如何在单个 POST 请求中向 Google Cloud 自然语言 API 发送多个文本字符串

Question

这是我的 Python 代码

def sentiment_local_file(text):
  """Detects sentiment in the local document"""
  language_client = language.Client()
  if isinstance(text, six.binary_type):
      text = text.decode('utf-8')
  with open("abhi.txt",'r') as fr:
      data = json.loads(fr.read())
  print ([data['document']['content']])
  document = language_client.document_from_text(data['document']['content'])
  result = document.annotate_text(include_sentiment=True,
                                         include_syntax=False,
                                         include_entities=False)

我正在尝试在单个 post 请求中发送字符串列表进行分析，但它给出了错误.这是我正在阅读的文本文件。在上面的代码文本中引用文件名，代码示例是一个函数

{
 "document":{
 "type":"PLAIN_TEXT",
 "language": "EN",
 "content":[
    "pretending to be very busy"
  ,
    "being totally unconcerned"
  ,
    "a very superior attitude"
  ,
    "calm, dignified and affectionate disposition" 
]},"encodingType":"UTF8"}

我阅读了文档和许多示例仍然无法弄清楚。

Answer 1

据我所知，没有办法发送要分析的字符串列表。返回的句子是一个数组，因为 GNL API 会破坏每个句子并对其进行分析。

假设您发送以下请求：

{
 "document": {
  "type": "PLAIN_TEXT",
  "content": "Terrible, I did not like the last updated."
 }
}

响应可能是：

{
    "language": "en",
    "sentences": [
        {
            "text": {
                "content": "Terrible, I did not like the last updated.",
                "beginOffset": -1
            },
            "sentiment": {
                "magnitude": 0.9,
                "score": -0.9
            }
        }
    ]
}

上面的响应，有一个名为"句子"的数组，但只有一个元素。发生这种情况是因为我们发送去分析的文本只有一句话。因此，下面是另一个示例：

请求：

{
 "document": {
  "type": "PLAIN_TEXT",
  "content": "Terrible, I did not like the last updated. Also, I would like to have access to old version"
 }
}

可能的响应将是：

{
    "language": "en",
    "sentences": [
        {
            "text": {
                "content": "Terrible, I did not like the last updated.",
                "beginOffset": -1
            },
            "sentiment": {
                "magnitude": 0.9,
                "score": -0.9
            }
        },
        {
            "text": {
                "content": "Also, I would like to have access to old version",
                "beginOffset": -1
            },
            "sentiment": {
                "magnitude": 0,
                "score": 0
            }
        }
    ]
}

在这种情况下，sentence数组有两个元素。发生这种情况是因为发送的文本有两个句子(用句点"."分隔(。

希望对您有所帮助。