打印的python的dict或list对象用FEHelper查看

Tech

利用Python的kubernetes库，可以查询出来某个namespace里下的pod,但是返回的结果集过于复杂，想找到其中的pod IP, 需要费点时间。

1
2
3
all_pods = client.CoreV1Api().list_namespaced_pod(
                CURRENT_RUNNING_NAMESPACE
                ).to_dict()["items"]

all_pods直接打的结果不是json，里面的None和datetime.datetime是无法被FEHelper识别的。
json直接dump的时候会报错误：TypeError: Object of type datetime is not JSON serializable。
这个时候只需要简单的加个参数就可以了。

1
json.dumps(all_pods, sort_keys=True, default=str)

打印出来的结果，拷贝到FEHelper然后去掉头尾的引号就可以识别了, 顺利找出pod IP的路径，拼出来想要的结果。

1
[s['status']['pod_ip'] for s in all_pods]

在K8S环境中构建基于python flask架构的websocket

Tech

业务上需要搭建一个能推送消息的架构，现有的服务端是基于python-flask构建的。

代码部分

服务端

服务端python需要集成socketio，参考这里https://github.com/miguelgrinberg/python-socketio/blob/main/examples/server/wsgi/app.py
但是需要记得async_mode改成gevent。

前端

前端测试的代码, 需要替换一下Server IP。

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
<!DOCTYPE HTML>
<html>
<head>
    <title>Flask-SocketIO Test</title>
    <script type="text/javascript" src="//code.jquery.com/jquery-2.1.4.min.js"></script>
    <script type="text/javascript" src="//cdnjs.cloudflare.com/ajax/libs/socket.io/3.0.3/socket.io.min.js"></script>
    <script type="text/javascript" charset="utf-8">
        $(document).ready(function(){
            var socket = io.connect("http://<Server IP>:5000/");

            socket.on('connect', function() {
                socket.emit('my_event', {data: 'I\'m connected!'});
            });
            socket.on('disconnect', function() {
                $('#log').append('<br>Disconnected');
            });
            socket.on('nuke_response', function(msg) {
                $('#log').append('<br>Received: ' + msg.data);
            });

            // event handler for server sent data
            // the data is displayed in the "Received" section of the page
            // handlers for the different forms in the page
            // these send data to the server in a variety of ways
            $('form#emit').submit(function(event) {
                socket.emit('my_event', {data: $('#emit_data').val()});
                return false;
            });
            $('form#broadcast').submit(function(event) {
                socket.emit('my_broadcast_event', {data: $('#broadcast_data').val()});
                return false;
            });
            $('form#join').submit(function(event) {
                socket.emit('join', {room: $('#join_room').val()});
                return false;
            });
            $('form#leave').submit(function(event) {
                socket.emit('leave', {room: $('#leave_room').val()});
                return false;
            });
            $('form#send_room').submit(function(event) {
                socket.emit('my_room_event', {room: $('#room_name').val(), data: $('#room_data').val()});
                return false;
            });
            $('form#close').submit(function(event) {
                socket.emit('close_room', {room: $('#close_room').val()});
                return false;
            });
            $('form#disconnect').submit(function(event) {
                socket.emit('disconnect_request');
                return false;
            });
        });
    </script>
</head>
<body>
    <h1>Flask-SocketIO Test</h1>
    <h2>Send:</h2>
    <form id="emit" method="POST" action='#'>
        <input type="text" name="emit_data" id="emit_data" placeholder="Message">
        <input type="submit" value="Echo">
    </form>
    <form id="broadcast" method="POST" action='#'>
        <input type="text" name="broadcast_data" id="broadcast_data" placeholder="Message">
        <input type="submit" value="Broadcast">
    </form>
    <form id="join" method="POST" action='#'>
        <input type="text" name="join_room" id="join_room" placeholder="Room Name">
        <input type="submit" value="Join Room">
    </form>
    <form id="leave" method="POST" action='#'>
        <input type="text" name="leave_room" id="leave_room" placeholder="Room Name">
        <input type="submit" value="Leave Room">
    </form>
    <form id="send_room" method="POST" action='#'>
        <input type="text" name="room_name" id="room_name" placeholder="Room Name">
        <input type="text" name="room_data" id="room_data" placeholder="Message">
        <input type="submit" value="Send to Room">
    </form>
    <form id="close" method="POST" action="#">
        <input type="text" name="close_room" id="close_room" placeholder="Room Name">
        <input type="submit" value="Close Room">
    </form>
    <form id="disconnect" method="POST" action="#">
        <input type="submit" value="Disconnect">
    </form>
    <h2>Receive:</h2>
    <div><p id="log"></p></div>
</body>
</html>

