DBOX

A multi database abstraction.

Database Support

We have driver support for:

• CSV
• Hive
• JDBC
• JSON
• Mongo
• MsSQL
• MySQL
• ODBC
• Oracle
• Postgres
• XLSX

Quick Start

Important Note

This quick start assuming mongodb as a driver. If you are using other driver, make sure you do some adjustment to your code. Especially insert and update section.

Import Database driver

First of all, we need to specify which driver we need. Just like golang sql driver, we just need to import with blank identifier. Dbox driver will initialize itself using import side-effect.

import (
"github.com/eaciit/dbox"
_ "github.com/eaciit/dbox/dbc/mongo"
)

Now our driver is initialized.

Create Connection

Next, we create our connection info struct. In this example we want to connect to MongoDB Database.

ci:= dbox.ConnectionInfo {
"127.0.0.1",
"dbname",
"user",
"pass",
nil,
}
conn, err:=dbox.NewConnection("mongo", &ci)
iferr!=nil {
panic("Connect Failed"); // Change with your error handling
}
err=conn.Connect()
iferr!=nil {
panic("Connect Failed"); // Change with your error handling
}
// Do something with database using conn

Select Query

With conn in our hand, now we can manipulate database at will. For example we want to pull data from collection "Users" that have field "_id" = "0123456789".

q, err:=conn.NewQuery().
From("Users").
Where(dbox.Eq("_id", "0123456789")).
Cursor(nil)
iferr!=nil {
panic("Query Failed");
}
// You need to close query// Protip: Make sure you check for error before defering Close()deferq.Close()

Now q contains our query cursor. We can fetch one by one or all of them at once.

// This map is only for quick example// It's better to use some struct to create stronger type checkusers:= []map[string]interface{}{}
q.Fetch(&users, 0, false)

Insert

Pulling out data is not gonna usefull without saving data. We will now learn how to insert data into our database.

First, make sure you have working conn. And then we create new insert query:

q:=conn.NewQuery().From("Users").SetConfig("multiexec", true).Save()
// Make sure q is closed when exiting functiondeferq.Close()
// This map is only for quick example// It's better to use some struct to create stronger type checkuser:=map[string]interface{}
// To enforce new data, you need to put some uniuserue identifier// Neglected to do this may replace all your data with this data// Instead of NewObjectId(), you can create your own primary key combinationuser["_id"] =bson.NewObjectId()
// Put all other field datauser["name"] ="EACIIT"// Wrap data querynewdata:=map[string]interface{}{"data": user}
// Execute insertionerr=q.Exec(newdata)
iferr!=nil {
panic("Query Failed");
}

Update

Update code is almost as same as insert. But instead of generating new Object Id, use the old Object Id instead.

// Get user object, please refer to Select Query sectionusers:= []map[string]interface{}{}
q.Fetch(&users, 0, false)
iflen(users) ==0 {
panic("User not found");
}
// Updating datanewUser:=map[string]interface{}
// Take _id from the first user row// The other way is direcly manipulate users[0] instead of creating new var// Just make sure _id is samenewUser["_id"] =users[0]["_id"]
newUser["name"] ="New User"// Create new cursorqInsert:=conn.NewQuery().From("Users").SetConfig("multiexec", true).Save()
// Make sure q is closed when exiting functiondeferqInsert.Close()
// Wrap data querynewdata:=map[string]interface{}{"data": newUser}
// Execute updateerr=qInsert.Exec(newdata)
iferr!=nil {
panic("Query Failed");
}

When mongodb encounter insertion with same _id, it will do replacement instead.

Delete

Delete is as simple as select.

// Delete queryq:=conn.NewQuery().
From("Users").
Where(dbox.Eq("_id", "0123456789")).
Delete()
// Make sure q is closed when exiting functiondeferq.Close()
// Execute deleteerr=q.Exec(nil)
iferr!=nil {
panic("Query Failed")
}