linkfuture-pg-api

PG-API

RESTful API for PostgreSQL
An easier way to query database

Table of Contents

• Installation
• Usage
• Query
• Query Composite
• Query Enum
• Configuration
• Events
• KeyWords
• Notice

Installation

The easiest way to install linkfuture-pg-api is with npm.

npm install linkfuture-pg-api

Usage

const $config = {
    "connection":"postgres://<user>:<password>@<host>:<port>/<dbname>",
    "tables":{
        "user":{
            delete:false,//disable delete operation on user table, other operation will be available as default. 
         }
    }
}
const $pgConnector = require("linkfuture-pg-api");
// access through Restful API
const $pgApi = $pgConnector.api($config);
app.use("/api/db/",$pgApi);  
// access through NodeJs
const $pgQuery = $pgConnector.query($config);
let result = await $pgQuery.select("user",{$where:{user_id:1}}); 

Query

SELECT (GET)

• Select by Primary Key

    GET http://[host]/api/db/[table-name or view-name]/[id]
    GET http://[host]/api/db/user/1

• Select by JSON Query

    GET http://[host]/api/db/[table-name or view-name]?$q=[JSON QUERY]
    GET http://[host]/api/db/user?$q={"$where":{"id":{"$any":[1,2,3]}}}

• Select by Query String

    GET http://[host]/api/db/[table-name or view-name]?[ColumnName]=[ColumnValue]&$limit=10
    GET http://[host]/api/db/user?age={"$gt":5,"$lt":50}&is_active=1&$limit=1

• Select in Node

    const $pgQuery = $pgConnector.query($config);
    let result = await $pgQuery.select("user",{$where:{user_id:1}});
    let result = await $pgQuery.selectOne("user",{$where:{user_id:1}});

• JSON Query Example
  Normal Query

{
  "*":true
  ,"unknown_field2": {"$multiply":["age","price","price"]}
  ,"unknown_field3": {"$multiply":["age",{"$divide":["age","price"]}]}
  ,"unknown_field4": {"$divide":["age","price"]}
  ,"unknown_field5": {"$plus":["age","price"]}
  ,"unknown_field6": {"$minus":["age","price"]}
  ,"unknown_field7": {"$module":["age","price"]}
  ,"$where":{
      "display_name": "UNIT TEST",
      "account": {"$similar":"account%"},
      "age":{"$gt":5,"$lt":50},
      "is_active":1,
      "roles":[1,2],
      "price":{"$between":[300,500]},
      "account_id":{"$any": [4,3]},
      "meta":{"$contain":{"b":4}},
      "$or":[
          { "account":"test_1"},
          { "account":"test_2"}
       ]
  }
  ,"$sort":{"data_registered":"DESC","account_id":"ASC"  }
  ,"$limit":10
  ,"$offset":0
}

Group by Query

{
  "gender":true
  ,"sum_operation": {"$sum":"age"}
  ,"sum_multiply_operation": {"$sum":{"$multiply":["age","price"]}}
  ,"count_operation": {"$count":1}
  ,"min_operation": {"$min":"age"}
  ,"max_operation": {"$max":"age"}
  ,"avg_operation": {"$avg":"age"}
  ,"$where":{
    "display_name": {"$like":"% display %"}
    ,"data_registered":{"$gt":"2015-09-30 21:21:31.647424+00"}
    ,"$or":[
        { "account":"account_1"}
      ,{ "account":"account_2"}
      ,{"display_name": {"$similar":"my display name"}}
    ]
  }
  ,"$limit":10
  ,"$offset":0
  ,"$group":["gender","age"]
}

Distinct Query

{
  "$distinct":["price",{"unknown_field2":{"$multiply":["age","price","price"]}}]
  ,"$limit":10
  ,"$offset":0
}

INSERT (POST)

• INSERT by JSON Query

    POST http://[host]/api/db/[table-name]
    POST http://[host]/api/db/user
    {
         "account":"my_account_1"
        ,"password":"my passowrd"
        ,"display_name":"my display name"
        ,"gender":"male"
        ,"date_registered":"2015-10-30 14:21:31.647424 -07:00:00"
        ,"struct":{"name":"full update","supplier_id":[10,50],"price":1.99} 
        ,"age":10
        ,"price":50
        ,"roles":[1,2]
        ,"is_active":true
        ,"struct":null
        ,"meta":null
      }

