"""

This script compares the data for social timeseries endpoints backed by datasources:
1. neo4j (old way)
2. snowflake (new way)

By default it uses graphql on QA.

In order to make this thing work you need to set ENV variables in .env

ORCHARD_IDENTITY_ID = (uuid)
which is your account id in the insights
ORCHARD_PROFILE_ID = (int)
can be found in dev console in chrome: application -> local storage -> user traits

"""
import os

import dotenv
import pytest

dotenv.load_dotenv()

@pytest.fixture
def graphql_endpoint():
    headers = {
        "orchard-identity-id": os.environ['ORCHARD_IDENTITY_ID'],
        "orchard-profile-id": os.environ['ORCHARD_PROFILE_ID'],
        "orchard-profile-type": "InsightsProfile",
        "apollographql-client-name": "test-frontend-insights",
        "apollographql-client-version": "1.0.0",
        "cache-control": "no-cache"
    }
    from sgqlc.endpoint.http import HTTPEndpoint

    url = os.environ.get("GRAPHQL_URL", "https://qa-graphql-gateway.theorchard.io/graphql")
    yield HTTPEndpoint(url, headers)


@pytest.mark.parametrize('participant_id, date_lte, date_gte', [
    ("37d93121-6b6b-44f9-a383-e70ebdcb0c22", "2021-03-20", "2021-03-15"),
    ("37d93121-6b6b-44f9-a383-e70ebdcb0c22", "2020-03-01", "2020-01-01"),
    ("37d93121-6b6b-44f9-a383-e70ebdcb0c22", "2020-05-01", "2020-03-01"),
    ("37d93121-6b6b-44f9-a383-e70ebdcb0c22", "2020-07-01", "2020-05-01"),
    ("37d93121-6b6b-44f9-a383-e70ebdcb0c22", "2020-09-01", "2020-07-01"),
    ("37d93121-6b6b-44f9-a383-e70ebdcb0c22", "2020-11-01", "2020-09-01"),
    ("37d93121-6b6b-44f9-a383-e70ebdcb0c22", "2021-01-01", "2020-11-01"),
    ("37d93121-6b6b-44f9-a383-e70ebdcb0c22", "2021-03-01", "2021-01-01"),

    ("bc5d3083-ec53-4ab5-9756-50511d39e33c", "2021-03-20", "2021-03-15"),
    ("bc5d3083-ec53-4ab5-9756-50511d39e33c", "2020-03-01", "2020-01-01"),
    ("bc5d3083-ec53-4ab5-9756-50511d39e33c", "2020-05-01", "2020-03-01"),
    ("bc5d3083-ec53-4ab5-9756-50511d39e33c", "2020-07-01", "2020-05-01"),
    ("bc5d3083-ec53-4ab5-9756-50511d39e33c", "2020-09-01", "2020-07-01"),
    ("bc5d3083-ec53-4ab5-9756-50511d39e33c", "2020-11-01", "2020-09-01"),
    ("bc5d3083-ec53-4ab5-9756-50511d39e33c", "2021-01-01", "2020-11-01"),
    ("bc5d3083-ec53-4ab5-9756-50511d39e33c", "2021-03-01", "2021-01-01"),

    ("7745c66a-e044-4330-9452-0b44d7362424", "2020-03-01", "2020-01-01"),
    ("7745c66a-e044-4330-9452-0b44d7362424", "2020-05-01", "2020-03-01"),
    ("7745c66a-e044-4330-9452-0b44d7362424", "2020-07-01", "2020-05-01"),
    ("7745c66a-e044-4330-9452-0b44d7362424", "2020-09-01", "2020-07-01"),
    ("7745c66a-e044-4330-9452-0b44d7362424", "2020-11-01", "2020-09-01"),
    ("7745c66a-e044-4330-9452-0b44d7362424", "2021-01-01", "2020-11-01"),
    ("7745c66a-e044-4330-9452-0b44d7362424", "2021-03-01", "2021-01-01"),
])
def test_social_accounts_timeseries(participant_id, date_lte, date_gte, graphql_endpoint):
    old_query = """
    query GlobalParticipantSocials($id: ID!, $startDate: String, $endDate: String){
        globalParticipantByGpId (id: $id) {
            id
            participant {
                id
                socialStats {
                    monthlyListeners
                    popularity
                    followers
                }
                socialAccounts {
                    id
                    platform
                    url
                    timeseries(filter: { timestamp_gte: { formatted:  $startDate }, timestamp_lte: { formatted:  $endDate } } ) {
                        followers
                        monthlyListeners
                        popularity
                        views
                        timestamp {
                            formatted
                        }
                    }
                }
            }
        }
    }
    """

