Curl转R代码后无法获取文件名列表，请求排查原因

我有一段可正常运行的Curl POST请求，调用https://historicdata.website.com/api/DownloadListOfFiles接口能正确获取文件名列表，但通过ChatGPT转换为使用httr和jsonlite包的R代码后，请求返回成功却得到错误页面。转换后的R代码存在多处参数不一致问题，现排查原因并给出修正方案。

原Curl代码 #

curl -X POST https://historicdata.website.com/api/DownloadListOfFiles
-H 'content-type: application/json' 
-H 'ssoid: INPUT_SESSION_TOKEN_HERE' 
-d '{
    "music":"Rock",
    "plan":"Basic Plan",
    "fromDay" : 1,
    "fromMonth": 3,
    "fromYear" : 2020,
    "toDay": 3,
    "toMonth" : 3,
    "toYear": 2020,
    "eventId": null,
    "eventName": null,
    "marketTypesCollection": [ "GUITAR" ],
    "countriesCollection" : [ "GB", "IE" ],
    "fileTypeCollection" : [ "M" ]
}'

转换后的R代码 #

library(httr)
library(jsonlite)

session_token <- "INPUT_SESSION_TOKEN_HERE"

# Define the URL and headers
url <- "https://historicdata.website.com/api/DownloadListOfFiles"
headers <- c(
  "content-type" = "application/json",
  "ssoid" = session_token)

# Define the request body
body <- list(
  music = "Rock",
  plan = "Basic Plan",
  fromDay = 1,
  fromMonth = 10,
  fromYear = 2023,
  toDay = 2,
  toMonth = 10,
  toYear = 2023,
  eventId = NULL,
  eventName = NULL,
  marketTypesCollection = "GUITAR",
  fileTypeCollection = "M"
)

# Perform the POST request
response <- httr::POST(url, httr::add_headers(.headers=headers), body = jsonlite::toJSON(body))

# Check response
if (httr::http_error(response)) {
  stop("HTTP error: ", httr::http_status(response)$reason)
} else {
  cat("Request successful.")
}

# Extract and print the content of the response
content <- httr::content(response, as = "text")
cat(content)

错误返回结果 #

Request successful.

<!DOCTYPE html>
<html>
<head>
<meta name="viewport" content="width=device-width" />
<title>ngErrorRedirect</title>
</head>
<body>
<div>
Error
</div>
<script defer src="https://static.cloud.com/beacon.min.js/v8c1696crossorigin="anonymous"></script>
</body>
</html>

问题原因分析 #

• 日期参数不匹配：原Curl请求的日期范围是2020年3月1日至3月3日，转换后的R代码错误使用了2023年10月1日至10月2日，参数不符可能触发后端的日期验证逻辑，导致返回错误。
• 数组参数处理错误：原Curl中的marketTypesCollection、countriesCollection、fileTypeCollection均为数组类型，但转换后的R代码将前两者中的两个设为单个字符串，还缺失了countriesCollection参数，后端无法正确解析参数结构，引发错误。
• JSON序列化配置问题：jsonlite::toJSON()默认会将单元素数组自动简化为单个值，导致后端接收的参数类型与接口要求不符，需要设置auto_unbox = FALSE来保留数组结构。

修正后的R代码 #

library(httr)
library(jsonlite)

session_token <- "INPUT_SESSION_TOKEN_HERE"

# 定义请求URL和头部
url <- "https://historicdata.website.com/api/DownloadListOfFiles"
headers <- c(
  "content-type" = "application/json",
  "ssoid" = session_token
)

# 定义请求体，严格匹配原Curl的参数
body <- list(
  music = "Rock",
  plan = "Basic Plan",
  fromDay = 1,
  fromMonth = 3,
  fromYear = 2020,
  toDay = 3,
  toMonth = 3,
  toYear = 2020,
  eventId = NULL,
  eventName = NULL,
  marketTypesCollection = list("GUITAR"),  # 保留数组结构
  countriesCollection = list("GB", "IE"),  # 补全缺失的数组参数
  fileTypeCollection = list("M")           # 保留数组结构
)

# 发起POST请求，设置auto_unbox=FALSE确保数组不被简化
response <- httr::POST(
  url = url,
  httr::add_headers(.headers = headers),
  body = jsonlite::toJSON(body, auto_unbox = FALSE)
)

# 检查响应状态
if (httr::http_error(response)) {
  stop("HTTP错误: ", httr::http_status(response)$reason)
} else {
  cat("请求成功。\n")
}

# 提取并打印响应内容
content <- httr::content(response, as = "text")
cat(content)

内容的提问来源于stack exchange，提问作者user2165379