clickhouse-orm (中文 README)


Clickhouse ORM for Nodejs. Send query over HTTP interface. Using TimonKK/clickhouse.

Install:

``
npm i clickhouse-orm
`

Usage

Create instance：

`typescript
const { ClickhouseOrm } = require("clickhouse-orm");

const chOrm = ClickhouseOrm({
db: {
name: "orm_test",
},
debug: true,
client: {
url: "localhost",
port: "8123",
basicAuth: {
username: "default",
password: "",
},
debug: false,
isUseGzip: true,
format: "json", // "json" || "csv" || "tsv"
},
});
`

Define Model：
`typescript
import { DATA_TYPE, ModelSyncTableConfig } from 'clickhouse-orm';
const table1Schema: ModelSyncTableConfig<{
status?: number;
time: Date;
browser?: string;
browser_v?: string;
}> = {
tableName: "table1",
schema: {
time: { type: DATA_TYPE.DateTime, default: Date },
status: { type: DATA_TYPE.Int32 },
browser: { type: DATA_TYPE.String },
browser_v: { type: DATA_TYPE.String },
},
options: ENGINE = MergeTree
PARTITION BY toYYYYMM(time)
ORDER BY time,
autoCreate: true,
autoSync: true,
};
`

Create data / Find：

`typescript
// create database 'orm_test'
await chOrm.createDatabase();
// register schema and create [if] table
const Table1Model = await chOrm.model(table1Schema);

// create data
const resCreate = await Table1Model.create({
status: 1,
time: new Date(),
browser: "chrome",
browser_v: "90.0.1.21",
});
console.log("create:", resCreate);

// find
Table1Model.find({
select: "*",
limit: 3,
}).then((res) => {
// SQL: SELECT * from orm_test.table1 LIMIT 3
console.log("find:", res);
});
`

More in Basic Example.

Document

Note: '?' is a Optional

$3

db : object<{name:string, engine?:string, cluster?:string}>

> name: database name

> engine?: database engine

> cluster?: cluster name

debug : boolean

> Default: false

client : object

> Drive configuration. More in TimonKK/clickhouse.

$3

| | required | type | description |
| ------ | ------ | ------ | ------ |
| tableName | true | string | It is the table name. |
| schema | true | { [column]: { type, default? } } | Type defines the data type, and default sets the default value |


`typescript
import { DATA_TYPE, ModelConfig } from 'clickhouse-orm';
const xxxSchema: ModelConfig<{
status?: number;
time: Date;
browser?: string;
browser_v?: string;
}> = {
// table name
tableName: "xxx",
// define column name
schema: {
time: { type: DATA_TYPE.DateTime, default: Date },
status: { type: DATA_TYPE.Int32 },
browser: { type: DATA_TYPE.LowCardinality(DATA_TYPE.String) },
browser_v: { type: DATA_TYPE.String }
},
};
`
----


2. ModelSyncTableConfig (Recommended)

| | required | type | description |
| ------ | ------ | ------ | ------ |
| tableName | true | string | It is the table name. |
| schema | true | { [column]: { type, default? } } | Type defines the data type, and default sets the default value |
| options | true | string | Create table setting |
| autoCreate | true | boolean | Auto create table |
| autoSync | false | boolean | Auto sync table structure(Careful use) |


Automatically create tables and automatically synchronize table field structures

`typescript
import { DATA_TYPE, ModelSyncTableConfig } from 'clickhouse-orm';
const oldSchema: ModelSyncTableConfig<{
time: Date;
will_typeChanged?: number;
will_deleted?: string;
}> = {
tableName: "xxx",
schema: {
time: { type: DATA_TYPE.DateTime, default: Date },
will_typeChanged: { type: DATA_TYPE.Int16 },
will_deleted: { type: DATA_TYPE.String },
},
options: ENGINE = MergeTree
PARTITION BY toYYYYMM(time)
ORDER BY time,
autoCreate: true,
autoSync: true,
};
`
Synchronize the table structure (just the field columns) when creating the model. When the model is created, the field structure of the remote database table will be pulled for comparison with the schema in the current code. Finally, the ORM will automatically execute (add, delete, and modify) statements.


