import csv
import os
import re
from datetime import datetime

import urllib3
from urllib3 import Timeout

JENKINS_USER = os.getenv('JENKINS_USER', 'jdenniss')
JENKINS_API_TOKEN = os.getenv('JENKINS_API_TOKEN')
JENKINS_URL = os.getenv('JENKINS_INSTANCE', 'https://scheduler.theorchard.io')

TOOL_TO_FIND = os.getenv('TOOL_TO_FIND', 'python3.8')

DOCKER_BUILD_REGEX = r'^(\[\d{4}-\d{2}-\d{2}T\d{2}:\d{2}:\d{2}\.\d{3}Z\] )?#\d+.*'

def main():
    job_info = get_json(f'{JENKINS_URL}/api/json')
    found_jobs = []
    for job in job_info['jobs']:
        found_jobs.extend(handle_job(job))

    with open(f'{TOOL_TO_FIND}.csv', 'w') as csvfile:
        writer = csv.DictWriter(csvfile, fieldnames=['name', 'url', 'tool', 'last_built', 'in_job_config', 'in_last_build_output'])
        writer.writeheader()
        for job in found_jobs:
            writer.writerow(job)


def get_data(url):
    response = get_request(url)
    return response.data.decode('utf8')


def get_json(url):
    response = get_request(url)
    return response.json()


def get_request(url):
    headers = urllib3.make_headers(basic_auth=f'{JENKINS_USER}:{JENKINS_API_TOKEN}')
    return urllib3.request('GET', url, headers=headers, timeout=Timeout(connect=5.0, read=30.0))


def handle_job(job):
    found_jobs = []

    job_type = job['_class']
    if job_type == 'org.jenkinsci.plugins.workflow.multibranch.WorkflowMultiBranchProject':
        found_jobs.extend(handle_multibranch_project(job))
    elif job_type == 'jenkins.branch.OrganizationFolder':
        found_jobs.extend(handle_org_folder(job))
    else:
        found_jobs.extend(handle_standalone_project(job))

    return found_jobs


def handle_org_folder(job):
    found_jobs = []

    print(f'Processing org folder {job["name"]}')
    job_info = get_json(f'{job["url"]}/api/json')
    for job in job_info['jobs']:
        found_jobs.extend(handle_job(job))

    return found_jobs


def handle_multibranch_project(job):
    print(f'Processing multibranch project {job["name"]}')
    job_info = get_json(f'{job["url"]}/api/json')
    for job in job_info['jobs']:
        if job['name'] == 'master':
            in_last_build_output = check_last_build(job)
            if in_last_build_output:
                return [{
                    'name': parse_name(job['url']),
                    'url': job['url'],
                    'tool': TOOL_TO_FIND,
                    'last_built': get_last_built(job),
                    'in_job_config': False,
                    'in_last_build_output': True
                }]

    return []


def handle_standalone_project(job):
    print(f'Processing standalone job {job["name"]}')
    job_config = get_data(f'{job["url"]}/config.xml')
    in_job_config = False
    if TOOL_TO_FIND in job_config:
        in_job_config = True

    in_last_build_output = check_last_build(job)

    if in_job_config or in_last_build_output:
        return [{
            'name': parse_name(job['url']),
            'url': job['url'],
            'tool': TOOL_TO_FIND,
            'last_built': get_last_built(job),
            'in_job_config': in_job_config,
            'in_last_build_output': in_last_build_output
        }]
    else:
        return []


def check_last_build(job):
    job_info = get_json(f'{job["url"]}/api/json')
    last_successful_build = job_info['lastSuccessfulBuild']

    if last_successful_build:
        try:
            last_build_output = get_data(f'{last_successful_build["url"]}/consoleText')

            for line in last_build_output.splitlines():
                # Only consider lines that have a space after the tool name, to filter out references which are not actually invocations of the tool
                # Also filter out lines that look like Docker build output
                if f'{TOOL_TO_FIND} ' in line and not re.match(DOCKER_BUILD_REGEX, line):
                    return True
        except Exception as e:
            print(f'Error processing build {last_successful_build["url"]}: {e}')

    return False


def get_last_built(job):
    job_info = get_json(f'{job["url"]}/api/json')
    last_build = job_info['lastBuild']

    if last_build:
        last_build_info = get_json(f'{last_build["url"]}/api/json')
        timestamp = last_build_info['timestamp']
        return datetime.fromtimestamp(timestamp / 1000).strftime('%Y-%m-%d %H:%M:%S')

    return 'N/A'


def parse_name(url):
    name = url.removeprefix(f'{JENKINS_URL}/job/')
    name = name.replace('/job', '')
    name = name.removesuffix('/')
    return name


if __name__ == '__main__':
    main()