extract values inside an array column in amazon athena-CodePudding

I have a table in athena aws where the column 'metadata_stopinfo' has the structure that you can see in the image.

I am trying to extract values that are inside that array, however when I try

SELECT 
   "json_extract_scalar"(metadata_stopinfo, '$.city')
FROM "table"

I have the following problem

SYNTAX_ERROR: line 2:5: Unexpected parameters (array(row("address" row("addressline" varchar,"city" varchar,"countrycode" varchar,"countrycodeoriginal" varchar,"state" varchar,"zipcode" varchar),"carrierreference" varchar,"contacts" array(row("contacttype" varchar,"email" varchar,"fax" varchar,"mobilephone" varchar,"name" varchar,"officephone" varchar,"userid" varchar)),"containerinfo" array(row("containerid" varchar,"containeridtype" varchar,"equipmentcode" varchar,"equipmenttype" varchar)),"conveyancelinenumber" varchar,"conveyancetype" varchar,"conveyancetypeoriginal" varchar,"dateinfo" row("arrivalestimateddate" varchar,"arrivalestimateddateend" varchar,"arrivalestimatedendoffset" varchar,"arrivalestimatedoffset" varchar,"arrivalrequesteddate" varchar,"deliveryestimateddate" varchar,"deliveryestimateddateend" varchar,"deliveryestimatedendoffset" varchar,"deliveryestimatedoffset" varchar,"deliveryrequesteddate" varchar,"deliveryrequesteddateend" varchar,"deliveryrequestedendoffset" varchar,"deliveryrequestedoffset" varchar,"departureestimateddate" varchar,"departureestimateddateend" varchar,"departureestimatedendoffset" varchar,"departureestimatedoffset" varchar,"departurerequesteddate" varchar,"pickuprequesteddate" varchar,"pickuprequesteddateend" varchar,"pickuprequestedendoffset" varchar,"pickuprequestedoffset" varchar,"pickupestimateddate" varchar,"pickupestimateddateend" varchar,"pickupestimatedendoffset" varchar,"pickupestimatedoffset" varchar),"deliverynotenumber" varchar,"instructions" array(row("customerspecificsubtype" varchar,"header" boolean,"instructionsubtype" varchar,"instructiontype" varchar,"text" varchar)),"locationid" varchar,"partnercarrieraddress" row("addressline" varchar,"city" varchar,"countrycode" varchar,"countrycodeoriginal" varchar,"state" varchar,"zipcode" varchar),"partnercarriercontacts" array(row("contacttype" varchar,"email" varchar,"fax" varchar,"name" varchar,"officephone" varchar)),"partnercarrierid" varchar,"partnercarriername" varchar,"partnerid" varchar,"partnername" varchar,"partnertimezone" varchar,"partnertype" varchar,"productquantity" row("number" double,"originalunitofmeasure" varchar,"quantitytype" varchar,"unitofmeasure" varchar),"sequencenumber" bigint,"shipmentidentifier" varchar,"stoptype" varchar,"transportinfo" row("description" varchar,"transportcode" varchar,"transportoriginalcode" varchar),"vesselinfo" row("lloydsnumber" varchar,"shipsradiocallnumber" varchar,"vesselname" varchar,"vesselnumber" varchar,"voyagetripnumber" varchar))), varchar(6)) for function json_extract_scalar. Expected: json_extract_scalar(varchar(x), JsonPath) , json_extract_scalar(json, JsonPath)

My question is, how can i extract values inside de column ?

CodePudding user response：

json_extract_scalar unsurprisingly works with json (note that even if yur data was in json format, json_extract_scalar(metadata_stopinfo, '$.city') still would not have worked cause your data is an array), while your column contains array's of row's, so you need to work with it correspondingly. For example you can use indexes to access elements in array (in presto array indexes start from 1):

SELECT 
   metadata_stopinfo[1] r
FROM "table"

And then access the fields:

The fields may be of any SQL type, and are accessed with field reference operator .

SELECT 
   metadata_stopinfo[1].city city
FROM "table"

Also you can flatten the array with unnest:

SELECT r.city
FROM "table",
    unnest(metadata_stopinfo) as t(r)