Flattening/unnesting a text field/array to multiple levels

Question

I need to flatten a text column to two separate columns. I'm using:

SELECT s.CoolerShelf,
       s.ShelfPosition,
 FROM planogram
  CROSS JOIN LATERAL UNNEST(string_to_array(shelves, ','))
      WITH ORDINALITY s(CoolerShelf,ShelfPosition)

Result:

[["8d2cf35d-5708-45e0-9cb6-acad358e0f92" 1
"5a91f7a2-029a-46d7-8440-9337dd1b87d3" 2
"521562a9-9d33-438d-8156-1e6b1874ec8e" 3
"e14817e4-6630-4dca-a188-ac71060dcac9" 4
"76967052-ba9d-43f5-afd4-b4bbe1452d7e" 5
"2e5a6fb2-071e-426b-ac55-69f16baa0b42" 6
"108f263d-ee78-4124-a94b-2c5641f90321" 7
"0dbe5016-9e78-4173-b6e6-ff3e0199ca2e"] 8
["9bd83b79-186d-4ae5-9373-956dbd515070" 9
"b6172191-fa44-436d-879d-c883e4d240ed" 10
"093b72ba-74cd-48b9-86df-7e7d9341ae53" 11
"88b6c7f8-1d23-4e82-b959-8cb3400cc039" 12
"8279d979-8a57-4595-b9d3-346f6b05924e" 13
"735e6139-0fce-4bb7-a4a2-00ceb86c9b07" 14
"0ad84c4f-e0d8-4606-b563-8b2e32cc632f" 15
"5a86f7ea-0763-4473-ba09-91398e938be7"]} 16

Desired result:

coolershelf	shelfposition
1	0
1	1
1	2
1	3
1	4
1	5
1	6
1	7
2	0
2	1
2	2
2	3
2	4
2	5
2	6
2	7

The numbering needs to start from the beginning after every open and closed parenthesis and I'd like to replace guid values with actual numbers. I tried UNENST(string_to_array) .

Answer 1

Considering that your text column uses [ and ] to separate the groups of guid on two levels, we can convert the text into json format so that these characters are recognized as array start and end and then use the json_array_elements function to split the json arrays. Try this:

SELECT s.id AS CoolerShelf,
       (t.id - 1) AS ShelfPosition,
 FROM planogram
  CROSS JOIN LATERAL json_array_elements(shelves :: json)
      WITH ORDINALITY AS s(content,id)
 CROSS JOIN LATERAL json_array_elements(s.content)
      WITH ORDINALITY AS t(content,id)