How to update multiple SQL records, with merge with multiple options, based on an XML document?

I'm relatively new to SQL (a week or so), and I have a bunch of xml files that have a structure somewhat like this:

<foods>
     <type>
       <food_id>123456789</food_id>
       <food_type>bacon</food_type>
       <food_information>
           <stuff>blah</stuff>
       </food_information>
       <more_food_info>
           <more_stuff>blah</more_stuff>
       </more_food_info>
     </type>
     <type>
       <food_id>123456780</food_id>
       <food_type>eggs</food>
       <food_information>
           <stuff>blahblah</stuff>
       </food_information>
       <more_food_info>
           <more_stuff>blahblahblah</more_stuff>
       </more_food_info>
     </type>
</foods>

I am doing Bulk inserts via SQL. Now, this database contains tens of millions of records of food. Almost everything is unique, but the one unique value is the "food_id" value. It is always unique. However, newer XML files contain updates for the old records, and I need to be able to use ONLY those new records, not the old ones.

food_information is a table containing a bunch of stuff.

more_food_info is also another table containing a bunch of stuff.

Everything has the proper primary and foreign key relationship set up. The main table has a primary key of Type_ID. The sub-tables have a primary key of Name_ID without (e.g.: FoodInformation_ID).

My insertion code also works perfectly:

DECLARE @xml_data xml
DECLARE @xdoc INT

SET @xml_data = (SELECT CONVERT(xml, BulkColumn, 2) FROM OPENROWSET(Bulk 'foods.xml', SINGLE_CLOB) [rowresults])

EXEC sp_xml_preparedocument @xdoc, OUTPUT, @xml_data

INSERT INTO [type]
    SELECT * FROM OPENXML(@xdoc, '/foods/type', 2)
    WITH
    (
        [food_id] VARCHAR(8) 'food_id',
        [food_type] VARCHAR(255) 'food_type',
    )

INSERT INTO [food_information]
    SELECT 
        [type].[Type_ID], 
        xml.[stuff] FROM [type]
    INNER JOIN OPENXML(@xdoc, '/foods/type', 2)
    WITH 
    ( 
        [food_id] VARCHAR(8) '../food_id', 
        [stuff] VARCHAR(255) 'stuff'
    )
    AS XML ON xml.[food_id] = [type].[food_id]
    -- Repeat the same style for every other table... (not the same as the [type] table)

EXEC sp_xml_removedocument

This works 100%, but I need to be able to match everything from these foreign keys to the primary key, and, of course, to the serial number, to update all records based on it. I don't know how to do this for multiple records.

I need all tables joined in a MERGE update, but I only know how to merge the FIRST table since I can't seem to insert other conditions from other tables. Or I need a way to find all records relating to that food_id and delete the lowest Type_ID number (primary key) for each record and match that to the corresponding foreign key to delete the other records in the other tables.

The problem is that food_id when reading bulk xml returns a giant list of ids. I only want the one we're currently on, so I can't compare it. And "IN (select statement)" is not what I'm looking for: it has to be one at a time.

TLDR: How do I do this? psuedo-code:

Read XML document.
If exists (SELECT [food_id] from [type] WHERE [food_id] = current_xml_food_id)
    THEN
    -- Update/MERGE stuff
    -- Include all the other tables that need updating: 
    -- ... food_information, more_food_info, etc.
    -- ...and make sure the foreign key relationship stays
    -- ...intact. We need to match Type_ID (main table 
    -- ...PrimaryKey) to TableName_ID (ForeignKey)
    ELSE
    -- Insert stuff
End read