Oulu

import boto3
import subprocess
import json

client = boto3.client("dynamodb")

customer_ids = [
    "34460669"
]

prefixes = [
    "cashflow-amount"
]

def fetch_timeseries(customer_id, prefix):
    exclusive_start_key = None
    items = []

    while True:
        kwargs = {
            "TableName": "tcb-ps5-data-uat1b-customer-features-ts",
            "KeyConditionExpression": "customer_code = :customer_code and begins_with(timeseries_key, :timeserie_prefix)",
            "ExpressionAttributeValues": {
                ":customer_code": { "S": customer_id },
                ":timeserie_prefix": { "S": customer_id + "__" + prefix },
            },
        }

        if exclusive_start_key:
            kwargs["ExclusiveStartKey"] = exclusive_start_key

        response = client.query(**kwargs)

        exclusive_start_key = response.get("LastEvaluatedKey", None)
        for item in response["Items"]:
            items.append({
                "customer_code": item["customer_code"]["S"],
                "timeseries_key": item["timeseries_key"]["S"],
            })

        if not exclusive_start_key:
            break

    return items


def parse_data(data):
    lines = data.split("\n")
    output = { "history": {} }
    for line in lines:
        if line.startswith("TimeSeriesKey"):
            item = line.split(": ")
            output["time_series_key"] = item[1]
        elif line.startswith("Customer Code"):
            item = line.split(": ")
            output["customer_code"] = item[1]
        elif line.startswith("Dimensions") or line.startswith("Dimemsions"):
            item = line.split(": ")
            output["dimensions"] = item[1]
        elif line[:4].isnumeric():
            item = line.split(" = ")
            output["history"][item[0]] = item[1]
    return output


def decrypt_data(item):
    process = subprocess.Popen(
        [
            "./timeseries-cli.sh",
            "read-partition",
            item["timeseries_key"],
        ],
        stdout=subprocess.PIPE,
        stderr=subprocess.PIPE,
    )

    stdout, stderr = process.communicate()
    return parse_data(stdout.decode("utf-8"))



result = []
for customer_id in customer_ids:
    for prefix in prefixes:
        items = fetch_timeseries(customer_id, prefix)
        for item in items:
            result.append(decrypt_data(item))

with open("result.json", "w") as f:
    json.dump(result, f)

print(json.dumps(result[:1], indent=2))
print("Count", len(result))


# aws s3 cp result.json s3://atasync1