`typescript
const newSchema = {
...oldSchema,
schema: {
time: { type: DATA_TYPE.DateTime, default: Date },
will_typeChanged: { type: DATA_TYPE.Int32 },
add_column: { type: DATA_TYPE.String },
}
}
chOrm.model(newSchema)
`

> clickhouse-orm-log: sync table structure: ALTER TABLE orm_test.xxx DROP COLUMN will_deleted

> clickhouse-orm-log: sync table structure: ALTER TABLE orm_test.xxx ADD COLUMN add_column String

> clickhouse-orm-log: sync table structure: ALTER TABLE orm_test.xxx MODIFY COLUMN will_typeChanged Int32

Unrecognized field name modification! The following configuration will delete column1 and add column2.

`typescript
oldSchema = {
column1: { type: DATA_TYPE.String },
}
newSchema = {
column2: { type: DATA_TYPE.String },
}
`

More in SyncTable Example.

----
3. ModelSqlCreateTableConfig

| | required | type | description |
| ------ | ------ | ------ | ------ |
| tableName | true | string | It is the table name. |
| schema | true | { [column]: { type, default? } } | Type defines the data type, and default sets the default value |
| createTable | true | string | It is the SQL for creating tables.When model is executed, this SQL will be executed. It is suggested to add 'IF NOT EXISTS'.
Watch out !!! >>>>> If the table already exists and you want to modify it. You need to execute the modification sql through other clients（Such as Remote terminal） and update the code of the Schema!!!|


Customized table creation statement, and the table will be created automatically when the model is created

`typescript
import { DATA_TYPE, ModelSqlCreateTableConfig } from 'clickhouse-orm';
const xxxSchema: ModelSqlCreateTableConfig<{
status?: number;
time: Date;
browser?: string;
browser_v?: string;
}> = {
// table name
tableName: "xxx",
// define column name
schema: {
time: { type: DATA_TYPE.DateTime, default: Date },
status: { type: DATA_TYPE.Int32 },
browser: { type: DATA_TYPE.LowCardinality(DATA_TYPE.String) },
browser_v: { type: DATA_TYPE.String }
},
// create table sql
createTable: (dbTableName) => {
// dbTableName = db + '.' + tableName = (orm_test.table1)
return
CREATE TABLE IF NOT EXISTS ${dbTableName}
(
time DateTime,
status Int32,
browser LowCardinality(String),
browser_v String
)
ENGINE = MergeTree
PARTITION BY toYYYYMM(time)
ORDER BY time;
},
};
`

$3

ClickHouse data type. Most of the following data will be validated by ORM, but only the basic data type number | string | boolean | date will be validated, not the most standard data type. For example, in Int8, ORM only verifies the number type.


`typescript
UInt8;
UInt16;
UInt32;
UInt64;
Int8;
Int16;
Int32;
Int64;
Float32;
Float64;
Boolean;
String;
UUID;
Date;
Date32;
DateTime;
DateTime64;
/**
*
* @param Number
* @example DATA_TYPE.FixedString(3)
*/
FixedString;
/**
*
* @param DATA_TYPE
* @example DATA_TYPE.LowCardinality(DATA_TYPE.String)
*/
LowCardinality;
/**
*
* @param string
* @example DATA_TYPE.Enum8('hello' = 1, 'world' = 2)
* @desc number [-128, 127]
*/
Enum8;
/**
*
* @param string
* @example DATA_TYPE.Enum16('hello' = 3000, 'world' = 3500)
* @desc number [-32768, 32767]
*/
Enum16;
/**
*
* @param columnType
* Clickhouse dataTypes: Array(T), JSON, Map(key, value), IPv4, Nullable(), more...
* @example DATA_TYPE.Other('Array(String)') , DATA_TYPE.Other('Int8')
* @desc No INSERT data validation provided
*/
Other;
`
More in Datatype Example.


$3

The setLogService is a global configuration method and will affect all instances.

Default: console.log

Custom example: winston

`javascript
const { setLogService } = require("clickhouse-orm");
const winston = require("winston");
const logger = winston.createLogger();

setLogService(logger.info);
`

$3

`javascript
chOrm.client
.query(select * from orm_test.table1 limit 3)
.toPromise()
.then((res) => {
console.log("Use sql:", res);
});
`

The chOrm.client is the TimonKK/clickhouse instance.

