Warning Project will be archived as read-only. This project still can be use, But there will be no update until unknown date announcement.
Parsing Kendo DataSource filter & sort request to any kind of data in go that can immediately serve to mongo query or aggregate and other ORMs. Already used in EACIIT's environment and can be scale-able for generic purpose. Currently there are few parser available:
- mongo-go-driver Filter (mainly used for primary test)
- eaciit/dbox Filter
- eaciit/dbox Pipe or Aggregation type
- sebar/dbflex Filter
- xorm/builder Filter This builder have nice ToSQL function, so you can transform into SQL
- Parse Kendo DataSource request into go struct
- Parse Kendo DataSource request into filter & sort database / orm-ish
- Transforming filter + Chaining
- Customizable filter operator
- Additional Pre-parse filter
- Adaptor based, customizable parse into other generic purpose
- Kendo Request Wrapper
- Go mod support
- Install
go get -v raditzlawliet/kendoparser- Import
import (
kp "github.com/raditzlawliet/kendoparser"
kpmongo "github.com/raditzlawliet/kendoparser/parser/kpmongo"
)- Use
request := struct {
Filter kp.Filter,
Sort kp.Sort,
}{}
filter := request.Filter.Parse(kpmongo.FilterParser)
sort := request.Sort.Parse(kpmongo.SortParser)- JSON Sample
{
"data": {
"filter": {
"field": "id",
"operator": "eq",
"value": "val"
}
}
}- GO Implementation
import kpmongo "github.com/raditzlawliet/kendoparser/parser/kpmongo"// just for information
// you can use gorilla mux or knot
var k = k *knot.WebContext
// retrieve payload
payload := &KendoRequest{}
k.GetPayload(payload)
// Usually payload the struct will be this
payload := KendoRequest {
Data: KendoData{
Filter: KendoFilter{
Field: "_id", Operator: "eq", Value: "val"
},
},
}
filter := payload.Data.Filter.Parse(kpmongo.FilterParser)
sort := payload.Data.Sort.Parse(kpmongo.SortParser)- More Example
{
"data": {
"filters": [
{
"filter": {
"field": "id",
"operator": "eq",
"value": "val"
}
},
{
"filter": {
"field": "abc",
"operator": "in",
"values": ["a", "b"]
}
}
],
"logic": "and"
}
}import kpmongo "github.com/raditzlawliet/kendoparser/parser/kpmongo"
// just for information
// you can use gorilla mux or knot
var k = k *knot.WebContext
// retrieve payload
payload := &KendoRequest{}
k.GetPayload(payload)
// Usually payload the struct will be like this
payload := KendoRequest {
Data: KendoData{
Filter: KendoFilter{
Filters: []KendoFilter{
Filter: KendoFilter{
Field: "_id", Operator: "eq", Value: "val"
},
Filter: KendoFilter{
Field: "abc", Operator: "in", Values: []interface{}{"a", "b"}
},
}
},
},
}
filter := payload.Data.Filter.Parse(kpmongo.FilterParser)
sort := payload.Data.Sort.Parse(kpmongo.SortParser)- kpmongo - mongo-go-driver
- equal, notequal, equaldate, notequaldate, contain, notcontain, in, notin, gt, lt, gte, lte, gtdate, ltdate, gtedate, ltedate, exists, dateexists, between
- kpdbox - eaciit/dbox
- equal, notequal, contain, notcontain, in, gt, lt, gte, lte, gtedate, ltedate, exists, between
- kpdboxpipe - eaciit/dbox
- equal, notequal, contain, notcontain, in, gt, lt, gte, lte, gtedate, ltedate, exists, between
- kpdbflex - sebar/dbflex
- equal, notequal, contain, notcontain, in, gt, lt, gte, lte, gtedate, ltedate, exists, between
- kpxorm - xorm/builder
- equal, notequal, contain, notcontain, in, gt, lt, gte, lte, gtedate, ltedate, exists, between
Operator Manager mainly used for filter parser to parsing the data based on operator. In available parser they also have each Operator Manager. You also can create your own Operator Manager and you can also set scoped Operator Manager for spesific payload
- Extending Existing Operator Manager
import kpmongo "github.com/raditzlawliet/kendoparser/parser/kpmongo"
// Create the func
func IsOne(kf kendoparser.Filter) interface{} {
return bson.M{kf.Field: bson.M{"$eq": 1}}
}
// Assign into Exisiting Operator Manager
kpmongo.OperatorManager.RegisterOperator(IsOne, "isone")
// Example kendo json request
"filter": {
"field": "id",
"operator": "isone"
}
// Parse
request := struct {
Filter kp.Filter,
}{}
filter := request.Filter.Parse(kpmongo.FilterParser)
// Result
{"id" : {"$eq": 1}}-
Creating Your own Operator Manager
Usually when you try to create your own Operator Manager, you should create your own Parser. if the parser already existed, i recommended you use the existing Operator Manager.
