DEPRECATED, WARNING: The Couchbase Group introduced collections. The main feature of this repository was to focus on one-bucket collection-like usage. Since the collections as feature already implemented in Couchbase v2, it's no longer necessary, so this repository no longer maintained.
Simple Couchbase framework.
Project specifically focuses on the one bucket as database approach, and makes it easier to manage complex data sets. It tries to get rid of the embedded jsons per document and separates them into different documents behind the scene.
This is still a work in progress. We will not take responsibility for any breaks in the code that happen after a new version comes out.
- Automatic index generator with indexable tags.
- Simple usage through the handler.
- Following best practices of Couchbase usage behind the scene, which doesn't affect the user of the library.
- Only struct can be referenced
Create a new handler with the New function, that needs a configuration.
bucket.New( {bucket.Configuration} )
type Configuration struct {
// The address of the couchbase server
ConnectionString string
// Username and password to access couchbase
Username string
Password string
// The name and password of the bucket you want to use
BucketName string
BucketPassword string
// The separator of your choice, this will separate the prefix from the field name
Separator string
}
After that you can use the Insert, Get, Remove, Upsert, Touch, GetAndTouch and Ping methods of the handler.
package main
import (
"context"
"fmt"
"github.com/PumpkinSeed/bucket"
"github.com/couchbase/gocb"
)
var conf = &bucket.Configuration{
ConnectionString: "myServer.org:1234",
Username: "cbUsername",
Password: "cbPassword",
BucketName: "testBucket",
BucketPassword: "testBucketPassword",
Separator: "::",
}
type myStruct struct {
justAField string `json:"just_a_field"`
}
func main() {
var in = &myStruct{
justAField: "basic",
}
var cas bucket.Cas
var typ = "prefix"
ctx := context.Background()
h, err := bucket.New(conf)
if err != nil {
// handle error
}
// insert
cas, id, err := h.Insert(ctx, typ, "myID", in, 0)
if err != nil {
// handle error
}
// get
var out = &myStruct{}
err = h.Get(ctx, typ, id, out)
if err != nil {
// handle error
}
// touch
err = h.Touch(ctx, typ, id, in, 0)
if err != nil {
// handle error
}
// get and touch
var secondOut = &myStruct{}
err = h.GetAndTouch(ctx, typ, id, secondOut, 0)
if err != nil {
// handle error
}
// ping
var services []gocb.ServiceType
res, err := h.Ping(ctx, services)
if err != nil {
// handle error
}
fmt.Println(res)
// upsert
in.justAField = "updated"
cas, newID, err := h.Upsert(ctx, typ, id, in, 0)
// remove
err = h.Remove(ctx, typ, newID, in)
if err != nil {
// handle error
}
}
Important:
- The typ parameter will be the prefix of the initial struct, so you should use the same value for the same types!
- IDs should be unique, if the parameter is an empty string (
""
) a globally unique ID will be automatically generated!
Embedded structs can be separated into a a new entry with the cb_referenced
tag. The value will decide the typ of the struct.
type example struct {
refMe *refMe `json:"ref_me" cb_referenced:"separate_entry"`
ignoreMe *ignoreMe `json:"ignore_me"`
}
type refMe struct {
referencedField int `json:"referenced_field"`
}
type ignoreMe struct {
notReferencedField int `json:"not_referenced_field"`
}
You can also index structs adding the cb_indexable:"true"
tag to the field, and then calling bucket.Index({context.Context}, yourStruct)
.
type example struct {
indexMe string `json:"index_me" cb_indexable:"true"`
butNotThisOne string `json:"but_not_this_one"`
}