使用 shuf 来打乱一个文件中的行或是选择文件中一个随机的行。

当内容中出现特殊字符比如 “-”的时候，拼音分词写入失败

Elasticsearch | 作者 slairmy | 发布于2022年11月29日 | 阅读数：2029
分享到：QQ空间新浪微博微信QQ好友印象笔记有道云笔记
ES 版本 7.7.1
pinying 插件版本 7.7.1
 
startOffset must be non-negative, and endOffset must be >= startOffset, and offsets must not go backwards startOffset=0,endOffset=2,lastStartOffset=7 for field 'user_name_ik_pinyin'
 
mapping
 
"mappings" : {

      "properties" : {

        "avatar" : {

          "type" : "keyword",

          "index" : false

        },

        "department_id" : {

          "type" : "keyword"

        },

        "department_name" : {

          "type" : "text",

          "analyzer" : "ik_max_word"

        },

        "id" : {

          "type" : "keyword"

        },

        "job_number" : {

          "type" : "keyword"

        },

        "job_position" : {

          "type" : "text",

          "analyzer" : "ik_max_word"

        },

        "user_id" : {

          "type" : "keyword"

        },

        "user_name_ik_pinyin" : {

          "type" : "text",

          "analyzer" : "ik_max_word_pinyin"

        },

        "user_name_standard" : {

          "type" : "text",

          "analyzer" : "standard"

        }

      }

    }

 
当我执行分词
GET test_name/_analyze
{
  "analyzer": "ik_max_word_pinyin",
  "text": "kk-lin-test"
}
 
下面是分词结果
 
{

  "tokens" : [

    {

      "token" : "k",

      "start_offset" : 0,

      "end_offset" : 11,

      "type" : "LETTER",

      "position" : 0

    },

    {

      "token" : "lin",

      "start_offset" : 0,

      "end_offset" : 11,

      "type" : "LETTER",

      "position" : 2

    },

    {

      "token" : "te",

      "start_offset" : 0,

      "end_offset" : 11,

      "type" : "LETTER",

      "position" : 3

    },

    {

      "token" : "s",

      "start_offset" : 0,

      "end_offset" : 11,

      "type" : "LETTER",

      "position" : 4

    },

    {

      "token" : "t",

      "start_offset" : 0,

      "end_offset" : 11,

      "type" : "LETTER",

      "position" : 5

    },

    {

      "token" : "kk",

      "start_offset" : 0,

      "end_offset" : 2,

      "type" : "LETTER",

      "position" : 6

    },

    {

      "token" : "lin",

      "start_offset" : 3,

      "end_offset" : 6,

      "type" : "LETTER",

      "position" : 7

    },

    {

      "token" : "test",

      "start_offset" : 7,

      "end_offset" : 11,

      "type" : "LETTER",

      "position" : 8

    },

    {

      "token" : "kklintest",

      "start_offset" : 7,

      "end_offset" : 11,

      "type" : "LETTER",

      "position" : 8

    },

    {

      "token" : "k",

      "start_offset" : 0,

      "end_offset" : 2,

      "type" : "ENGLISH",

      "position" : 9

    },

    {

      "token" : "kk",

      "start_offset" : 0,

      "end_offset" : 2,

      "type" : "ENGLISH",

      "position" : 9

    },

    {

      "token" : "lin",

      "start_offset" : 3,

      "end_offset" : 6,

      "type" : "ENGLISH",

      "position" : 10

    },

    {

      "token" : "te",

      "start_offset" : 7,

      "end_offset" : 11,

      "type" : "ENGLISH",

      "position" : 11

    },

    {

      "token" : "s",

      "start_offset" : 7,

      "end_offset" : 11,

      "type" : "ENGLISH",

      "position" : 12

    },

    {

      "token" : "t",

      "start_offset" : 7,

      "end_offset" : 11,

      "type" : "ENGLISH",

      "position" : 13

    },

    {

      "token" : "test",

      "start_offset" : 7,

      "end_offset" : 11,

      "type" : "ENGLISH",

      "position" : 13

    }

  ]

}

好像是因为kk这个重复了，出现了相同的 start_offset 和 end_offset 但是 position 不同导致的，查了一下资料说要注释掉部分代码重新编译（6.8 版本的一个方案）
 
请问一下有其他的方式可以在 ES层面去解决这个问题嘛？
 
2022-11-29 添加评论

没有找到相关结果

    已邀请:

    slairmy

    赞同来自:

    排查了好久，发现是settings的一个配置解决的，原settings
     
    {
    
    "mappings": {
    
        "properties": {
    
            "id": {
    
                "type": "keyword"
    
            },
    
            "user_id": {
    
                "type": "keyword"
    
            },
    
            "user_name_ik_pinyin": {
    
                "type": "text",
    
                "analyzer": "ik_max_word_pinyin",
    
                "search_analyzer": "ik_max_word_pinyin"
    
            },
    
            "user_name_standard": {
    
                "type": "text",
    
                "analyzer": "standard",
    
                "search_analyzer": "standard"
    
            },
    
            "department_id": {
    
                "type": "keyword"
    
            },
    
            "department_name": {
    
                "type": "text",
    
                "analyzer": "ik_max_word",
    
                "search_analyzer": "ik_max_word"
    
            },
    
            "job_number": {
    
                "type": "keyword"
    
            },
    
            "job_position": {
    
                "type": "text",
    
                "analyzer": "ik_max_word",
    
                "search_analyzer": "ik_max_word"
    
            },
    
            "avatar": {
    
                "type": "keyword",
    
                "index": false
    
            }
    
        }
    
    },
    
    "settings": {
    
        "number_of_shards": 1,
    
        "number_of_replicas": 0,
    
        "analysis": {
    
            "analyzer": {
    
                "ik_smart_pinyin": {
    
                    "type": "custom",
    
                    "tokenizer": "ik_smart",
    
                    "filter": [
    
                        "my_pinyin",
    
                        "word_delimiter"
    
                    ]
    
                },
    
                "ik_max_word_pinyin": {
    
                    "type": "custom",
    
                    "tokenizer": "ik_max_word",
    
                    "filter": [
    
                        "my_pinyin",
    
                        "word_delimiter"
    
                    ]
    
                },
    
                "standard_pinyin": {
    
                    "type": "custom",
    
                    "tokenizer": "standard",
    
                    "filter": [
    
                        "my_pinyin",
    
                        "word_delimiter"
    
                    ]
    
                }
    
            },
    
            "filter": {
    
                "my_pinyin": {
    
                    "type": "pinyin",
    
                    "keep_full_pinyin": true,
    
                    "remove_duplicated_term": true,
    
                    "keep_separate_first_letter": true,
    
                    "limit_first_letter_length": 16,
    
                    "keep_original": true,
    
                    "lowercase": true
    
                }   
    
            }
    
        }
    
    }
    
}

    是因为 "keep_original": true 这个配置项引起的，但是具体为什么不知道，去掉这个配置项或者改成false就好了！
    2022-12-02 0 0
    分享
    为什么被折叠? 0 个回复被折叠

    要回复问题请先登录或注册

    Copyright © 2024 · CC BY-NC-SA 3.0

    本站服务器及带宽由 提供赞助