More Examples

$3

`javascript
import * as dayjs from "dayjs";

queryExample1({
Model: Table1Model,
status: 1,
beginTime: dayjs().subtract(1, "day").format("YYYY-MM-DD HH:mm:ss"),
endTime: dayjs().format("YYYY-MM-DD HH:mm:ss"),
}).then((res) => {
console.log("queryExample1:", res);
});

const queryExample1 = ({ Model, status, beginTime, endTime }) => {
let wheres = [],
where;
if (status) wheres.push(status='${status}');
if (beginTime) wheres.push(time>='${beginTime}');
if (endTime) wheres.push(time<='${endTime}');
if (wheres.length > 0) where = wheres.join(" and ");

return Model.find({
where,
select: *,
orderBy: "time ASC",
limit: 5,
});
};
`

Final executed SQL:

`sql
SELECT * from orm_test.table1 where status='1' and time>='2022-02-04 15:34:22' and time<='2022-02-05 15:34:22'  ORDER BY time ASC LIMIT 5
`

$3

`javascript
countExample1({
Model: Table1Model,
}).then((res) => {
console.log("countExample1:", res);
});

const countExample1 = ({ Model }) => {
return Model.find({
select: count(*) AS total,
});
};
`

Final executed SQL:

`sql
SELECT count(*) AS total from orm_test.table1
`

$3

`javascript
Table1Model.find({
select: status,browser,
groupBy: "status,browser",
});
`

Final executed SQL:

`sql
SELECT status,browser from orm_test.table1  GROUP BY status,browser
`

$3

`javascript
Table1Model.find([
{
select: browser,
groupBy: "browser",
},
{
select: count() as browserTotal,
},
]);
`

Final executed SQL:

`sql
SELECT count() as browserTotal from (SELECT browser from orm_test.table1  GROUP BY browser  )
`

$3

`javascript
// new data model
const data = Table1Model.build();

// set value
data.time = new _Date_();
data.status = 1;
data.browser = "chrome";
data.browser_v = "90.0.1.21";

// do save
data.save().then((res) => {
console.log("save:", res);
});
`

Final executed SQL:

`
INSERT INTO orm_test.table1 (time,status,browser,browser_v) [{"time":"2022-02-05T07:51:16.919Z","status":1,"browser":"chrome","browser_v":"90.0.1.21"}]\
`

$3

`javascript
//do create
await Table1Model.create({
status: 1,
time: new Date(),
browser: "chrome",
browser_v: "90.0.1.21",
});
`

Final executed SQL:

`
INSERT INTO orm_test.table1 (time,status,browser,browser_v) [{"time":"2022-02-05T07:51:16.919Z","status":1,"browser":"chrome","browser_v":"90.0.1.21"}]\
`

$3

`javascript
const list = [
{ status: 2, browser: "IE", browser_v: "10.0.1.21" },
{ status: 2, browser: "FF", browser_v: "2.0.3" },
{ status: 3, browser: "IE", browser_v: "1.1.1" },
];

Table1Model.insertMany(list);
// or
Table1Model.insertMany(
list.map((item) => {
const data = Table1Model.build();
// set value
data.time = new Date();
data.status = item.status;
data.browser = item.browser;
data.browser_v = item.browser_v;
return data;
})
);
`

Final executed SQL:

`sql
INSERT INTO orm_test.table1 (time,status,browser,browser_v) [{"time":"2022-02-05T07:34:22.226Z","status":2,"browser":"IE","browser_v":"10.0.1.21"},{"time":"2022-02-05T07:34:22.226Z","status":2,"browser":"FF","browser_v":"2.0.3"},{"time":"2022-02-05T07:34:22.226Z","status":3,"browser":"IE","browser_v":"1.1.1"}]
`

$3

`javascript

Table1Model.delete({
where: browser='Chrome',
})
`

Final executed SQL:

`sql
ALTER TABLE orm_test.table1 DELETE WHERE browser='Chrome'
`

cluster

Create a cluster instance：

`javascript
const { ClickhouseOrm, DATA_TYPE, setLogService } = require("clickhouse-orm");

const chOrm = ClickhouseOrm({
db: {
name: "orm_cluster_test",
cluster: "default_cluster",
},
// ...other
});
const table2Schema = {
// table name
tableName: "table2",
...other,
};

// create database 'orm_cluster_test'
// SQL: CREATE DATABASE IF NOT EXISTS orm_cluster_test ON CLUSTER default_cluster
await chOrm.createDatabase();

// register schema and create [if] table
// createTable() SQL: CREATE TABLE IF NOT EXISTS orm_cluster_test.table2 ON CLUSTER default_cluster...
const Table2Model = await chOrm.model(table2Schema);
``

Wechat Discussion

Click to join

##### or