聚合api编程之自定义kubectl输出字段的方法:metav1.Table

字数统计: 920阅读时长: 4 min

 2022/05/22   Share

kubectl输出的格式一般是设置好的写死的也就是各个资源的列表头的定义都是写死的具体代码在

github.com/kubernetes/kubernetes/-/blob/pkg/printers/internalversion/printers.go

验证: 当我们输入

1	kubectl describe pod xxx -v=10

可以看到kubectl的输出是转换成了table这种type形式输出的

如果kubectl无法解析成table的话, 他会自己解析成一个 no-Table Object 也就是说只显示NAME和AGE

下面是个例子

....
import (
	"github.com/haojianxun/k8s-aa/pkg/apis/myingress/v1beta1"
	metav1 "k8s.io/apimachinery/pkg/apis/meta/v1"
)
var (
	Table_ListColumns=[]string{"name","namespace","path","host"}
)
// 把列表 或单资源 变成表格化
func ConvertToTable(obj interface{}) *metav1.Table{

	t:= &metav1.Table{}
	t.Kind="Table"
	t.APIVersion="meta.k8s.io/v1"
	if v,ok:=obj.(*v1beta1.MyIngressList);ok{  //代表取列表
		//设置表头
		th:=make([]metav1.TableColumnDefinition,len(Table_ListColumns))
		for i,h:=range Table_ListColumns{
			format:="name"

			th[i]=metav1.TableColumnDefinition{Name: h,Type: "string",Format:format}
		}
		t.ColumnDefinitions=th //设置表头
		//设置 行  数据
		rows:=make([]metav1.TableRow,len(v.Items))
		for i,item:=range v.Items{
			rows[i]=metav1.TableRow{
				Cells: []interface{}{item.Name,item.Namespace,
					item.Spec.Path,item.Spec.Host},
			}
		}
		t.Rows=rows
	}
	return t

}

....

总之想显示自定义的列的话先构造一个metav1.Table

那Table的type长什么样子呢

type Table struct {
	TypeMeta `json:",inline"`
	// Standard list metadata.
	// More info: <https://git.k8s.io/community/contributors/devel/sig-architecture/api-conventions.md#types-kinds>
	// +optional
	ListMeta `json:"metadata,omitempty"`

	// columnDefinitions describes each column in the returned items array. The number of cells per row
	// will always match the number of column definitions.
	ColumnDefinitions []TableColumnDefinition `json:"columnDefinitions"`
	// rows is the list of items in the table.
	Rows []TableRow `json:"rows"`
}

可以看到我们要做的就是把 ColumnDefinitions 去填充自己定义的想打印的列

ColumnDefinitions的定义如下

type TableColumnDefinition struct {
	// name is a human readable name for the column.
	Name string `json:"name"`
	// type is an OpenAPI type definition for this column, such as number, integer, string, or
	// array.
	// See <https://github.com/OAI/OpenAPI-Specification/blob/master/versions/2.0.md#data-types> for more.
	Type string `json:"type"`
	// format is an optional OpenAPI type modifier for this column. A format modifies the type and
	// imposes additional rules, like date or time formatting for a string. The 'name' format is applied
	// to the primary identifier column which has type 'string' to assist in clients identifying column
	// is the resource name.
	// See <https://github.com/OAI/OpenAPI-Specification/blob/master/versions/2.0.md#data-types> for more.
	Format string `json:"format"`
	// description is a human readable description of this column.
	Description string `json:"description"`
	// priority is an integer defining the relative importance of this column compared to others. Lower
	// numbers are considered higher priority. Columns that may be omitted in limited space scenarios
	// should be given a higher priority.
	Priority int32 `json:"priority"`
}

至于想定义那些列你可以自己去定义 , 下面给个源码中kubectl输出deployment资源的列的定义

可以参考下deployment是怎么定义的

....

deploymentColumnDefinitions := []metav1.TableColumnDefinition{
		{Name: "Name", Type: "string", Format: "name", Description: metav1.ObjectMeta{}.SwaggerDoc()["name"]},
		{Name: "Ready", Type: "string", Description: "Number of the pod with ready state"},
		{Name: "Up-to-date", Type: "string", Description: extensionsv1beta1.DeploymentStatus{}.SwaggerDoc()["updatedReplicas"]},
		{Name: "Available", Type: "string", Description: extensionsv1beta1.DeploymentStatus{}.SwaggerDoc()["availableReplicas"]},
		{Name: "Age", Type: "string", Description: metav1.ObjectMeta{}.SwaggerDoc()["creationTimestamp"]},
		{Name: "Containers", Type: "string", Priority: 1, Description: "Names of each container in the template."},
		{Name: "Images", Type: "string", Priority: 1, Description: "Images referenced by each container in the template."},
		{Name: "Selector", Type: "string", Priority: 1, Description: extensionsv1beta1.DeploymentSpec{}.SwaggerDoc()["selector"]},
	}
	t := &metav1.Table{
		ColumnDefinitions: deploymentColumnDefinitions,
	}
	t.Kind = "Table"
	t.APIVersion = "meta.k8s.io/v1"

.....

table的Kind固定为”Table” 里面的metav1.TableColumnDefinition 里面的东西自己去自定义格式

之后定义完了格式之后就可以填充数据了填充数据的话用的就是Table中的Rows

上面的例子的填充的地方在

for i,item:=range v.Items{
		rows[i]=metav1.TableRow{
			Cells: []interface{}{item.Name,item.Namespace,
				item.Spec.Path,item.Spec.Host},
		}
	}

就是构造出一个

1
2
3

metav1.TableRow{
	Cells: []interface{}{这里按照列的顺序依次填入想填的数据即可}
}

原文作者：HAOJX

原文链接：haojianxun.github.io/2022/05/22/聚合api编程之自定义kubectl输出字段的方法metav1.Table/

发表日期：May 22nd 2022, 12:00:00 am

更新日期：October 17th 2022, 1:42:52 am

Next Post

cue基本语法
Previous Post

取任意yaml资源转化为k8s资源的完全通用写法(kubectl源码的写法)

CATALOG

