This demo shows how to create an Azure Search Instance and index JSON files stored in an associated Blob Storage account.
- The ARM template will deploy both Search and Blob Storage using the button below. Note - keep the search and blob names short as there is a maximum length of 24 characters and the ARM template will add a guid to the end of the value you provide.
.
You will need 4 things for the subsequent REST calls, so gather these now from the Azure Portal:
- mysearch: the name of your search instance. This will be the name you gave to search in the ARM provisioning process, plus some unique characters.
- apikey: the Primary/Secondary Admin Key for your search instance.
- blobname: the name of your blob storage instance. This will be the name you gave to blob storage in the ARM provisioning process, plus some unique characters.
- blobaccountkey: the Azure Storage Access Key (either key1 or key2).
-
Create a container in the blob storage account called "allblobs" (you can call it something different, but you'll have to change the REST calls to reflect it), and upload a sample file to it. I used a publicly available food hygeine dataset from DATA.GOV.UK. The data is available in XML format so I used a simple XML to JSON transformation - the converted file is available here if you want to use this. I recommend using Azure Storage Explorer for uploading the file.
-
Next, you need to create a search index.
POST https://<mysearch>.search.windows.net/indexes?api-version=2015-02-28-Preview
Content-Type: application/json
api-key: <apikey>
{
"name": "default",
"fields": [
{"name": "id", "type": "Edm.String", "key": true, "searchable": false},
{"name": "businessname", "type": "Edm.String", "searchable": true, "filterable": true, "sortable": true, "facetable": true}
]
}
- Then, add a data source to point to your blob storage account.
POST https://<mysearch>.search.windows.net/datasources?api-version=2015-02-28-Preview
Content-Type: application/json
api-key: <apikey>
{
"name" : "blob-datasource",
"type" : "azureblob",
"credentials" : { "connectionString" : "DefaultEndpointsProtocol=https;AccountName=<blobname>;AccountKey=<blobaccountkey>" },
"container" : { "name" : "allblobs" }
}
- Once the data source has been added, then you need to add an indexer, which maps the data source to the index.
POST https://<mysearch>.search.windows.net/indexers?api-version=2015-02-28-Preview
Content-Type: application/json
api-key: <apikey>
{
"name" : "bob-json-indexer",
"dataSourceName" : "blob-datasource",
"targetIndexName" : "default",
"schedule" : { "interval" : "PT2H" },
"parameters" : { "configuration" : { "parsingMode" : "jsonArray", "documentRoot" : "/FHRSEstablishment/EstablishmentCollection/EstablishmentDetail" } },
"fieldMappings" : [
{ "sourceFieldName" : "/FHRSID", "targetFieldName" : "id" },
{ "sourceFieldName" : "/BusinessName", "targetFieldName" : "businessname" }
]
}
- Test! You can use the Azure Portal to test. The Search instance has a Search Explorer that you can use to type in a search and run the query - or you can query the REST endpoint directly. Here I show test results being returned using Search Explorer:
