发布到Atlassian Confluence api时出现意外的grunt-http错误 [英] Unexpected grunt-http error when posting to Atlassian Confluence api

问题描述

试图在Atlassian wiki上创建一个wiki页面。我之前使用的是python脚本，这段代码没有任何问题：

  data = json.dumps（{type： data：testData，title：postTitle，space：{key：EB}，body：{storage：{value：content，representation ：storage}}}）
r = requests.post（https://estech.atlassian.net/wiki/rest/api/content/，data = data，headers = headers，auth =（confluenceLogin ['username']，confluenceLogin ['password']））
  

现在我试着使用以下grunt任务配置：

  http：{
 atlassianwiki：{
 options：{
 uri：atlassianURL +/ wiki / rest / api / content /，
标题：{Content-Type：application / json}，
 auth：{
user ：confluencelogin，
pass：confluencepass 
}，
方法：POST，
 body：JSON.stringify（wi kijson）
} 
} 
} 
  

：

  wikijson = {
type：page，
data：testData ，
title：testtitle，
space：{key：EB}，
body：{
storage：{
value：< p>测试发布< / p>，
representation：storage
} 
} 
} 
  

运行此任务时出现以下错误：

 致命错误：500 {statusCode：500，message：java.io.EOFException：由于输入结束而无法映射到Object的内容} 
  

经过一番Google-fu，我发现有些人声称他们通过在卷曲中添加--post302来解决这个问题命令行。但我真的不知道或理解这是如何适用于此。

谢谢，我希望这是有道理的。

解决方案

我在使用合流REST API时遇到了问题，在我的情况下，问题出现在内容类型头文件中，但您似乎已经拥有了它。

我没有'尝试创建新页面，但更新现有的一个
Confluence API对我来说似乎有点神奇，所以我只是在这里离开这里我开始工作之前必须做的所有步骤，也许其中一个会帮助你。 / p>

function composeRequest（method）{
var auth = new Buffer（user +'： '+ pass）.toString（'base64'）;
var request = {
host：'confluence.myserver.com'，
port：443，
contentType：application / json; charset = utf-8，
'path'：path，
method：method || GET，
标题：{
'授权'：'基本'+验证，
'内容类型'：'application / json'
}，
rejectUnauthorized：false，
requestCert：true，
agent：false
};


退货请求;
}


而且它出现了页面更新请求JSON必须包含

• pageId（即使它在路径中，您需要重复它）


• type


• title


• 版本（这很奇怪，但你应该设置它，0或1，我不记得）

而且当你的数据被填充时，你应该将它转换为字符串，并填充请求中的content-type字段！

  data = JSON.stringify（data）; 
 request.headers ['Content-Length'] = data.length; 
 https.request（request，respondHandler）
  

Attempting to create a wiki page on an Atlassian wiki. I previously was using a python script and this code worked no problem:

data = json.dumps({"type":"page", "data":"testData", "title":postTitle,"space":{"key":"EB"}, "body":{"storage":{"value": content,"representation":"storage"}}})
r = requests.post("https://estech.atlassian.net/wiki/rest/api/content/", data=data, headers=headers, auth=(confluenceLogin['username'], confluenceLogin['password']))

Now I'm trying to use the following grunt task configuration:

    http: {
        atlassianwiki: {
            options: {
                uri: atlassianURL + "/wiki/rest/api/content/",
                headers: {"Content-Type": "application/json"},
                auth: {
                    "user": confluencelogin,
                    "pass": confluencepass
                },
                method:"POST",
                body: JSON.stringify(wikijson)
            }
        }
    }

with wikijson looking like:

wikijson = {
            "type": "page",
            "data": "testData",
            "title": "testtitle",
            "space": {key:"EB"},
            "body": {
                "storage": {
                    "value": "<p>testing posting</p>",
                    "representation": "storage"
                }
            }
        }

And I get the following error when this task runs:

Fatal error: 500 {"statusCode":500, "message":"java.io.EOFException: No content to map to Object due to end of input"}

Upon a bit of google-fu, I found that some people claim they fixed this by adding "--post302" to their curl command line. But I don't really know or understand how that applies here.

Thanks and I hope this makes sense.

解决方案

i was fighting with confluence REST API and in my case the problem was in content-type header, but you seem to have it already.
I didn't try to create new page but to update existing one Confluence API seemed a little bit magic to me, so i just leave here all steps i had to make before it started working, maybe one of them will help you.

function composeRequest(method) {
  var auth = new Buffer(user + ':' + pass).toString('base64');
  var request = {
  host: 'confluence.myserver.com',
  port: 443,
  contentType: "application/json; charset=utf-8",
  'path': path,
  method: method || "GET",
  headers: {
    'Authorization': 'Basic ' + auth,
    'Content-Type': 'application/json'
  },
  rejectUnauthorized: false,
  requestCert: true,
  agent: false
};


  return request;
}

And it appeared that page update request JSON MUST contain

• pageId (even it is inside path, you need to repeat it)
• type
• title
• version (it's weird, but you should set it. 0 or 1, i don't remember)

And when your data if filled, you should convert it to string and fill content-type field in your request!

data = JSON.stringify(data);
request.headers['Content-Length'] = data.length;
https.request(request, respondHandler)

这篇关于发布到Atlassian Confluence api时出现意外的grunt-http错误的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！