I have a table of posts, each with an insight jsonb column containing something like the data sample below.
Data sample (comes from Facebook, so cannot change format)
[
{
"name": "post_story_adds_unique",
"values": [
{ "value": 93 }
],
},
{
"name": "post_story_adds",
"values": [
{ "value": 100 }
]
},
{
"name": "post_impressions_organic_unique",
"values": [
{ "value": 123 }
]
},
...
]
I want to have a calculated sum of reach and viral and then have it ordered by the total.
Desired Results
id message post_created reach viral total
69 This World Family dablah... 2016-05-11 18:44:16 6683 646 7329
...
I managed to get the results so far but I cannot figure how to get the sum of the two columns. I don't know where to add another join or select to sum the two columns.
Results so far
id message post_created reach viral
69 This World Family dablah... 2016-05-11 18:44:16 6683 646
58 blah blah flip flop blah... 2016-05-22 11:00:01 4880 403
55 This is another message ... 2016-05-24 10:00:00 4417 109
I've tried various ways such as including the SUM (reach + viral) as total in the first SELECT but mostly I get back errors saying columns don't exist.
Heres my SQL so far:
SELECT
id,
message,
post_created,
obj.value->'values'->0->'value' AS reach,
obj2.value->'values'->0->'value' AS viral
FROM (
SELECT
id,
message,
post_created,
insights
FROM posts
WHERE (
page_id = 4 AND
post_created >= '2016-05-01 00:00:00' AND
post_created <= '2016-05-31 23:59:59' AND
insights @> '[{"name":"post_impressions_organic_unique"}, {"name":"post_impressions_viral_unique"}]'
)
) e1
JOIN LATERAL jsonb_array_elements(insights) obj(value) ON obj.value->>'name' = 'post_impressions_organic_unique'
JOIN LATERAL jsonb_array_elements(insights) obj2(value) ON obj2.value->>'name' = 'post_impressions_viral_unique'
ORDER BY reach DESC;