Azure AI Search index field mapping

Question

Azure AI Search index field mapping

Rayne 20

I have an excel in Azure blob storage and now want to create an index with the fields as the column names in the excel and then populate the values of each field in the index. Let me know if this possible through field mapping or any other way.

SnehaAgrawal-MSFT 22,721 Reputation points Moderator

2024-08-26T12:12:54.37+00:00
@Rayne Yes, it is possible to create an index with fields as column names in the excel and populate the values of each field in the index using field mapping. You can use the Azure Cognitive Search indexer to extract data from your data source, map it to an index schema, and then load it into the index.

To create an index with fields as column names in the excel, you can define the field mappings in the indexer definition. You can use the FieldMappings property of the SearchIndexer class to define the field mappings.

{ "dataSourceName" : "mydatasource", "targetIndexName" : "myindex", "fieldMappings" : [ { "sourceFieldName" : "column1", "targetFieldName" : "field1" }, { "sourceFieldName" : "column2", "targetFieldName" : "field2" }, { "sourceFieldName" : "column3", "targetFieldName" : "field3" } ] }

In this example, the dataSourceName property specifies the name of the data source, the targetIndexName property specifies the name of the target index, and the fieldMappings property specifies the field mappings.

You can replace the column1, column2, and column3 with the actual column names in your excel file, and replace field1, field2, and field3 with the desired field names in your index schema.

Once you have defined the field mappings, you can create the indexer and run it to extract data from your data source, map it to the index schema, and load it into the index.

See- https://learn.microsoft.com/en-us/azure/search/search-indexer-field-mappings?tabs=rest

Let us know.
Rayne 20 Reputation points

2024-08-29T02:50:14.8566667+00:00

Already tried this but it didn't worked. I have multiple fields like id, name, category, content, etc.. So whats happening its mapping the fields with metadata of the file. When I executed the indexer it was showing 1 doc succeeded but the count should be equal to the number of rows in that excel. And when I checked the content id field was having default document id like aHR0cHM6...... And rest of the fields were null. Then to test from where it's picking the value I added name as metadata of the file and then executed the indexer and it was showing same id and name with rest of the fields as null.