这样就基本完成了正常环境下的websocket下的代码部分的功能。

架构部分

架构图

arch

如图所示，几条重点：

  1. 系统发布在一个K8S环境中，一个application的deployment里包含三个pod和一个service。
  2. K8S Sevice用NodePort的方式进行服务暴漏。
  3. K8S系统外面是用一个HAProxy进行代理，域名解析到HAProxy所在的虚拟机的IP上。
  4. HAProxy和K8S service形成了两层的LoadBalance。

目的是让客户端Client A, B, C访问到一个POD后，以后就一直绑定到这个pod上。
需要改动的配置：

  1. HAProxy的balance策略改成source
    这样就让HAProxy进行转发的时候根据客户端ip进行选择目的IP。
  2. K8S Service的配置里增加sessionAffinity: ClientIP
    K8S Service进行转发的时候根据客户端ip进行选择目的IP
  3. K8S Service的配置里增加externalTrafficPolicy: Local
    如果不加这一条，那么所有的ClientIP都会被认为是来自HAProxy的那个IP，加上这个配置后，会根据请求的header里的X-Forwarded-For里的客户端IP进行判断。

消息队列

正常情况下，客户端A和B链接到第一个pod，客户端C链接到第三个pod，如果有一个时间发生在第二个POD，或者Job Pod上，是无法直接发送消息给所有的pod上的。
所以需要一个消息总线，可以选择Redis/MQ/kube-event来实现，三个API pod侦听消息队列的某个时间，所有需要发送给客户的消息直接发送给消息队列，然后消息队列转发给三个API pod，
API pod收到消息队列的提醒后，然后推送给自己链接的客户端，这样就完成了一个整体回路。

pip安装是的时候遇到Microsoft Visual C++ 14.0 or greater is required

Tech

pip安装部分包的时候，会遇到下面的错误提示，这是因为缺少的C++的编译环境。

1
2
error: Microsoft Visual C++ 14.0 or greater is required. Get it with "Microsoft C++ Build Tools": https://visualstudio.microsoft.com/visual-cpp-build-tools/
 [end of output]

网上找到的许多信息看起来已经过期了，目前能解决的办法是，打开这个错误提示里的网站。

https://visualstudio.microsoft.com/visual-cpp-build-tools/

下载一个“Download Build Tools”， 然后安装，从里面选择“Windows”里的“Desktop development with C++”, 然后确认安装，等待就可以了。

