Search code examples
postgresqljsonb

PostgreSQL open JSONB column with slashes


How is it possible to open the JSONB column where JSONB string with slashes?

"{\"id\":\"c39fe0f5f9b7c89b005bf3491f2a2ce1\",\"token\":\"c39fe0f5f9b7c89b005bf3491f2a2ce1\",\"line_items\":[{\"id\":32968150843480,\"properties\":{},\"quantity\":2,\"variant_id\":32968150843480,\"key\":\"32968150843480:4a6f6b7d19c7aef119af2cd909f429f1\",\"discounted_price\":\"40.00\",\"discounts\":[],\"gift_card\":false,\"grams\":0,\"line_price\":\"80.00\",\"original_line_price\":\"80.00\",\"original_price\":\"40.00\",\"price\":\"40.00\",\"product_id\":4638774493272,\"sku\":\"36457537-mud-yellow-28\",\"taxable\":false,\"title\":\"Knee Length Summer Shorts - Camel / 28\",\"total_discount\":\"0.00\",\"vendor\":\"Other\",\"discounted_price_set\":{\"shop_money\":{\"amount\":\"40.0\",\"currency_code\":\"USD\"},\"presentment_money\":{\"amount\":\"40.0\",\"currency_code\":\"USD\"}},\"line_price_set\":{\"shop_money\":{\"amount\":\"80.0\",\"currency_code\":\"USD\"},\"presentment_money\":{\"amount\":\"80.0\",\"currency_code\":\"USD\"}},\"original_line_price_set\":{\"shop_money\":{\"amount\":\"80.0\",\"currency_code\":\"USD\"},\"presentment_money\":{\"amount\":\"80.0\",\"currency_code\":\"USD\"}},\"price_set\":{\"shop_money\":{\"amount\":\"40.0\",\"currency_code\":\"USD\"},\"presentment_money\":{\"amount\":\"40.0\",\"currency_code\":\"USD\"}},\"total_discount_set\":{\"shop_money\":{\"amount\":\"0.0\",\"currency_code\":\"USD\"},\"presentment_money\":{\"amount\":\"0.0\",\"currency_code\":\"USD\"}}}],\"note\":null,\"updated_at\":\"2022-03-15T13:24:02.787Z\",\"created_at\":\"2022-03-15T13:23:31.912Z\",\"controller\":\"custom_webhooks\",\"action\":\"store_data\",\"custom_webhook\":{\"id\":\"c39fe0f5f9b7c89b005bf3491f2a2ce1\",\"token\":\"c39fe0f5f9b7c89b005bf3491f2a2ce1\",\"line_items\":[{\"id\":32968150843480,\"properties\":{},\"quantity\":2,\"variant_id\":32968150843480,\"key\":\"32968150843480:4a6f6b7d19c7aef119af2cd909f429f1\",\"discounted_price\":\"40.00\",\"discounts\":[],\"gift_card\":false,\"grams\":0,\"line_price\":\"80.00\",\"original_line_price\":\"80.00\",\"original_price\":\"40.00\",\"price\":\"40.00\",\"product_id\":4638774493272,\"sku\":\"36457537-mud-yellow-28\",\"taxable\":false,\"title\":\"Knee Length Summer Shorts - Camel / 28\",\"total_discount\":\"0.00\",\"vendor\":\"Other\",\"discounted_price_set\":{\"shop_money\":{\"amount\":\"40.0\",\"currency_code\":\"USD\"},\"presentment_money\":{\"amount\":\"40.0\",\"currency_code\":\"USD\"}},\"line_price_set\":{\"shop_money\":{\"amount\":\"80.0\",\"currency_code\":\"USD\"},\"presentment_money\":{\"amount\":\"80.0\",\"currency_code\":\"USD\"}},\"original_line_price_set\":{\"shop_money\":{\"amount\":\"80.0\",\"currency_code\":\"USD\"},\"presentment_money\":{\"amount\":\"80.0\",\"currency_code\":\"USD\"}},\"price_set\":{\"shop_money\":{\"amount\":\"40.0\",\"currency_code\":\"USD\"},\"presentment_money\":{\"amount\":\"40.0\",\"currency_code\":\"USD\"}},\"total_discount_set\":{\"shop_money\":{\"amount\":\"0.0\",\"currency_code\":\"USD\"},\"presentment_money\":{\"amount\":\"0.0\",\"currency_code\":\"USD\"}}}],\"note\":null,\"updated_at\":\"2022-03-15T13:24:02.787Z\",\"created_at\":\"2022-03-15T13:23:31.912Z\"}}"

this real JSONB column enter image description here

I can not find any example of how to deal with this type of JSONB


Solution

  • Whatever is inserting your data is screwing it up. It is taking the string representation of a JSON object and stuffing that into a JSON string scalar. You need to fix that or it will just keep happening.

    To fix what is already there, you need to extract the real PostgreSQL string out of the JSON string, then cast that to JSONB. Extracting a JSON string can be done unintuitively with #>>'{}', or even less intuitively with ->>0.

    select (data#>>'{}')::jsonb from table_name.
    

    Of course you should fix it permanently, not just do it on the fly all the time, which is both slow and confusing.

    update table_name set data=(data#>>'{}')::jsonb;
    

    Of course fixing the tool which screws this up in the first place, and fixing the historical data, need to be done in a coordinated fashion or you will have a glorious mess on your hands.