import os

import sentry_sdk

from config import config
from queries.mysql import get_error_jobs_with_queued_claims, update_claim_status_for_job
from queries.mysql import update_claims_status
from s3.read_s3 import get_file_list, get_file_content

log = config.get_logger()
sentry_sdk.init(config.SENTRY_DSN)


def fix_claims_for_error_job(env):
    batch_size = 50
    bucket_name = f'{env.lower()}-neighbouring-rights'
    job_count = 0
    has_results = True
    offset = 0

    while has_results:
        log.info(f'Getting jobs that need to be fixed at offset: {offset}')
        ids, rowcount = get_error_jobs_with_queued_claims(
            batch_size, offset
        )
        has_results = bool(ids)
        log.info(f'Found #{rowcount} jobs')
        if not has_results:
            break

        offset += batch_size
        job_count += rowcount
        all_files = []
        for job in ids:
            # for each job, get a list of files that are on s3 folder.
            # it will have error and maybe an excluded files.
            # all these are old jobs so use the old folder structure.
            prefix = f"delivery/claims/order_{job['order_id']}_job_{job['job_id']}/"
            files = get_file_list(bucket_name, prefix)
            log.info(f"Files in {job['job_id']}: {files}")
            if len(files) == 0:
                log.info(f"No files found for job: {job['job_id']}")
                log.info(f"Going to update all the claims for job: {job['job_id']}")
                # If there are no files, then update the job status
                # This function has been giving us a SQL error in the past for jobs with no files.
                # This should be fixed now but have a look to https://theorchard.atlassian.net/browse/NR-1123 and
                # https://theorchard.atlassian.net/jira/software/c/projects/NR/boards/488?selectedIssue=NR-1643
                # for more information.
                update_claim_status_for_job(job['job_id'], 'error')
            else: # if there are files, then update the individual claims statuses
                all_files += files

        for file in all_files:
            rows = get_file_content(bucket_name, file, log)
            status = 'excluded' if '_excluded.csv' in file else 'error'
            log.info(
                f"Updating #{len(rows)} claims from file: {file} with status: {status}")
            update_claims_status(rows, status)

    log.info(f'Total {job_count} jobs fixed.')


if __name__ == "__main__":
    """Main entrypoint function."""
    env = os.environ.get("ENV", 'qa')
    try:
        log.info(f'Starting fix_claims_for_error_job for env={env}')
        fix_claims_for_error_job(env)
    except Exception as e:
        log.error(e)
        raise e
    finally:
        log.info("Finished.")