2
回答
关于elasticsearch支持搜索中文和拼音的索引如何构建构建
终于搞明白,存储TCO原来是这样算的>>>   

环境一台 centos6.5 192.168.1.82 es-node2 一台 centos6.6 192。168.1.136 es-node1

{
"status": 200,
"name": "es-node1",
"cluster_name": "elasticsearch-cluster-centos",
"version": {
"number": "1.4.4",
"build_hash": "c88f77ffc81301dfa9dfd81ca2232f09588bd512",
"build_timestamp": "2015-02-19T13:05:36Z",
"build_snapshot": false,
"lucene_version": "4.10.3"
},
"tagline": "You Know, for Search"
}



问题:关键字"苍穹的法芙娜:逝者的权利” 拼音首字母 CQDFFNSZDQL

条件: 逝者的权利,DQL 

目的:关键词命中

以目前的索引结构

[body] => Array
(
[name] => 非诚勿扰

[letter] => FCWR

搜索中文测试代码:

$json = '{
   "query" : {
       "match" : {
           "name" : "'. $q. '"
       }
   }
}';
$params['body']  = $json;
$params['size'] = 100;

$query = $es->search($params);
测试结果:正确返回

搜索拼音测试代码:

$json = '{
   "query" : {
       "match" : {
           "letter" : "'. $q. '"
       }
   }
}';
$params['body']  = $json;
$params['size'] = 100;

$query = $es->search($params);

测试结果为空

于是我有了第二种方案:索引结构
[body] => Array
(
[name] => 苍穹的法芙娜:逝者的权利
[alias] => 蒼穹のファフナー Right of Left/苍穹之战神:逝者的权利/苍穹之巨龙:逝者的权利
[letter] => CQDFFNSZDQL
[kewords] => Array
(
[0] => C
[1] => CQ
[2] => CQD
[3] => CQDF
[4] => CQDFF
[5] => CQDFFN
[6] => CQDFFNS
[7] => CQDFFNSZ
[8] => CQDFFNSZD
[9] => CQDFFNSZDQ
[10] => CQDFFNSZDQL
[11] => Q
[12] => QD
[13] => QDF
[14] => QDFF
[15] => QDFFN
[16] => QDFFNS
[17] => QDFFNSZ
[18] => QDFFNSZD
[19] => QDFFNSZDQ
[20] => QDFFNSZDQL
[21] => D
[22] => DF
[23] => DFF
[24] => DFFN
[25] => DFFNS
[26] => DFFNSZ
[27] => DFFNSZD
[28] => DFFNSZDQ
[29] => DFFNSZDQL
[30] => F
[31] => FF
[32] => FFN
[33] => FFNS
[34] => FFNSZ
[35] => FFNSZD
[36] => FFNSZDQ
[37] => FFNSZDQL
[38] => F
[39] => FN
[40] => FNS
[41] => FNSZ
[42] => FNSZD
[43] => FNSZDQ
[44] => FNSZDQL
[45] => N
[46] => NS
[47] => NSZ
[48] => NSZD
[49] => NSZDQ
[50] => NSZDQL
[51] => S
[52] => SZ
[53] => SZD
[54] => SZDQ
[55] => SZDQL
[56] => Z
[57] => ZD
[58] => ZDQ
[59] => ZDQL
[60] => D
[61] => DQ
[62] => DQL
[63] => Q
[64] => QL
[65] => L
[66] => 苍
[67] => 苍穹
[68] => 苍穹的
[69] => 苍穹的法
[70] => 苍穹的法芙
[71] => 苍穹的法芙娜
[72] => 苍穹的法芙娜逝
[73] => 苍穹的法芙娜逝者
[74] => 苍穹的法芙娜逝者的
[75] => 苍穹的法芙娜逝者的权
[76] => 苍穹的法芙娜逝者的权利
[77] => 穹
[78] => 穹的
[79] => 穹的法
[80] => 穹的法芙
[81] => 穹的法芙娜
[82] => 穹的法芙娜逝
[83] => 穹的法芙娜逝者
[84] => 穹的法芙娜逝者的
[85] => 穹的法芙娜逝者的权
[86] => 穹的法芙娜逝者的权利
[87] => 的
[88] => 的法
[89] => 的法芙
[90] => 的法芙娜
[91] => 的法芙娜逝
[92] => 的法芙娜逝者
[93] => 的法芙娜逝者的
[94] => 的法芙娜逝者的权
[95] => 的法芙娜逝者的权利
[96] => 法
[97] => 法芙
[98] => 法芙娜
[99] => 法芙娜逝
[100] => 法芙娜逝者
[101] => 法芙娜逝者的
[102] => 法芙娜逝者的权
[103] => 法芙娜逝者的权利
[104] => 芙
[105] => 芙娜
[106] => 芙娜逝
[107] => 芙娜逝者
[108] => 芙娜逝者的
[109] => 芙娜逝者的权
[110] => 芙娜逝者的权利
[111] => 娜
[112] => 娜逝
[113] => 娜逝者
[114] => 娜逝者的
[115] => 娜逝者的权
[116] => 娜逝者的权利
[117] => 逝
[118] => 逝者
[119] => 逝者的
[120] => 逝者的权
[121] => 逝者的权利
[122] => 者
[123] => 者的
[124] => 者的权
[125] => 者的权利
[126] => 的
[127] => 的权
[128] => 的权利
[129] => 权
[130] => 权利
[131] => 利
)

测试代码:

$json = '{
   "query" : {
       "match" : {
           "kewords" : "'. $q. '"
       }
   }
}';
$params['body']  = $json;
$params['size'] = 100;

$query = $es->search($params);

结果全部命中。

缺点,每条信息需要耗费时间拼凑关键词


有没有其他的好办法

有人会使用

Pinyin Analysis for ElasticSearch插件的么?

请联系我:iposou@163.com,526161488



举报
歪歪1985
发帖于3年前 2回/4K+阅

以下是问题补充:

顶部