python - 如何读取搜索分析器的配置？

Question

在我看来，Elasticsearch 有两个分析器，一个用于索引，另一个用于查询。

我成功地添加了索引分析器来标记文档，但不知道如何将分析器添加到搜索查询中。

这是我用 Python 编写的初步文件

from elasticsearch import Elasticsearch
es_conn = Elasticsearch(config.ES_HOSTS)


def analyze_query(text, es_conn, index_name):
    '''
        analyzes any text with my_analyzer defined in es_settings.json
        input:
            - text: a query text
            - es_conn: elasticsearch connection
            - index_name: name of index
        output:
            - a list of tokens
    '''

    tokens = es_conn.indices.analyze(
        index = index_name,
        body = {"text": text},
        # how to point the analyzer to the json file???
        analyzer = 'my_analyzer')["tokens"]
    return [token_row["token"].encode('utf-8') for token_row in tokens]

出现错误

elasticsearch.exceptions.RequestError: TransportError(400, u'illegal_argument_exception', u'[1aVYakX][127.0.0.1:9300][indices:admin/analyze[s]]')

问题出现在 analyzer = 'my_analyzer' 部分，我不确定如何使其指向定义 Elasticsearch 设置的 json 文件。

添加

json 文件

{
  "settings": {
    "number_of_shards": 1,
    "analysis": {
      "analyzer": {
        "my_analyzer": {
          "filter": [
            "lowercase",
            "stop"
          ],
          "char_filter": [
            "html_strip"
          ],
          "type": "custom",
          "tokenizer": "uax_url_email"
        }
      }
    }
  },
  "mappings": {
    "doc": {
      "properties": {
        "filename": {
          "type": "keyword",
          "index": false,
          "doc_values": false
        },
        "path": {
          "type": "keyword",
          "index": false,
          "doc_values": false
        },
        "text": {
          "type": "text",
          "analyzer": "my_analyzer",
          "search_analyzer": "my_analyzer"
        }
      }
    }
  }
}

Answer 1

analyze 函数不将分析器作为参数。更多详细信息请参见:http://elasticsearch-py.readthedocs.io/en/master/api.html#elasticsearch.client.IndicesClient.analyze .

相反，您应该传递 body 参数。将analyze_query 函数替换为下面的函数。

def analyze_query(text, es_conn, index_name, settings_path = config.SETTINGS_PATH):

    with open(settings_path) as json_data:
        settings = json.load(json_data)["settings"]["analysis"]

        filter_settings = settings["filter"]
        analyzer_settings = settings["analyzer"]["my_analyzer"]

        body = {}
        body["text"] = text
        body["tokenizer"] = analyzer_settings["tokenizer"]
        if "char_filter" in analyzer_settings:
          body["char_filter"] = analyzer_settings["char_filter"]

        body["filter"] = [ filter_settings[f] if f in filter_settings else f \
            for f in analyzer_settings["filter"]]
        tokens = es_conn.indices.analyze(
            index = index_name,
            body = body)["tokens"]

    return [token_row["token"].encode('utf-8') for token_row in tokens]

祝你作业顺利。