import csv

import json

import re

"""

Pulling the set of functions out of Postgres to build a raw testing set:

SELECT nspname || '.' || proname as name,

json_agg(

json_build_object(

'name',

(string_to_array(longstring.params, ' '))[1],

'type',

array_to_string((string_to_array(longstring.params, ' '))[2:],

' ')

)

) AS params

FROM pg_catalog.pg_namespace n

JOIN pg_catalog.pg_proc p

ON pronamespace = n.oid

JOIN (

SELECT unnest(regexp_split_to_array(pg_get_function_arguments(p.oid), ', ')) AS params, proname AS funname

FROM pg_catalog.pg_namespace n

JOIN pg_catalog.pg_proc p

ON pronamespace = n.oid

WHERE nspname IN ('ap','da','ecg','ti','ts')

) AS longstring

ON longstring.funname = p.proname

WHERE nspname IN ('ap','da','ecg','ti','ts')

GROUP BY nspname, proname;

"""

with open('datachecks/neotoma_fun.csv') as f:

funcfile = csv.reader(f)

silent = next(f)

functions = list(funcfile)

test_funs = []

for i in functions:

i[1] = re.sub("\'", '"', i[1])

i[1]=re.sub("\: None", ':""', i[1])

keys = list(map(lambda x: {x.get('name'):x.get('type')}, json.loads(i[1])))

dictout = {}

dictout.update(list(d.items())[0] for d in keys)

test_funs.append([i[0], dictout])

with open('datachecks/neotoma_fun_raw.json', 'w') as f:

json.dump(test_funs, f)