如何使用 Azure 认知搜索搜索字符串的一部分

Question

我是 Azure 认知搜索的新手，并且已成功配置我的索引以进行自动完成（感谢 this article 使用部分搜索）。

但现在我有另一个用例，我将许多文件存储在带有元数据的 Azure Blob 容器中：

（每个文件的）元数据字段之一称为 partnumbers，其值是一串用逗号分隔的产品 SKU（例如“123456,78901,102938, 09876”）。我建立了索引，以便将此信息存储为 Edm.String，如下所示：

{
  "name": "my-index",
  "fields": [
    {
      "name": "partnumbers",
      "type": "Edm.String",
      "facetable": true,
      "filterable": true,
      "key": false,
      "retrievable": true,
      "searchable": true,
      "sortable": true,
      "analyzer": null,
      "indexAnalyzer": null,
      "searchAnalyzer": null,
      "synonymMaps": [],
      "fields": []
    },
    {
      "name": "metadata_storage_name",
      "type": "Edm.String",
      "facetable": true,
      "filterable": true,
      "key": false,
      "retrievable": false,
      "searchable": true,
      "sortable": true,
      "analyzer": null,
      "indexAnalyzer": null,
      "searchAnalyzer": null,
      "synonymMaps": [],
      "fields": []
    },
    {
      "name": "metadata_storage_content_type",
      "type": "Edm.String",
      "facetable": true,
      "filterable": true,
      "key": false,
      "retrievable": false,
      "searchable": true,
      "sortable": true,
      "analyzer": null,
      "indexAnalyzer": null,
      "searchAnalyzer": null,
      "synonymMaps": [],
      "fields": []
    },
    {
      "name": "metadata_storage_last_modified",
      "type": "Edm.String",
      "facetable": true,
      "filterable": true,
      "key": false,
      "retrievable": false,
      "searchable": true,
      "sortable": true,
      "analyzer": null,
      "indexAnalyzer": null,
      "searchAnalyzer": null,
      "synonymMaps": [],
      "fields": []
    },
    {
      "name": "metadata_storage_path",
      "type": "Edm.String",
      "facetable": true,
      "filterable": true,
      "key": false,
      "retrievable": false,
      "searchable": true,
      "sortable": true,
      "analyzer": null,
      "indexAnalyzer": null,
      "searchAnalyzer": null,
      "synonymMaps": [],
      "fields": []
    },
    {
      "name": "metadata_storage_size",
      "type": "Edm.Int64",
      "facetable": true,
      "filterable": true,
      "retrievable": false,
      "sortable": true,
      "analyzer": null,
      "indexAnalyzer": null,
      "searchAnalyzer": null,
      "synonymMaps": [],
      "fields": []
    },
    {
      "name": "key",
      "type": "Edm.String",
      "facetable": true,
      "filterable": true,
      "key": true,
      "retrievable": true,
      "searchable": true,
      "sortable": true,
      "analyzer": null,
      "indexAnalyzer": null,
      "searchAnalyzer": null,
      "synonymMaps": [],
      "fields": []
    },
    {
      "name": "partialPartnumbers",
      "type": "Edm.String",
      "facetable": false,
      "filterable": false,
      "key": false,
      "retrievable": false,
      "searchable": true,
      "sortable": false,
      "analyzer": null,
      "indexAnalyzer": "prefixCmAnalyzer",
      "searchAnalyzer": "standardCmAnalyzer",
      "synonymMaps": [],
      "fields": []
    },
  ],
  "suggesters": [
    {
      "name": "my-index_suggester",
      "searchMode": "analyzingInfixMatching",
      "sourceFields": [
        "partnumbers"
      ]
    }
  ],
  "scoringProfiles": [
    {
      "name": "exactFirst",
      "functions": [],
      "functionAggregation": null,
      "text": {
        "weights": {
          "partnumbers": 2,
          "partialPartnumbers": 1,
        }
      }
    }
  ],
  "defaultScoringProfile": "exactFirst",
  "corsOptions": null,
  "analyzers": [
    {
      "@odata.type": "#Microsoft.Azure.Search.CustomAnalyzer",
      "name": "standardCmAnalyzer",
      "tokenizer": "standard_v2",
      "tokenFilters": [
        "lowercase",
        "asciifolding"
      ],
      "charFilters": []
    },
    {
      "@odata.type": "#Microsoft.Azure.Search.CustomAnalyzer",
      "name": "prefixCmAnalyzer",
      "tokenizer": "standard_v2",
      "tokenFilters": [
        "lowercase",
        "asciifolding",
        "edgeNGramCmTokenFilter"
      ],
      "charFilters": []
    }
  ],
  "charFilters": [],
  "tokenFilters": [
    {
      "@odata.type": "#Microsoft.Azure.Search.EdgeNGramTokenFilterV2",
      "name": "edgeNGramCmTokenFilter",
      "minGram": 2,
      "maxGram": 20,
      "side": "front"
    }
  ],
  "tokenizers": [],
  "@odata.etag": "\"0x8D8184F367A74XX\""
}