查看kubectl config文件中的权限信息

Tech

有时候调试问题的时候，只有一份kubeconfig文件，可以根据这个文件获取当前用户的权限，方便下一步的调试。

把config文件中的client-certificate-data内容，或者.crt文件拿出来用openssl解析一下。

1
2
3
4
5
6
# 直接处理内容
$ echo "LS0tLS1CRUdJTiBDRVJUSUZJQ0FURS0tLS0tCk1JSURJVENDQWdtZ0F3SUJBZ0lJWkVMQmxCUDl4dmN3RFFZSktvWklodmNOQVFFTEJRQXdGVEVUTUJFR0ExVUUKQXhNS2EzVmlaWEp1WlhSbGN6QWVGdzB5TXpBeU1EZ3dOak0xTWpKYUZ3MHlOREF6TURVd016QXdNREZhTURReApGekFWQmdOVkJBb1REbk41YzNSbGJUcHRZWE4wWlhKek1Sa3dGd1lEVlFRREV4QnJkV0psY201bGRHVnpMV0ZrCmJXbHVNSUlCSWpBTkJna3Foa2lHOXcwQkFRRUZBQU9DQVE4QU1JSUJDZ0tDQVFFQXRubG93TUYxeXVHS1c2R2QKVVp3M09OQ3pqVGcyUDRpbU00SHZlU29ZNFdHRmRMQitnQ2tPSmtCTUpYYU92c2UrYStwOVhMYVJjVUwwN2RWdQo5NUVQYStIUWx3VE5UYVRZRXRtNGFHZHhOMUY5L0pxejZGejVXeUZKSTBnOEVCNDVEVGw1WVhuV0E0YzlXcHBDCmF3Y0N2c0lvK0RoRWowSjFMeFlMTHhDSTVha05pWWdKYlJ5NDhiallSengrdDBoUjJEUjZCRmtibUg4a1Z6NmwKNFduSDRHZGhJRWNhV1g5UUxaQnY1QmNDRDRoN3J0V0J1WFd5NVRtUGViRE1nTC9aQmVERVBUTGFRRUVZSDNkbwpGM1dFcGV5QXFsMHc4UHVzWnllRnpIcUY3c2lzYnZWaFAxN2xhSzNtRmloSDNkVG0rN3RieWNtRnQzTk1RVzZjCmtwbkVrd0lEQVFBQm8xWXdWREFPQmdOVkhROEJBZjhFQkFNQ0JhQXdFd1lEVlIwbEJBd3dDZ1lJS3dZQkJRVUgKQXdJd0RBWURWUjBUQVFIL0JBSXdBREFmQmdOVkhTTUVHREFXZ0JRaWN3TzROU2hyeU1ib0VVVUNCTjNVU1RMZgo4akFOQmdrcWhraUc5dzBCQVFzRkFBT0NBUUVBVDF5MEtFa1Y5Sm05aU5TRklpeE5raXRjTFVwamJCS2FOSVNNCjI3T2hLWFZWRXdORzRuaERHanQrZFZNa2RhTTFvVlpqRVg2M09CeFdxYnFZNUhSWnZGZi9PUkpGaXo2cUxlbmIKSlJxMk5zSWswdHdpdTZXeFB5WERHOFcxenEvZHdxcmFQOVFlZGd1SFRQMGgrSHd6TFVzalJFSE9SanNOYzUvawpUbWM4andnN2VSckNUamU2Ym41RlNuam1CU0ZOaGM0TGlFZXllVXVpREViOVI2ZG96U1RrdFZHWEFldnlHT1BzCk1PcWJTRTRsT3FsbnZWcTVFd3J0N3ZhZEdmYzRNRDdCWnFLdzBNdkhPckFnb251T1BnM1UrWDlUODNhQ2NZVkgKcmxKaWdybkowc3ZkYjI2WEhKWDZiU3ZnWFpZb3c2S1pGY1liOTdJQURUVDBma0pNZUE9PQotLS0tLUVORCBDRVJUSUZJQ0FURS0tLS0tCg==" | base64 -d | openssl x509 -noout -text

# 处理.crt文件

$ openssl x509 -in cluster.crt -noout -text

openssl输出的内容里包含如下一段：

1
Subject: O=system:masters, CN=kubernetes-admin

这里面的O=对应的就是K8S里的Group，CN=对应的就是user, 剩下的就是cluster里去找对应rolebinding, clusterrolebinding, role和clusterrole就可以了。

用sed替换掉特定字符串后的指定内容

Tech

工作中遇到了这么一个需求，有一段没有格式的json字符串，需要把其中的key为“name”的值改成“new-webhook-config”, 用sed可以完成这个工作。

1
{"apiVersion":"admissionregistration.k8s.io/v1beta1","kind":"MutatingWebhookConfiguration","metadata":{"annotations":{},"name":"webhook-config"}}
1
sed -ri 's/(\"name\":\")[^"]*/\1new-webhook-config/g' 1.txt

重点的是其中的\1，这里表示的第一个子串，如果没有这个\1，那么就会全部替换掉。
可以参考这里：
https://www.cnblogs.com/maxincai/p/5146338.html