    new_query = """
query GlobalParticipantSocialsV2($id: ID!, $startDate: String, $endDate: String){
    globalParticipantByGpId (id: $id) {
        id
        participant {
            id
            socialStats {
                monthlyListeners
                popularity
                followers
            }
            socialAccounts {
                id
                platform
                url
                timeseriesV2(startDate: $startDate, endDate: $endDate) {
                    followers
                    monthlyListeners
                    popularity
                    views
                    timestamp
                }
            }
        }
    }
}
    """
    variables = {
        "id": participant_id,
        "startDate": date_gte,
        "endDate": date_lte,
        "filter": {
            "timestamp_gte": {
                "formatted": date_gte
            },
            "timestamp_lte": {
                "formatted": date_lte
            }
        }
    }

    response = graphql_endpoint(old_query, variables)
    assert not 'errors' in response, response['errors']
    old_data = response['data']['globalParticipantByGpId'][0]['participant']['socialAccounts']
    response = graphql_endpoint(new_query, variables)
    assert not 'errors' in response, response['errors']
    new_data = response['data']['globalParticipantByGpId'][0]['participant']['socialAccounts']
    # convert new data to the old data structure
    for social_account in new_data:
        social_account['timeseries'] = social_account['timeseriesV2']
        for ts in social_account['timeseries']:
            ts['timestamp'] = {'formatted': ts['timestamp']}
        del social_account['timeseriesV2']

    new_social_account_by_url = {a['url']:a for a in new_data}
    old_social_account_by_url = {a['url']:a for a in old_data}
    # let's compare the data
    counter = 0
    for url in old_social_account_by_url:
        old_social_account = old_social_account_by_url[url]
        new_social_account = new_social_account_by_url[url]
        counter += 1
        print(f"{counter} OLD: {old_social_account['url']} NEW: {new_social_account['url']}")
        # extract timeseries
        old_timeseries_list = old_social_account['timeseries']
        new_timeseries_list = new_social_account['timeseries']
        # create dicts for timeseries
        old_timeseries = {v['timestamp']['formatted']:v for v in old_timeseries_list}
        new_timeseries = {v['timestamp']['formatted']:v for v in new_timeseries_list}
        for timestamp, old_value in old_timeseries.items():
            if timestamp.startswith('2020-09-2') \
                    or timestamp.startswith('2020-09-3') \
                    or timestamp.startswith('2020-10-0') \
                    or timestamp.startswith('2020-10-20'):
                # DS-3330 - missing data in chartmetrics
                continue
            assert new_timeseries[timestamp] == old_value

        # compare all of the data excpet timeseries
        del old_social_account['timeseries']
        del new_social_account['timeseries']
        assert old_social_account == new_social_account


@pytest.mark.parametrize('participant_id, date_lte, date_gte', [
    ("37d93121-6b6b-44f9-a383-e70ebdcb0c22", "2021-03-20", "2021-03-15"),
    ("37d93121-6b6b-44f9-a383-e70ebdcb0c22", "2020-03-01", "2020-01-01"),
    ("37d93121-6b6b-44f9-a383-e70ebdcb0c22", "2020-05-01", "2020-03-01"),
    ("37d93121-6b6b-44f9-a383-e70ebdcb0c22", "2020-07-01", "2020-05-01"),
    ("37d93121-6b6b-44f9-a383-e70ebdcb0c22", "2020-09-01", "2020-07-01"),
    ("37d93121-6b6b-44f9-a383-e70ebdcb0c22", "2020-11-01", "2020-09-01"),
    ("37d93121-6b6b-44f9-a383-e70ebdcb0c22", "2021-01-01", "2020-11-01"),
    ("37d93121-6b6b-44f9-a383-e70ebdcb0c22", "2021-03-01", "2021-01-01"),

    ("bc5d3083-ec53-4ab5-9756-50511d39e33c", "2021-03-20", "2021-03-15"),
    ("bc5d3083-ec53-4ab5-9756-50511d39e33c", "2020-03-01", "2020-01-01"),
    ("bc5d3083-ec53-4ab5-9756-50511d39e33c", "2020-05-01", "2020-03-01"),
    ("bc5d3083-ec53-4ab5-9756-50511d39e33c", "2020-07-01", "2020-05-01"),
    ("bc5d3083-ec53-4ab5-9756-50511d39e33c", "2020-09-01", "2020-07-01"),
    ("bc5d3083-ec53-4ab5-9756-50511d39e33c", "2020-11-01", "2020-09-01"),
    ("bc5d3083-ec53-4ab5-9756-50511d39e33c", "2021-01-01", "2020-11-01"),
    ("bc5d3083-ec53-4ab5-9756-50511d39e33c", "2021-03-01", "2021-01-01"),