• BULK INSERT by JSON Query

    POST http://[host]/api/db/[table-name]
    POST http://[host]/api/db/user
    [
      {
         "account":"my_account_1"
        ,"password":"my passowrd"
        ,"display_name":"my display name"
        ,"gender":"male"
        ,"date_registered":"2015-10-30 14:21:31.647424 -07:00:00"
        ,"age":10
        ,"price":50
        ,"roles":[1,2]
        ,"is_active":true
        ,"struct":null
        ,"meta":null
      }
      ,{
         "account":"my_account_2"
        ,"password":"my passowrd"
        ,"gender":"female"
        ,"price":50
        ,"age":10
        ,"display_name":"my display name"
        ,"date_registered":"2015-10-30 14:21:31.647424 -07:00:00"
        ,"meta":
        {
          "img":"https://scontent-ord1-1.xx.fbcdn.net/v/t1.0-1/c9.0.40.40/p40x40/1618502_10203352692842640_430525865_n.jpg?oh=10b7e45293509d2b667a27f21985891f&oe=582C74C9"
        ,"gender":"male"
        ,"languages":"english"
        }
      }
    ]

• Insert in Node

    const $pgQuery = $pgConnector.query($config);
    let result = await $pgQuery.insert("user",[{"account":"my_account_1",,"password":"my passowrd"}}]);

UPSERT (PUT)

The UPDATE action to be performed in case of a conflict,otherwise do insert.
Reference https://www.postgresql.org/docs/9.5/static/sql-insert.html for more

• UPSERT by JSON Query

    PUT http://[host]/api/db/[table-name]/[constraint_name]
    PUT http://[host]/api/db/city/city_pkey
    {
        "id":1  //update if id is exist
        ,"name":"my_account_2"
        ,"district":"11213"
        ,"countrycode":"my passowrd"
        ,"population":6000
    }

• BULK INSERT by JSON Query

    PUT http://[host]/api/db/[table-name]/[constraint_name]
    PUT http://[host]/api/db/city/city_pkey
    [
      {
        "id":1 //update if id is exist
        ,"name":"my_account_2" 
        ,"countrycode":"my passowrd"
        ,"district":"11213"
        ,"population":6000
      },
      {
        "id":2 //update if id is exist
        ,"name":"my_account_1"
        ,"countrycode":"my passowrd"
        ,"district":"22222"
        ,"population":7000
      }
    ]

• Upsert in Node

    const $pgQuery = $pgConnector.query($config);
    let result = await $pgQuery.upsert("user",[{"account":"my_account_1",,"password":"my passowrd"}}],"city_pkey");

PARTIALLY UPDATES (PATCH)

• Update by JSON Query

    PATCH http://[host]/api/db/[table-name]
    PATCH http://[host]/api/db/user
    {
        "display_name":"new name",
        "age":10,
        "$where":{
            "id":{"$any":[1,2,3]}
        }
    }

• Update in Node

    const $pgQuery = $pgConnector.query($config);
    let result = await $pgQuery.update("user",{"display_name":"new name","$where":{"id":{"$any":[1,2,3]}}}});

DELETE (DELETE)

• Delete by Primary Key

    DELETE http://[host]/api/db/[table-name or view-name]/[id]
    DELETE http://[host]/api/db/user/1

• Delete by JSON Query (no need [$where])

    DELETE http://[host]/api/db/[table-name]?$q=[JSON QUERY]
    DELETE http://[host]/api/db/user?$q={"id":{"$any":[1,2,3]}}

• Delete by Query String

    DELETE http://[host]/api/db/[table-name]?[ColumnName]=[ColumnValue]
    DELETE http://[host]/api/db/user?age={"$gt":5,"$lt":50}&is_active=1

• Delete in Node

    const $pgQuery = $pgConnector.query($config);
    let result = await $pgQuery.delete("user",{"id":{"$any":[1,2,3]}});