调用Jira API统计一段时间每个账户内ticket被resolved的数量

Tech

Jira是不能分组统计的，只能用dashboard里用二维表格统计，但是如果查询语句是“status changed to resolved by”，统计表格里是不能显示出来“resolved by”这个用户名的。
之前同事是手动做了特别多的查询，然后把这些查询放到confluence里进行统计的，所以就想用jira api分账户统计一下，代码如下。

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
import json
from functools import reduce

import requests
from prettytable import PrettyTable

token = 'xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx' # 从JIRA的Porta里profile页面创建一个Personal Access Tokens
headers = {
    "Authorization": "Bearer %s" % token,
    "Accept": "application/json, text/plain, */*",
    "Content-Type": "application/json;charset=UTF-8",
}
duiring_time = "2022-01-01, 2023-01-01"
jira_user_accounts = [
    'zhangsan',
    'lisi',
    'wangwu'
]


if __name__ == '__main__':
    requests.packages.urllib3.disable_warnings()
    total_result = []
    for jira_user in jira_user_accounts:
        params = {
            "startAt": 0,
            "maxResults": 2000,
            "jql": "project = CSFS AND status changed to resolved by (%s) during (%s)" % (jira_user, duiring_time),
            "fields": ['customfield_40669']
        }
        res = requests.post("https://jiradc2.ext.net.nokia.com/rest/api/2/search", headers=headers,
                            data=json.dumps(params, ensure_ascii=False).encode("utf-8"), verify=False)
        issues = res.json()['issues']
        if len(issues) > 0:
            display_name = "Unknown"
            # 想办法找到所有的display name
            try:
                for issue in issues:
                    resolve_users = issue['fields']["customfield_40669"]
                    for resolve_user in resolve_users:
                        # 这个ticket可能会被多个人resolve，取最后一个resolver
                        if resolve_user['name'] == jira_user:
                            display_name = resolve_user["displayName"]
                            raise StopIteration
            except StopIteration:
                total_result.append({"name": jira_user, 'count': len(issues), "display_name": display_name})
        else:
            total_result.append({"name": jira_user, 'count': 0})

    total_result.sort(key=lambda u: u["count"], reverse=True)

    total_count = reduce(lambda x, y: x + y["count"], total_result, 0)
    print("total: " + str(total_count))
    myTable = PrettyTable(["user", "display name", "count"])
    for result in total_result:
        if 'display_name' in result:
            myTable.add_row([result['name'], result['display_name'], result['count']])
        else:
            myTable.add_row([result['name'], '', result['count']])
    print(myTable)

查询后的结果：

1
2
3
4
5
6
7
+----------+-----------------------+-------+
|   user   |        dispaly        | count |
+----------+-----------------------+-------+
| zhangsa  |   San Zhang (Company) |  302  |
| lisi     |     Si Li   (Company) |  113  |
| wangwu   |    Wu Wang (Company)  |   93  |
+----------+-----------------------+-------+

效果不错，值得分享一下。

学习Python的列表生成器以及Map-Reduce

Tech

最近写了很多的接口，需要反复处理json数据，抽离数据，重组格式。
花时间研究了一些python的列表生成，以及更高级的Map/Reduce, 并给大家分享了一下，记录一下share的过程写下的代码。

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
from functools import reduce

