5] - 2025-01-20 ### Changed - Updated `list_all_tables` and `table_data_schema` methods to accept `db_name` as a function parameter instead of using `self.valves.db_name`. - Improved flexibility by decoupling database name from class variables, allowing dynamic database selection at runtime. ## [0.1.4] - 2025-01-17 ### Added - Added support for Oracle database using `cx_Oracle` driver. - Added dynamic engine creation in each method to ensure fresh database connections for every operation. - Added support for Oracle-specific queries in `list_all_tables` and `table_data_schema` methods. ### Changed - Moved `self._get_engine()` from `__init__` to individual methods for better flexibility and tool compatibility. - Updated `_get_engine` method to support Oracle database connection URL. - Improved `table_data_schema` method to handle Oracle-specific column metadata. ### Fixed - Fixed potential connection issues by ensuring each method creates its own database engine. - Improved error handling for Oracle-specific queries and edge cases. ## [0.1.3] - 2025-01-17 ### Added - Added support for multiple database types (e.g., MySQL, PostgreSQL, SQLite) using SQLAlchemy. - Added configuration flexibility through environment variables or external configuration files. - Enhanced query security with stricter validation and SQL injection prevention. - Improved error handling with detailed exception messages for better debugging. ### Changed - Replaced `pymysql` with SQLAlchemy for broader database compatibility. - Abstracted database connection logic into a reusable `_get_engine` method. - Updated `table_data_schema` method to support multiple database types. ### Fixed - Fixed potential SQL injection vulnerabilities in query execution. - Improved handling of edge cases in query validation and execution. ## [0.1.2] - 2025-01-16 ### Added - Added support for specifying the database port with a default value of `3306`. - Abstracted database connection logic into a reusable `_get_connection` method. ## [0.1.1] - 2025-01-16 ### Added - Support for additional read-only query types: `SHOW`, `DESCRIBE`, `EXPLAIN`, and `USE`. - Enhanced query validation to block sensitive keywords (e.g., `INSERT`, `UPDATE`, `DELETE`, `CREATE`, `DROP`, `ALTER`). ### Fixed - Improved handling of queries starting with `WITH` (CTE queries). - Fixed case sensitivity issues in query validation. ## [0.1.0] - 2025-01-09 ### Initial Release - Basic functionality for listing tables, describing table schemas, and executing `SELECT` queries. - Query results returned in CSV format. """ import os from typing import List, Dict, Any from pydantic import BaseModel, Field import re from sqlalchemy import create_engine, text from sqlalchemy.engine.base import Engine from sqlalchemy.exc import SQLAlchemyError class Tools: class Valves(BaseModel): db_host: str = Field( default="localhost", description="The host of the database. Replace with your own host.", ) db_user: str = Field( default="admin", description="The username for the database. Replace with your own username.", ) db_password: str = Field( default="admin", description="The password for the database. Replace with your own password.", ) db_name: str = Field( default="db", description="The name of the database. Replace with your own database name.", ) db_port: int = Field( default=3306, # Oracle 默认端口 description="The port of the database. Replace with your own port.", ) db_type: str = Field( default="mysql", description="The type of the database (e.g., mysql, postgresql, sqlite, oracle).", ) def __init__(self): """ Initialize the Tools class with the credentials for the database. """ print("Initializing database tool class") self.citation = True self.valves = Tools.Valves() def _get_engine(self) -> Engine: """ Create and return a database engine using the current configuration. """ if self.valves.db_type == "mysql": db_url = f"mysql+pymysql://{self.valves.db_user}:{self.valves.db_password}@{self.valves.db_host}:{self.valves.db_port}/{self.valves.db_name}" elif self.valves.db_type == "postgresql": db_url = f"postgresql://{self.valves.db_user}:{self.valves.db_password}@{self.valves.db_host}:{self.valves.db_port}/{self.valves.db_name}" elif self.valves.db_type == "sqlite": db_url = f"sqlite:///{self.valves.db_name}" elif self.valves.db_type == "oracle": db_url = f"oracle+cx_oracle://{self.valves.db_user}:{self.valves.db_password}@{self.valves.db_host}:{self.valves.db_port}/?service_name={self.valves.db_name}" else: raise ValueError(f"Unsupported database type: {self.valves.db_type}") return create_engine(db_url) def list_all_tables(self, db_name: str) -> str: """ List all tables in the database. :param db_name: The name of the database. :return: A string containing the names of all tables. """ print("Listing all tables in the database") engine = self._get_engine() # 动态创建引擎 try: with engine.connect() as conn: if self.valves.db_type == "mysql": result = conn.execute(text("SHOW TABLES;")) elif self.valves.db_type == "postgresql": result = conn.execute( text( "SELECT table_name FROM information_schema.tables WHERE table_schema = 'public';" ) ) elif self.valves.db_type == "sqlite": result = conn.execute( text("SELECT name FROM sqlite_master WHERE type='table';") ) elif self.valves.db_type == "oracle": result = conn.execute(text("SELECT table_name FROM user_tables;")) else: return "Unsupported database type." tables = [row[0] for row in result.fetchall()] if tables: return ( "Here is a list of all the tables in the database:\n\n" + "\n".join(tables) ) else: return "No tables found." except SQLAlchemyError as e: return f"Error listing tables: {str(e)}" def get_table_indexes(self, db_name: str, table_name: str) -> str: """ Get the indexes of a specific table in the database. :param db_name: The name of the database. :param table_name: The name of the table. :return: A string describing the indexes of the table. """ print(f"Getting indexes for table: {table_name}") engine = self._get_engine() try: key, cloumn = 0, 1 with engine.connect() as conn: if self.valves.db_type == "mysql": query = text(f"SHOW INDEX FROM {table_name}") key, cloumn = 2, 4 elif self.valves.db_type == "postgresql": query = text( """ SELECT indexname, indexdef FROM pg_indexes WHERE tablename = :table_name; """ ) elif self.valves.db_type == "sqlite": query = text( """ PRAGMA index_list(:table_name); """ ) elif self.valves.db_type == "oracle": query = text( """ SELECT index_name, column_name FROM user_ind_columns WHERE table_name = :table_name; """ ) else: return "Unsupported database type." result = conn.execute(query) indexes = result.fetchall() if not indexes: return f"No indexes found for table: {table_name}" description = f"Indexes for table '{table_name}':\n" for index in indexes: description += f"- {index[key]}: {index[cloumn]}\n" return description # result = conn.execute(query) # description = result.fetchall() # if not description: # return f"No indexes found for table: {table_name}" # column_names = result.keys() # description = f"Query executed successfully. Below is the actual result of the query {query} running against the database in CSV format:\n\n" # description += ",".join(column_names) + "\n" # for row in description: # description += ",".join(map(str, row)) + "\n" # return description except SQLAlchemyError as e: return f"Error getting indexes: {str(e)}" def table_data_schema(self, db_name: str, table_name: str) -> str: """ Describe the schema of a specific table in the database, including column comments. :param db_name: The name of the database. :param table_name: The name of the table to describe. :return: A string describing the data schema of the table. """ print(f"Database: {self.valves.db_name}") print(f"Describing table: {table_name}") engine = self._get_engine() # 动态创建引擎 try: with engine.connect() as conn: if self.valves.db_type == "mysql": query = text( " SELECT COLUMN_NAME, COLUMN_TYPE, IS_NULLABLE, COLUMN_KEY, COLUMN_COMMENT " " FROM INFORMATION_SCHEMA.COLUMNS " f" WHERE TABLE_SCHEMA = '{self.valves.db_name}' AND TABLE_NAME = '{table_name}';" ) elif self.valves.db_type == "postgresql": query = text( """ SELECT column_name, data_type, is_nullable, column_default, '' FROM information_schema.columns WHERE table_name = :table_name; """ ) elif self.valves.db_type == "sqlite": query = text("PRAGMA table_info(:table_name);") elif self.valves.db_type == "oracle": query = text( """ SELECT column_name, data_type, nullable, data_default, comments FROM user_tab_columns LEFT JOIN user_col_comments ON user_tab_columns.table_name = user_col_comments.table_name AND user_tab_columns.column_name = user_col_comments.column_name WHERE user_tab_columns.table_name = :table_name; """ ) else: return "Unsupported database type." # result = conn.execute( # query, {"db_name": db_name, "table_name": table_name} # ) result = conn.execute(query) columns = result.fetchall() if not columns: return f"No such table: {table_name}" description = ( f"Table '{table_name}' in the database has the following columns:\n" ) for column in columns: if self.valves.db_type == "sqlite": column_name, data_type, is_nullable, _, _, _ = column column_comment = "" elif self.valves.db_type == "oracle": ( column_name, data_type, is_nullable, data_default, column_comment, ) = column else: ( column_name, data_type, is_nullable, column_key, column_comment, ) = column description += f"- {column_name} ({data_type})" if is_nullable == "YES" or is_nullable == "Y": description += " [Nullable]" if column_key == "PRI": description += " [Primary Key]" if column_comment: description += f" [Comment: {column_comment}]" description += "\n" return description except SQLAlchemyError as e: return f"Error describing table: {str(e)}" def execute_read_query(self, query: str) -> str: """ Execute a read query and return the result in CSV format. :param query: The SQL query to execute. :return: A string containing the result of the query in CSV format. """ print(f"Executing query: {query}") normalized_query = query.strip().lower() if not re.match( r"^\s*(select|with|show|describe|desc|explain|use)\s", normalized_query ): return "Error: Only read-only queries (SELECT, WITH, SHOW, DESCRIBE, EXPLAIN, USE) are allowed. CREATE, DELETE, INSERT, UPDATE, DROP, and ALTER operations are not permitted." sensitive_keywords = [ "insert", "update", "delete", "create", "drop", "alter", "truncate", "grant", "revoke", "replace", ] for keyword in sensitive_keywords: if re.search(rf"\b{keyword}\b", normalized_query): return f"Error: Query contains a sensitive keyword '{keyword}'. Only read operations are allowed." engine = self._get_engine() # 动态创建引擎 try: with engine.connect() as conn: result = conn.execute(text(query)) rows = result.fetchall() if not rows: return "No data returned from query." column_names = result.keys() csv_data = f"Query executed successfully. Below is the actual result of the query {query} running against the database in CSV format:\n\n" csv_data += ",".join(column_names) + "\n" for row in rows: csv_data += ",".join(map(str, row)) + "\n" return csv_data except SQLAlchemyError as e: return f"Error executing query: {str(e)}" 将上面的工具连接到Microsoft sql server
最新发布
07-23
评论
添加红包

请填写红包祝福语或标题

红包个数最小为10个

红包金额最低5元

当前余额3.43前往充值 >
需支付:10.00
成就一亿技术人!
领取后你会自动成为博主和红包主的粉丝 规则
hope_wisdom
发出的红包
实付
使用余额支付
点击重新获取
扫码支付
钱包余额 0

抵扣说明:

1.余额是钱包充值的虚拟货币,按照1:1的比例进行支付金额的抵扣。
2.余额无法直接购买下载,可以购买VIP、付费专栏及课程。

余额充值