从tibble转换为JSON时遇到格式不符问题

你当前的代码生成的JSON中，fields是包含所有行的数组，而目标格式要求每个records元素对应一个独立的fields对象，同时需要额外添加顶层的fieldKey字段。问题根源在于rowwise() + nest()的嵌套方式导致结构不符合预期，且未补充所需的顶层字段。

修正代码 #

直接将每一行数据转换为含fields的独立列表，再组合成目标结构，同时添加fieldKey：

library(tibble)
library(jsonlite)
library(dplyr)
library(purrr)

# 原始数据框
original_tibble <- tibble::tibble(
  matter = c("「Organization Module - Organization of the address book panel", 
             "「Member」 module - set the department"),
  `Problem description` = c("It is essentially the same as the above requirements", 
                            "After selecting a person, you can adjust multiple departments to which he belongs\nAfter selecting a department, you can also add members to the current department"),
  classification = c("Product demand", "product demand"),
  `review date` = c("2019-10-30T00:00:00.000Z", "2019-10-29T16:00:00.000Z")
)

# 可选：调整列名以匹配目标JSON的字段大小写
original_tibble <- original_tibble %>%
  rename(
    "problem description" = `Problem description`,
    "Review date" = `review date`
  )

# 构造records列表：每一行对应一个带fields的对象
records <- original_tibble %>%
  pmap(function(...) {
    list(fields = list(...))
  })

# 转换为JSON，auto_unbox确保单个对象不被转为数组
result_json <- toJSON(
  list(records = records, fieldKey = "name"),
  pretty = TRUE,
  auto_unbox = TRUE
)

# 输出结果
cat(result_json)

代码说明 #

1. 列名调整：如果需要完全匹配目标JSON的字段大小写，使用rename()修改列名（可根据实际需求跳过）。
2. 构造records：通过pmap()遍历每一行，将行内字段打包成fields对应的列表，保证每个records元素是独立对象。
3. JSON转换：添加顶层fieldKey字段，auto_unbox = TRUE避免将单个fields对象转为数组，确保结构符合要求。

输出结果 #

运行后将生成你需要的JSON格式：

{
  "records": [
    {
      "fields": {
        "matter": "「Organization Module - Organization of the address book panel",
        "problem description": "It is essentially the same as the above requirements",
        "classification": "Product demand",
        "Review date": "2019-10-30T00:00:00.000Z"
      }
    },
    {
      "fields": {
        "matter": "「Member」 module - set the department",
        "problem description": "After selecting a person, you can adjust multiple departments to which he belongs\nAfter selecting a department, you can also add members to the current department",
        "classification": "product demand",
        "Review date": "2019-10-29T16:00:00.000Z"
      }
    }
  ],
  "fieldKey": "name"
}

内容的提问来源于stack exchange，提问作者Netmarketr