datas = [
    {
        "date": "22日星期四",
        "sunrise": "06:17",
        "high": "高温 17.0℃",
        "low": "低温 1.0℃",
        "sunset": "18:27",
        "aqi": 98,
        "fx": "西南风",
        "fl": "<3级",
        "type": "晴",
        "notice": "愿你拥有比阳光明媚的心情"
    },
    {
        "date": "23日星期五",
        "sunrise": "06:16",
        "high": "高温 18.0℃",
        "low": "低温 5.0℃",
        "sunset": "18:28",
        "aqi": 118,
        "fx": "无持续风向",
        "fl": "<3级",
        "type": "多云",
        "notice": "阴晴之间，谨防紫外线侵扰"
    },
    {
        "date": "24日星期六",
        "sunrise": "06:14",
        "high": "高温 21.0℃",
        "low": "低温 7.0℃",
        "sunset": "18:29",
        "aqi": 52,
        "fx": "西南风",
        "fl": "<3级",
        "type": "晴",
        "notice": "愿你拥有比阳光明媚的心情"
    },
    {
        "date": "25日星期日",
        "sunrise": "06:13",
        "high": "高温 22.0℃",
        "low": "低温 7.0℃",
        "sunset": "18:30",
        "aqi": 71,
        "fx": "西南风",
        "fl": "<3级",
        "type": "晴",
        "notice": "愿你拥有比阳光明媚的心情"
    },
    {
        "date": "26日星期一",
        "sunrise": "06:11",
        "high": "高温 21.0℃",
        "low": "低温 8.0℃",
        "sunset": "18:31",
        "aqi": 97,
        "fx": "西南风",
        "fl": "<3级",
        "type": "多云",
        "notice": "阴晴之间，谨防紫外线侵扰"
    }
]


def add(x):
    return x + 'Z'


def sum(x, y):
    return x + y


if __name__ == '__main__':
    zz = []
    for data in datas:
        zz.append(data['date'] + data['high'] + data['low'])
    print(zz)

    p = 9
    a = 10
    zz = p if p > a else a
    print(zz)

    zz = [data['date'] + data['high'] + data['low'] for data in datas if data["aqi"] > 90]
    print(zz)

    zz = ['不好' if data["aqi"] > 90 else '不好' for data in datas]
    print(zz)

    zz = [{data['date']: data['high']} for data in datas]
    print(zz)

    # zz = {data['date']: data['high'],data['sunrise']: data['sunset'] for data in datas}
    # print(zz)

    # MAP
    list_a = ['A', 'B', 'C']

    list_b = [1, 2, 3]
    # map --  ['A+','B+','C+']
    # reduce -- D= 'A+'+'B+'+'C+'

    zz = list(map(add, list_a))
    print(zz)

    zz = reduce(sum, list_b, 6)
    print(zz)

    zz = list(map(lambda x: x + 'Z', list_a))
    print(zz)

    zz = reduce(lambda x, y: x + y, list_b, 6)
    print(zz)

    zz = list(
        map(lambda x: {x['date']: x['fx'], x['aqi']: x['sunset'], 'status': '好天气' if x['aqi'] < 90 else '坏天气'},
            datas))
    print(zz)

    zz = []
    for x in datas:
        d = {x['date']: x['fx'], x['aqi']: x['sunset']}
        if x['aqi'] < 90:
            d['status'] = '好天气'
        else:
            d['status'] = '坏天气'
        zz.append(d)
    print(zz)

    zz = reduce(lambda d, k: {**d, **{k['date']: k['fx'], k['sunrise']: k['sunset']}}, datas, {})
    print(zz)

    a = {'22日星期四': '西南风', 98: '18:27', 'status1': '坏天气'}
    b = {'23日星期五': '无持续风向', 118: '18:28', 'status2': '天气'}

    print(a.update(b))
    print({**a, **b})

    datas = [(3, 5), (6, 8), (1, 9), (4, 8)]

    zz = reduce(lambda d, k: (d[0] + k[0], d[1] + k[1]), datas, (0, 0))
    print(zz)

通过docker registry的文件夹获取到所有的镜像

Tech

工作需要获取docker registry里所有的镜像和tag，但是系统在磁盘被占满的情况下出现无法工作的情况，无法通过dockers registry的API去获取。
所以尝试通过，registry的文件夹目录结构中取出想要的结果，最后证明可行。

1
2
3
4
5
6
export docker_registry_dir=/opt/bcmt/storage/docker-registry/ # docker registry 的安装目录。

cd ${docker_registry_dir}docker/registry/v2/repositories/; ls -R  | grep tags/ | grep -v "current"| grep -v "index" | sort | awk '{a=index($0,"./");b=index($0,"/_manifests/tags/");c=index($0,":"); print substr($0,a+2,b-a-2)":"substr($0,b+17,c-b-17)}' 