// creating global package variable
var (
// OperatorManager of Mongo Parser
OperatorManager = new(kendoparser.OperatorManager)
// Operator bundle of Mongo Parser
Operator = OperatorBundle{}
)
// Wrapper Bundle, so you can put all the func filter into this struct
type OperatorBundle struct{}
// Register when import
func init() {
RegisterOperator()
}
// Registration Process
func RegisterOperator() {
OperatorManager.SetDefaultOperator(Operator.Equal)
OperatorManager.RegisterOperator(Operator.Equal, "eq", "equal")
}
// Some Filter
func (o *OperatorBundle) Equal(kf kendoparser.Filter) interface{} {
if value, ok := kf.Value.(string); ok && kf.IgnoreCase {
value := regexp.QuoteMeta(value)
return bson.M{kf.Field: primitive.Regex{Pattern: "^" + strings.ToLower(value) + "$", Options: "i"}}
}
return bson.M{kf.Field: bson.M{"$eq": kf.Value}}
}Need modify your field? lowercase all field before processing? don't worry, you can use Transform to modify and apply to your all field. See kendo_test.go) for more uses
import kpmongo "github.com/raditzlawliet/kendoparser/parser/kpmongo"
kendoFilter := KendoFilter{}
// transform filter field or all field in filters into lower case
kendoFilter.TransformField(strings.ToLower) // only current filter
kendoFilter.TransformFieldAll(strings.ToLower) // include filters
// custom transform include value (if needed), it use pointer KendoFilter, so you can direct modify. No need to return
transformIDMongo := func(kf *KendoFilter) {
if kf.Field == "id" {
kf.Field = "_id"
}
}
kendoFilter.Transform(transformIDMongo) // only current filter
kendoFilter.TransformAll(transformIDMongo) // include filters
// chaining is possible
kendoFilter.TransformFieldAll(strings.ToLower).TransformAll(transformIDMongo).Parse(kpmongo.FilterParser)You can also add additional single handler before building filter by registered operator. This approach you can add custom direct filter within loop filter. pre-filter it self act same like parser, so you can chain the filter with other filter. But if your pre-filter already return value, the last parse only act return the value
import kpmongo "github.com/raditzlawliet/kendoparser/parser/kpmongo"
beforeParser := func (kf *KendoFilter) interface{} {
return nil
}
// dbox filter
resultFilter := kendoFilter.TransformAllField(strings.ToLower).
TransformAll(func(kf *KendoFilter) {
if kf.Field == "id" {
kf.Field = "_id"
}
}).
AddParse(beforeParser).
Parse(kpmongo.FilterParser)
// reset if needed another
kendoFilter.ResetAdditionalParsers()
kendoFilter.ResetAllAdditionalParsers() // Recursive to all filter childs
// dbox pipe
resultFilterPipe := kendoFilter.TransformAllField(strings.ToLower).
TransformAll(func(kf *KendoFilter) {
if kf.Field == "id" {
kf.Field = "_id"
}
}).
AddParse(beforeParser).
Parse(kpmongo.FilterParser)do you need sort? You can do it easly.
import kpmongo "github.com/raditzlawliet/kendoparser/parser/kpmongo"
sort := kData.Sort.Parse(kpmongo.SortParser)-
go mod tidy keep error
git.eaciitapp.com/sebar/dbflex: no matching versions for query "latest"This error appear because I use some private database driver in this parser, but you can exclude this using go.mod.
Modify go.mod file, if you dont have the package, just use this dummy in your go.mod, AND If you have the package and using it, replace it into your correct path package:
replace git.eaciitapp.com/sebar/dbflex => ./pkg/git.eaciitapp.com/sebar/dbflex
Feel free to contribute, don't forget to mention if needed
MIT License
Radityo