现在我正在努力寻找一种方法（通过特定语法？分析器？分词器？）能够找到所有具有 partnumbers 元数据字段的文件其中包含一个 SKU（以便我可以检索与一种产品相关的所有文档）：我想将 SKU“102938”传递给 Azure 搜索，它将 return 我所有包含此 SKU 的文件它的 partnumbers 元数据字段（可能包括其他 SKU）。

但我很难在 Google 上找到示例，并且文档似乎 - 目前 - 有点超出我的理解范围（我不确定是否正确理解什么是分析器、分词器等以及它们是如何工作的！这是我第一次深入“搜索”世界...)。

所以我真的很感谢社区能在这方面帮助我，我很想阅读让初学者理解一切的文章、教程或任何可以帮助我前进的东西！

提前致谢。

Answer 1

好的，我只是尝试了一些可行的方法：我定义了 pattern analyzer on my partnumbers field, and when I tested with the Analyzer Text API，它确实将我的 SKU 分成了几个标记。之后我可以搜索一个 SKU，它会返回我想要的所有文件！这是我的索引 JSON 定义：

{
  "name": "my-index",
  "fields": [
    {
      "name": "partnumbers",
      "type": "Edm.String",
      "facetable": true,
      "filterable": true,
      "key": false,
      "retrievable": true,
      "searchable": true,
      "sortable": true,
      "analyzer": "pattern",
      "indexAnalyzer": null,
      "searchAnalyzer": null,
      "synonymMaps": [],
      "fields": []
    },
    {
      "name": "metadata_storage_name",
      "type": "Edm.String",
      "facetable": true,
      "filterable": true,
      "key": false,
      "retrievable": true,
      "searchable": true,
      "sortable": true,
      "analyzer": null,
      "indexAnalyzer": null,
      "searchAnalyzer": null,
      "synonymMaps": [],
      "fields": []
    },
    {
      "name": "metadata_storage_content_type",
      "type": "Edm.String",
      "facetable": true,
      "filterable": true,
      "key": false,
      "retrievable": true,
      "searchable": true,
      "sortable": true,
      "analyzer": null,
      "indexAnalyzer": null,
      "searchAnalyzer": null,
      "synonymMaps": [],
      "fields": []
    },
    {
      "name": "metadata_storage_last_modified",
      "type": "Edm.String",
      "facetable": true,
      "filterable": true,
      "key": false,
      "retrievable": true,
      "searchable": true,
      "sortable": true,
      "analyzer": null,
      "indexAnalyzer": null,
      "searchAnalyzer": null,
      "synonymMaps": [],
      "fields": []
    },
    {
      "name": "metadata_storage_path",
      "type": "Edm.String",
      "facetable": true,
      "filterable": true,
      "key": false,
      "retrievable": true,
      "searchable": true,
      "sortable": true,
      "analyzer": null,
      "indexAnalyzer": null,
      "searchAnalyzer": null,
      "synonymMaps": [],
      "fields": []
    },
    {
      "name": "metadata_storage_size",
      "type": "Edm.Int64",
      "facetable": true,
      "filterable": true,
      "retrievable": true,
      "sortable": true,
      "analyzer": null,
      "indexAnalyzer": null,
      "searchAnalyzer": null,
      "synonymMaps": [],
      "fields": []
    },
    {
      "name": "key",
      "type": "Edm.String",
      "facetable": true,
      "filterable": true,
      "key": true,
      "retrievable": true,
      "searchable": true,
      "sortable": true,
      "analyzer": null,
      "indexAnalyzer": null,
      "searchAnalyzer": null,
      "synonymMaps": [],
      "fields": []
    },
    {
      "name": "name",
      "type": "Edm.String",
      "facetable": true,
      "filterable": true,
      "key": false,
      "retrievable": true,
