How to insert/update `ARRAY` column in Postgres using Bookshelf/Knex

Question

I have a table in my Postgres database which contains a column with datatype ARRAY . I am using Bookshelf to perform operations on the database. Now, I want to insert/update (append new data to previous data in an array) data in this column and I could not find a way to do this. Can anybody guide me on how I can achieve this? I think, one way of doing this could be using raw() function of Knex but I am not sure about how to use raw() function so please guide me on that too. Thanks.

Answer 1

I found the solution for this problem here . It seems BookshelfJS does not have a way to handle such operations so I had to use KnexJS. I implemented it something like this -

knex('users')                               //users table
        .where('id', id)
        .update({
            array_column_name: knex.raw('array_append(array_column_name, ?)', [data_to_append])
        })
        .then(function (user) {
            //Do something 
        });

Hope this helps other people in future.

Answer 2

Assuming the following table schema

CREATE TABLE test (id SERIAL PRIMARY KEY, data TEXT[] NOT NULL);

with example data

INSERT INTO test (data) VALUES (array[ 'def', 'ghi' ]);

which can be queried like

SELECT * FROM test ;
 id |   data    
----+-----------
  1 | {def,ghi}

You can use thearray functions to array_prepend( 'abc', array ) and array_append( array, 'xyz' ) like so

UPDATE test SET data = array_prepend( 'abc', data );
UPDATE test SET data = array_append( data, 'xyz' );

to get

SELECT * FROM test;
id  |       data        
----+-------------------
  1 | {abc,def,ghi,xyz}

You should be aware that the data column is not atomic and therefore this table schema does not adhere to first normal form (violation of 1NF). It will be more difficult to filter out values from the data column. You cannot use the WHERE clause easily, for example. Consider adapting the table schema to adhere to 1NF, better 3NF at least.

Answer 3

UPDATE tableName SET ColumnName = array_prepend( 'Value', ColumnName) WHERE ColumnName = 'Value' UPDATE tblTest SET TestColumnName = array_prepend( 'TestData', TestColumnName) WHERE TestColumnId = '1'