# 多行换成一行，用逗号分隔
cd ${docker_registry_dir}docker/registry/v2/repositories/; ls -R  | grep tags/ | grep -v "current"| grep -v "index" | sort | awk '{a=index($0,"./");b=index($0,"/_manifests/tags/");c=index($0,":"); print substr($0,a+2,b-a-2)":"substr($0,b+17,c-b-17)}' | tr '\n' ','

之前都是用notepad++手动替换处理，这次发现结合这几个命令awk, print, substr, tr就可以搞定，不错！

Cobra支持动态自定义flag

Tech

用GO实现了一个binary，这个binary会发送API请求去服务端，服务端完成对K8S环境的部署工作。
由于后端进行了大量封装，并且参数是非常多的，所以有了一个想法，具体的要求如下：

  1. 支持自定义flag列表
  2. help信息支持展示自定义的flag列表
  3. 列表信息来自于一个专门的API

基于这些需求，查询了很多文档后，并没有成功的例子，经过一下午的尝试，最后实现了预期的需求，分享一下。
回头准备写一篇post给cobra。

在Cobra的github repo里开了issue，询问是否有更加优雅和方便的方法，没想到Cobra的Collaborator竟然回复了，看起来现在的实现方法是一个正确的方向。
https://github.com/spf13/cobra/issues/1758

API部分， python flask实现的，简单定义了数据结构，然后返回一个json结果。

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
@ns.route('/custom-api')
class CustomAPI(Resource):
    def get(self):
        flags_list = {
            "cert": {
                "type": "string",
                "des": "add self-signed certs for private registries. Format: registry-name1:/path/to/cert-file1,registry-name2:/path/to/cert-file2,...",
                "default": ""
            },
            "csf-grocery": {
                "type": "bool",
                "des": "set up CSF Grocery",
                "default": True
            }
        }
        return json_success(flags_list)
utils/customized_flags.go
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
package utils

import (
	"fmt"
	"os"
	"strings"

	resty "github.com/go-resty/resty/v2"
	"github.com/spf13/cobra"
	"github.com/spf13/viper"
	"github.com/tidwall/gjson"
)

var customizedMap = make(map[string]interface{})

func ParseCutomizedFlags(url string, requiredFlags []string, cmd *cobra.Command, args []string) map[string]interface{} {

	if len(requiredFlags) > 0 && len(args) == 0 {
		cmd.Help()
		fmt.Println()
		os.Exit(0)
	}

	argsArrayString := strings.Join(args, "#")
	for _, flag := range requiredFlags {
		if !strings.Contains(argsArrayString, "--"+flag) {
			fmt.Printf("required flag(s) \"%v\" not set\n", flag)
			cmd.Help()
			fmt.Println()
			os.Exit(1)
		}
	}

	fullURL := viper.GetString("datacenter-server-api") + url
	var customizedData gjson.Result
	client := resty.New()
	resp, err := client.R().ForceContentType("application/json").EnableTrace().Get(fullURL)
	if err == nil {
		if gjson.Get(resp.String(), "status").String() == "OK" {
			customizedData = gjson.Get(resp.String(), "data")
		}
	} else {
		fmt.Println(err.Error())
		os.Exit(1)
	}

	if len(customizedData.Map()) > 0 {
		for key_name, valueMap := range customizedData.Map() {
			value_type := valueMap.Get("type").String()
			if value_type == "string" {
				cmd.Flags().String(string(key_name), valueMap.Map()["default"].String(), valueMap.Map()["des"].String())
			} else if value_type == "bool" {
				cmd.Flags().Bool(string(key_name), valueMap.Map()["default"].Bool(), valueMap.Map()["des"].String())
			}
		}
	}

	cmd.DisableFlagParsing = false
	err = cmd.ParseFlags(args)
	if err != nil {
		fmt.Println(err)
		cmd.Help()
		fmt.Println()
		os.Exit(1)
	}

	if cmd.Flag("help").Value.String() == "true" {
		cmd.Help()
		fmt.Println()
		os.Exit(0)
	}

	if len(customizedData.Array()) > 0 {
		for key_name := range customizedData.Map() {
			if cmd.Flag(key_name).Value.String() != cmd.Flag(key_name).DefValue {
				customizedMap[key_name] = cmd.Flag(key_name).Value
			}
		}
	}

	return customizedMap
}


调用：

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
import (
	"fmt"

