"""A repeatable process for measuring analytics latency.

Process -
1. Create new tab in Google Chrome browser.
2. Open DevTools to track this page.
3. Enable Preserve log option if you want to save data
    for several pages in one file.
4. Start recording in DevTools.
5. Load page/pages using links.
    https://workstation.theorchard.com/analytics/engagement
    https://workstation.theorchard.com/analytics/audience
    https://workstation.theorchard.com/analytics/index
    https://workstation.theorchard.com/analytics/overview
    https://workstation.qaorch.com/analytics/comparison?ids=USJMZ1800051&ids=USJMZ1800057

    This is very important especially for getting correct results from
    one file for several pages.
6. Click Export Har in DevTools and save a file.
7. Launch script python measure.py path --result filename.csv
    path - can be directory with multiple .har files
    --result filename.csv - The name of the result file(default result.csv).

An example of result file
https://docs.google.com/spreadsheets/d/14eit5nCMqQ-siMn2F6srV0t5-EOQAEjtbQ0_A_jyUd4/edit#gid=0
"""
import argparse
from collections import defaultdict
from collections import namedtuple
import csv
import json
import os


fieldnames = [
    'filename', 'date', 'page', 'type',  'url', 'time',
    'onContentLoad', 'onLoad']

RequestInfo = namedtuple('RequestInfo', 'type url time')


def get_request_type(url):
    """Analyze url and return type of request."""
    if '/images/'in url or url.endswith('js') or url.endswith('css'):
        return 'frontend'
    elif url.startswith('https://workstation.theorchard.com'):
        return 'api'
    return 'other'


def parse_file(filename, pages):
    """Analyze .har file with json data from DevTools."""
    with open(filename) as json_file:
        try:
            data = json.load(json_file)
            # save overall information about each page in the file
            for page in data['log']['pages']:
                pages[page['id']] = {
                    'title': page['title'],
                    'onContentLoad': page['pageTimings']['onContentLoad'],
                    'onLoad': page['pageTimings']['onLoad'],
                    'date': page['startedDateTime'],
                    'requests': []}

            # save information about each request in the file
            for d in data['log']['entries']:
                r = RequestInfo(
                    type=get_request_type(d['request']['url']),
                    url=d['request']['url'],
                    time=d['time'])
                # relate a request with particular page
                try:
                    pages[d['pageref']]['requests'].append(r)
                except KeyError:
                    print('url {} doesn\'t have pageref'.format(r.url))
            print('File {file} was successfully loaded')
        except Exception as err:
            print('An error occurred while loading {file} {error}'.format(
                file=filename, error=err))


def parse_data(path):
    """Parse one file or all files in the directory."""
    result = defaultdict(dict)
    # try to analyze all files in the directory if path is a folder
    if os.path.isdir(path):
        for file in os.listdir(path):
            parse_file(os.path.join(path, file), result[file])
    elif os.path.isfile(path):
        filename = os.path.basename(path)
        parse_file(path, result[filename])
    return result


def write_results(filename, pages):
    """Save results in csv file."""
    with open(filename, 'w', newline='') as csvfile:

        writer = csv.writer(csvfile)
        writer.writerow(fieldnames)

        for filename, data in pages.items():
            for page, info in data.items():
                for r in sorted(
                        info['requests'], key=lambda x: (x.type, x.url)):
                    if r.url == info['title']:
                        onContentLoad = info['onContentLoad']
                        onLoad = info['onLoad']
                    else:
                        onContentLoad, onLoad = None, None
                    row = [
                        filename, info['date'], info['title'],
                        r.type, r.url, r.time, onContentLoad, onLoad]
                    writer.writerow(row)


if __name__ == '__main__':
    parser = argparse.ArgumentParser(
        description='Collect information about latency')

    parser.add_argument(
        'path', help='name of the directory or filename with data to analyze')

    parser.add_argument(
        '--result', help='The result file name', default='result.csv')

    args = parser.parse_args()

    pages = parse_data(args.path)
    write_results(args.result, pages)