      "searchable": true,
      "sortable": true,
      "analyzer": null,
      "indexAnalyzer": null,
      "searchAnalyzer": null,
      "synonymMaps": [],
      "fields": []
    },
    {
      "name": "partialPartnumbers",
      "type": "Edm.String",
      "facetable": false,
      "filterable": false,
      "key": false,
      "retrievable": false,
      "searchable": true,
      "sortable": false,
      "analyzer": null,
      "indexAnalyzer": "prefixCmAnalyzer",
      "searchAnalyzer": "standardCmAnalyzer",
      "synonymMaps": [],
      "fields": []
    },
    {
      "name": "partialName",
      "type": "Edm.String",
      "facetable": false,
      "filterable": false,
      "key": false,
      "retrievable": false,
      "searchable": true,
      "sortable": false,
      "analyzer": null,
      "indexAnalyzer": "prefixCmAnalyzer",
      "searchAnalyzer": "standardCmAnalyzer",
      "synonymMaps": [],
      "fields": []
    }
  ],
  "suggesters": [
    {
      "name": "conformity-certificates-index_suggester",
      "searchMode": "analyzingInfixMatching",
      "sourceFields": [
        "name"
      ]
    }
  ],
  "scoringProfiles": [
    {
      "name": "exactFirst",
      "functions": [],
      "functionAggregation": null,
      "text": {
        "weights": {
          "partnumbers": 4,
          "partialPartnumbers": 3,
          "name": 2,
          "partialName": 1
        }
      }
    }
  ],
  "defaultScoringProfile": "exactFirst",
  "corsOptions": null,
  "analyzers": [
    {
      "@odata.type": "#Microsoft.Azure.Search.CustomAnalyzer",
      "name": "standardCmAnalyzer",
      "tokenizer": "standard_v2",
      "tokenFilters": [
        "lowercase",
        "asciifolding"
      ],
      "charFilters": []
    },
    {
      "@odata.type": "#Microsoft.Azure.Search.CustomAnalyzer",
      "name": "prefixCmAnalyzer",
      "tokenizer": "standard_v2",
      "tokenFilters": [
        "lowercase",
        "asciifolding",
        "edgeNGramCmTokenFilter"
      ],
      "charFilters": []
    }
  ],
  "charFilters": [],
  "tokenFilters": [
    {
      "@odata.type": "#Microsoft.Azure.Search.EdgeNGramTokenFilterV2",
      "name": "edgeNGramCmTokenFilter",
      "minGram": 2,
      "maxGram": 20,
      "side": "front"
    }
  ],
  "tokenizers": [],
  "@odata.etag": "\"0x8D818EC80CXXXX\""
}

Answer 2

这应该可以通过正则表达式和通配符搜索来实现

这可以应用于在索引上配置了 Lucene 查询分析器的任何可搜索字段。

"...通过设置 queryType=full 获得的 Full Lucene 查询语言通过添加对更多运算符和查询类型（如通配符、模糊、正则表达式和字段）的支持来扩展默认的简单查询语言作用域查询。例如，以简单查询语法发送的正则表达式将被解释为查询字符串而不是表达式。本文中的示例请求使用 Full Lucene 查询语言。"

字段名：搜索表达式

例如searchFields=零件编号&$select=零件编号&search=零件编号:102938*

https://docs.microsoft.com/en-us/azure/search/query-lucene-syntax

Answer 3

您可以使用常规过滤器搜索零件号。

$filter=search.in(partnumbers, '102938', ',')

您可以在此处的文档中找到更多示例：https://docs.microsoft.com/en-us/azure/search/search-query-odata-filter

请勿在此用例中使用通配符或正则表达式。您的示例具有不同长度的部件号。因此，对 102938* 的通配符搜索会无意中匹配 1029381、10293810、102938123 等。

您的数据已经明确且准确地列出了一组零件号。您可以查询该列表。

如何使用 Azure 认知搜索搜索字符串的一部分

How to search for part of a string with Azure Cognitive Search

search

azure

azure-cognitive-search

azure-blob-storage