    ("7745c66a-e044-4330-9452-0b44d7362424", "2020-03-01", "2020-01-01"),
    ("7745c66a-e044-4330-9452-0b44d7362424", "2020-05-01", "2020-03-01"),
    ("7745c66a-e044-4330-9452-0b44d7362424", "2020-07-01", "2020-05-01"),
    ("7745c66a-e044-4330-9452-0b44d7362424", "2020-09-01", "2020-07-01"),
    ("7745c66a-e044-4330-9452-0b44d7362424", "2020-11-01", "2020-09-01"),
    ("7745c66a-e044-4330-9452-0b44d7362424", "2021-01-01", "2020-11-01"),
    ("7745c66a-e044-4330-9452-0b44d7362424", "2021-03-01", "2021-01-01"),
])
def test_social_accounts_v2_timeseries(participant_id, date_lte, date_gte, graphql_endpoint):
    """
    Test improved socialAccountsV2, the improved version of socialAccounts
    """
    old_query = """
    query GlobalParticipantSocials($id: ID!, $startDate: String, $endDate: String){
        globalParticipantByGpId (id: $id) {
            id
            participant {
                id
                socialStats {
                    monthlyListeners
                    popularity
                    followers
                }
                socialAccounts {
                    platform
                    url
                    timeseries(filter: { timestamp_gte: { formatted:  $startDate }, timestamp_lte: { formatted:  $endDate } } ) {
                        followers
                        monthlyListeners
                        popularity
                        views
                        timestamp {
                            formatted
                        }
                    }
                }
            }
        }
    }
    """

    new_query = """
query GlobalParticipantSocialsV2($id: ID!, $startDate: String, $endDate: String){
    globalParticipantByGpId (id: $id) {
        id
        participant {
            id
            socialStats {
                monthlyListeners
                popularity
                followers
            }
            socialAccountsV2(startDate: $startDate, endDate: $endDate) {
                platform
                url
                timeseriesV2 {
                    followers
                    monthlyListeners
                    popularity
                    views
                    timestamp
                }
            }
        }
    }
}
    """
    variables = {
        "id": participant_id,
        "startDate": date_gte,
        "endDate": date_lte,
        "filter": {
            "timestamp_gte": {
                "formatted": date_gte
            },
            "timestamp_lte": {
                "formatted": date_lte
            }
        }
    }

    response = graphql_endpoint(old_query, variables)
    assert not 'errors' in response, response['errors']
    old_data = response['data']['globalParticipantByGpId'][0]['participant']['socialAccounts']
    response = graphql_endpoint(new_query, variables)
    assert not 'errors' in response, response['errors']
    new_data = response['data']['globalParticipantByGpId'][0]['participant']['socialAccountsV2']
    # convert new data to the old data structure
    for social_account in new_data:
        social_account['timeseries'] = social_account['timeseriesV2']
        for ts in social_account['timeseries']:
            ts['timestamp'] = {'formatted': ts['timestamp']}
        del social_account['timeseriesV2']

    new_social_account_by_url = {a['url']:a for a in new_data}
    old_social_account_by_url = {a['url']:a for a in old_data}
    # let's compare the data
    counter = 0
    for url in old_social_account_by_url:
        old_social_account = old_social_account_by_url[url]
        if not old_social_account.get('timeseries'):
            # skip if there is no data in old result
            print(f'SKIP no data for {url}')
            continue
        new_social_account = new_social_account_by_url[url]
        counter += 1
        print(f"{counter} OLD: {old_social_account['url']} NEW: {new_social_account['url']}")
        # extract timeseries
        old_timeseries_list = old_social_account['timeseries']
        new_timeseries_list = new_social_account['timeseries']
        # create dicts for timeseries
        old_timeseries = {v['timestamp']['formatted']:v for v in old_timeseries_list}
        new_timeseries = {v['timestamp']['formatted']:v for v in new_timeseries_list}
        for timestamp, old_value in old_timeseries.items():
            if timestamp.startswith('2020-09-2') \
                    or timestamp.startswith('2020-09-3') \
                    or timestamp.startswith('2020-10-0') \
                    or timestamp.startswith('2020-10-20'):
                # DS-3330 - missing data in chartmetrics
                continue
            assert new_timeseries[timestamp] == old_value

        # compare all of the data excpet timeseries
        del old_social_account['timeseries']
        del new_social_account['timeseries']
        assert old_social_account == new_social_account