-
Notifications
You must be signed in to change notification settings - Fork 18
paracraft download page
原有的过程
-
连接 http://www.paracraft.cn/?lang=zh ，点击下载
-
转到 http://www.paracraft.cn/download?lang=zh ，再点击三个不同版本的下载按钮
-
都会转到同一个地址 http://www.nplproject.com/zh/download/download-paracraft ，再选择相应的下载
新的过程（理想）
从 http://www.paracraft.cn/?lang=zh 直接到 http://keepwork.com/dreamanddead/downloads/index
dreamanddead为自己的个人帐户，实际上线的时候，要使用存放官方页面的一个帐户，帐户名可能为 paracraft?official?
ps: 在keepwork.com每注册一个新用户，都会在gitlab.keepwork.com生成一个同样帐号密码的帐户，存储了所有keepwork的页面数据
实际的下载页面是作为个人网站的一个single page，每个page都是一个markdown文件，都存储于gitlab后端。想要自动化的更新页面，就需要用代码进行md文件更新的操作。
实际上，每一个用户编辑页面并保存，都是git的一个新的提交，提交的增量为页面的改动。
在gitlab访问刚才的下载页： http://gitlab.keepwork.com/gitlab_rls_dreamanddead/keepworkdownloads/blob/master/dreamanddead/downloads/index.md
好在gitlab提供了api来对数据进行操作。
每个帐户都有一个private token，在进行api操作时必须在http header中带上token. ref: https://docs.gitlab.com/ee/api/README.html#private-tokens
ref: https://docs.gitlab.com/ee/api/commits.html#create-a-commit-with-multiple-files-and-actions
用api来创建提交，是我们最终的目的。根据api的要求，创建http post
POST /projects/:id/repository/commits
这里的id标识了是哪一个project。每一个project都是一个git仓库，每一个页面都存在于一个git repo中。所以project+path可以定位每一个页面。
post数据根据api要求，update操作可以满足我们的需求
{
"branch": "master",
"commit_message": "update download page from CI",
"actions": [
{
"action": "update",
"file_path": "dreamanddead/downloads/index.md",
"content": "page content here"
}
]
}
content是我们每次更新的页面内容
上面的逻辑非常清楚，剩下的问题在于，如何得到 project id
ps: 虽然我们可以用新申请的keepwork帐户登陆gitlab.keepwork.com，但是实际在gitlab后台的用户名为 gitlab_rls_${keepwork_name}
从用户名开始，得到用户id，ref：https://docs.gitlab.com/ee/api/users.html#for-admins
以我的用户名dreamanddead为例
GET /users?username=gitlab_rls_dreamanddead
返回
[
{
"name": "gitlab_rls_dreamanddead",
"username": "gitlab_rls_dreamanddead",
"id": 1630,
"state": "active",
"avatar_url": "http://www.gravatar.com/avatar/19969b73d47d4ac197141bcc2e965193?s=80&d=identicon",
"web_url": "http://git.keepwork.com/gitlab_rls_dreamanddead"
}
]
用户id为1630
进一步，取得用户的项目信息，ref：https://docs.gitlab.com/ee/api/projects.html#list-a-user-39-s-projects
GET /users/1630/projects
返回
[
{
"id": 3432,
"description": null,
"default_branch": "master",
"tag_list": [],
"archived": false,
"visibility": "public",
"ssh_url_to_repo": "git@git.keepwork.com:gitlab_rls_dreamanddead/keepworkdownloads.git",
"http_url_to_repo": "http://git.keepwork.com/gitlab_rls_dreamanddead/keepworkdownloads.git",
"web_url": "http://git.keepwork.com/gitlab_rls_dreamanddead/keepworkdownloads",
"owner": {
"name": "gitlab_rls_dreamanddead",
"username": "gitlab_rls_dreamanddead",
"id": 1630,
"state": "active",
"avatar_url": "http://www.gravatar.com/avatar/19969b73d47d4ac197141bcc2e965193?s=80&d=identicon",
"web_url": "http://git.keepwork.com/gitlab_rls_dreamanddead"
},
"name": "keepworkdownloads",
"name_with_namespace": "gitlab_rls_dreamanddead / keepworkdownloads",
"path": "keepworkdownloads",
"path_with_namespace": "gitlab_rls_dreamanddead/keepworkdownloads",
"container_registry_enabled": true,
"issues_enabled": true,
"merge_requests_enabled": true,
"wiki_enabled": true,
"jobs_enabled": true,
"snippets_enabled": true,
"created_at": "2017-08-31T07:00:37.569Z",
"last_activity_at": "2017-09-04T06:52:00.315Z",
"shared_runners_enabled": true,
"lfs_enabled": true,
"creator_id": 1630,
"namespace": {
"id": 1874,
"name": "gitlab_rls_dreamanddead",
"path": "gitlab_rls_dreamanddead",
"kind": "user",
"full_path": "gitlab_rls_dreamanddead",
"parent_id": null
},
"import_status": "none",
"avatar_url": null,
"star_count": 0,
"forks_count": 0,
"open_issues_count": 0,
"public_jobs": true,
"ci_config_path": null,
"shared_with_groups": [],
"only_allow_merge_if_pipeline_succeeds": false,
"request_access_enabled": true,
"only_allow_merge_if_all_discussions_are_resolved": false,
"printing_merge_request_link_enabled": true,
"permissions": {
"project_access": {
"access_level": 40,
"notification_level": 3
},
"group_access": null
}
},
....
....
....
....
....
....
....
]
project id即为 3432
脚本地址： https://github.com/tatfook/KeepMonitor/tree/master/paracraft/download-page
脚本的进行逻辑很简单
- 取得最新的版本号
- 用版本号与相关下载地址，与页面模板，生成下载页内容
- 更新下载页
其中用模板生成下载页的内容，委托给python3脚本来写，bash在这方面太弱