	"github.com/spf13/cobra"
	"nuke.csf.nokia.net/nuke-cli/utils"
)

var Region string
var requiredFlags []string

var DemoClusterCreate = &cobra.Command{
	Use:                "demo [string to print]",
	Short:              "demo anything to the screen",
	Long:               `demo is for printing anything back to the screen. For many years people have printed back to the screen.`,
	RunE:               createDemoCluster,
	DisableFlagParsing: true,
}

func createDemoCluster(cmd *cobra.Command, args []string) error {

	var customizedMap = utils.ParseCutomizedFlags("/api/v1/demo/custom-api", requiredFlags, cmd, args)

	for va := range customizedMap {
		fmt.Println(va)
		fmt.Println(customizedMap[va])
	}

	fmt.Println(Region)
	return nil
}

func init() {
	// DemoClusterCreate.Flags().StringVarP(&demoVersion, "version", "v", "1.14", "Venilla version, current can be set as 1.10 , 1.11 1.12")
	DemoClusterCreate.Flags().StringVarP(&Region, "region", "r", "zone-a", "(required) AWS region")
	// DemoClusterCreate.MarkFlagRequired("region")
	//特殊处理required的参数
	requiredFlags = append(requiredFlags, "region") 

}

最后执行的结果：

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
skeetw@N-20HEPF12NB62 MINGW64 /d/CSF_NCM/CSF-NUKE-CLI (1.0)
$ go run nuke.go create cluster demo --cert xxxx --csf-grocery=false
required flag(s) "region" not set
demo is for printing anything back to the screen. For many years people have printed back to the screen.

Usage:
  nuke create cluster demo [string to print] [flags]

Flags:
  -h, --help            help for demo
  -r, --region string   (required) AWS region (default "zone-a")

Global Flags:
      --datacenter-server-api string   data center server api (default "http://127.0.0.1:5678")
      --log string                     nuke log level [debug, info] (default "info")
  -o, --output string                  result output format. [json, yaml, file] (default "json")
      --output-folder string           when output type if file, all result files will be save to this folder. (default "./output/")
      --wrap-processbar                if need to wrap process bar when running a long task (default true)

exit status 1

skeetw@N-20HEPF12NB62 MINGW64 /d/CSF_NCM/CSF-NUKE-CLI (1.0)
$ go run nuke.go create cluster demo --cert xxxx --csf-grocery=false --region=989
csf-grocery
false
cert
xxxx
989

skeetw@N-20HEPF12NB62 MINGW64 /d/CSF_NCM/CSF-NUKE-CLI (1.0)
$ go run nuke.go create cluster demo --cert xxxx --region=989
cert
xxxx
989

skeetw@N-20HEPF12NB62 MINGW64 /d/CSF_NCM/CSF-NUKE-CLI (1.0)
$ go run nuke.go create cluster demo --region=989
989

skeetw@N-20HEPF12NB62 MINGW64 /d/CSF_NCM/CSF-NUKE-CLI (1.0)
$ go run nuke.go create cluster demo
demo is for printing anything back to the screen. For many years people have printed back to the screen.

Usage:
  nuke create cluster demo [string to print] [flags]

Flags:
  -h, --help            help for demo
  -r, --region string   (required) AWS region (default "zone-a")

Global Flags:
      --datacenter-server-api string   data center server api (default "http://127.0.0.1:5678")
      --log string                     nuke log level [debug, info] (default "info")
  -o, --output string                  result output format. [json, yaml, file] (default "json")
      --output-folder string           when output type if file, all result files will be save to this folder. (default "./output/")
      --wrap-processbar                if need to wrap process bar when running a long task (default true)

Powered by Hexo and Hexo-theme-hiker

Copyright © 2013 - 2024 Tiaobug All Rights Reserved.

本站总访问量    次