Composite

Query composite

    GET http://[host]/api/db/composite/[composite name]
    GET http://[host]/api/db/composite/type_struct

Enum

Query enum

    GET http://[host]/api/db/enum/[enum name]
    GET http://[host]/api/db/enum/type_gender

Configuration

For security reason, sometimes you may want to disable the operation on specific table, like disable delete operation on user table. You can leverage following configuration to reslove this issue.
By default, the API will enable all operations(select,delete,insert,update) on all tables and views

const $config = {
    "connection":"postgres://<user>:<password>@<host>:<port>/<dbname>",
    "tables":{
        "user":{ 
            select:true,
            delete:false,//disable delete operation on user table,
            max_limit:5000, //set default select max results, by default is 1000
            limit:12 //set default select limit, by default is 10
         }
    },
    "composites":{
        "type_struct":false //disable type_struct composite query, 
    },
    "enum":{
        "type_gender":false //disable type_gender enum query, 
    },
    "events":{
        onRequest:function () {
            $logger.info("onRequest =>",JSON.stringify(arguments));
        },
        on_select_city_request:function () {
            $logger.info("on_select_city_request =>",JSON.stringify(arguments));
        },
    }
    "custom":{ //custom query, you can define your own script with transaction 
            "find-user":{
                "query":[
                    "select * from public.user where account_id=${id};",
                    "select * from public.city where id=${cityId}",
                    "insert into public.user(account,display_name) VALUES(${name1},${display_name1}),(${name2},${display_name2}) returning account_id",
                    "update public.user set display_name = ${updated_display_name} where account=${name1}",
                    "delete from public.user where account=${deletename}",
                ],
                "method":["GET","post"]
            }
     }
}

Events

Events life cycle, you can catch on either global level (i.e onRequest) or specific action level(i.e on_select_city_request).
Request => Build => Query => Complete

• Request: when api load
• Build: before build TSQL and verify column and parameters
• Query: before db operation
• Complete: after DB operation

        "events":{
            onRequest:function () {
                $logger.info("onRequest =>",JSON.stringify(arguments));
            },
            on_city_request:function () {
                $logger.info("on_city_request =>",JSON.stringify(arguments));
            },
            on_select_city_request:function () {
                $logger.info("on_select_city_request =>",JSON.stringify(arguments));
            },
            onBuild:function () {
                $logger.info("onBuild =>",JSON.stringify(arguments));
            },
            on_city_build:function () {
                $logger.info("on_city_build =>",JSON.stringify(arguments));
            },
            on_select_city_build:function () {
                $logger.info("on_select_city_build =>",JSON.stringify(arguments));
            },
            onQuery:function () {
                $logger.info("onQuery =>",JSON.stringify(arguments));
            },
            on_city_query:function () {
                $logger.info("on_city_query =>",JSON.stringify(arguments));
            },            
            on_select_city_query:function () {
                $logger.info("on_select_city_query =>",JSON.stringify(arguments));
            },
            onComplete:function () {
                $logger.info("onComplete =>",JSON.stringify(arguments));
            },
            on_city_complete:function () {
                $logger.info("on_select_city_complete =>",JSON.stringify(arguments));
            },            
            on_select_city_complete:function () {
                $logger.info("on_select_city_complete =>",JSON.stringify(arguments));
            },
            on_delete_city_complete:function () {
                $logger.info("on_delete_city_complete =>",JSON.stringify(arguments));
            },
        },

KeyWords

• $q
• $or
• $where
• $sort
• $limit
• $offset
• $group
• $gt
• $gte
• $lt
• $lte
• $ne
• $like
• $ilike
• $similar
• $contain
• $any
• $between
• $in
• $multiply
• $divide
• $plus
• $minus
• $module
• $sum
• $count
• $min
• $max
• $avg
• $distinct

Notice

• Column name it will be ignore if the column name does not exist in current table, but the query will be continue.

• Bit Bool type will auto convert to bit if the column type is bit.

• Node 7+ Only The library is writen on Node 7+, heavily use await/async.