#!/usr/bin/env python

""" Scrape a recipe, convert it to Markdown and store it in a Zettelkasten.

    A free recipe-box.

    1. https://obsidian.md/
    2. https://www.ourstate.com/a-kitchens-riches/
"""
import json
import optparse
import os
import requests
import sys

from recipe_scrapers import scrape_me, WebsiteNotImplementedError, SCRAPERS


ROOT = os.path.dirname(os.path.abspath(__file__))


def ensure_directory_exists(path, expand_user=True, file=False):
    """ Create a directory if it doesn't exists.

        Expanding '~' to the user's home directory on POSIX systems.
    """
    if expand_user:
        path = os.path.expanduser(path)

    if file:
        directory = os.path.dirname(path)
    else:
        directory = path

    if not os.path.exists(directory) and directory:
        try:
            os.makedirs(directory)
        except OSError as e:
            # A parallel process created the directory after the existence check.
            pass

    return(path)


def valid_filename(directory, filename=None, ascii=False):
    """ Return a valid "new" filename in a directory, given a filename/directory=path to test.

        Deal with duplicate filenames.
    """
    def test_filename(filename, count):
        """ Filename to test for existence.
        """
        fn, ext = os.path.splitext(filename)
        return fn + '({})'.format(count) + ext

    return_path = filename is None

    # Directory is a path.
    if filename is None:
        filename = os.path.basename(directory)
        directory = os.path.dirname(directory)

    # if ascii:
    #     filename = unidecode(unicode(filename))
    #     filename = ' '.join(filename.splitlines()).strip()
    #     filename = filename.decode('ascii', 'ignore')

    # Allow for directories.
    items = {item: True for item in os.listdir(directory)}
    if filename in items:
        count = 1
        while test_filename(filename, count) in items:
            count += 1
        if return_path:
            return os.path.join(directory, test_filename(filename, count))
        return test_filename(filename, count)
    else:
        if return_path:
            return os.path.join(directory, filename)
        return filename


if __name__ == '__main__':

    parser = optparse.OptionParser('%prog url')

    parser.add_option('-l',
                      dest='list',
                      action='store_true',
                      default=False,
                      help='list all available sites')

    options, args = parser.parse_args()

    if options.list:
        for host in sorted(SCRAPERS):
            print(host)
        sys.exit()

    config_path = os.path.join(ROOT, 'recipe_box.json')
    if not os.path.exists(config_path):
        config = {'recipe_box': '~/recipe_box/'}
        with open(config_path, 'w') as f:
            json.dump(config, f)
    else:
        with open(config_path, 'r') as f:
            config = json.load(f)

    for url in args:
        try:
            scraper = scrape_me(url)
        except WebsiteNotImplementedError:
            print('No scraper defined for {url}'.format(url=url))
            print('It is recommended you add it to recipe-scrapers site, that way everybody gains from the effort.')
            print('https://github.com/hhursev/recipe-scrapers#if-you-want-a-scraper-for-a-new-site-added')
            print('')
            print('Once someone has added the new scraper:')
            print('pip install --upgrade recipe-scrapers')
        else:
            recipe_box = ensure_directory_exists(config['recipe_box'])
            media = ensure_directory_exists(os.path.join(config['recipe_box'], 'media'))

            prefix = scraper.title().lower()
            path = os.path.join(recipe_box, prefix + '.md')
            path = valid_filename(path)
            recipe = open(path, 'w')

            try:
                response = requests.get(scraper.image())
            except:
                filename = None
            else:
                # Not sure about image urls without filename extensions, might need python-magic.
                # Also, os.path.splitext(url), probably not a good idea. ;)
                filename = os.path.splitext(os.path.basename(path))[0] + os.path.splitext(scraper.image())[1]
                image = open(os.path.join(media, filename), 'wb')
                image.write(response.content)
                image.close()

            recipe.write('# {title}\n'.format(title=scraper.title()))
            if filename:
                recipe.write('![[{filename}]]\n'.format(filename=filename))
            recipe.write('\n')
            recipe.write('## Notes\n')
            recipe.write('\n')
            recipe.write('## Metadata\n')
            recipe.write('Yields: {yields}\n'.format(yields=scraper.yields()))
            recipe.write('Total Time: {total_time}\n'.format(total_time=scraper.total_time()))
            recipe.write('\n')
            recipe.write('## Ingredients\n')
            for ingredient in scraper.ingredients():
                recipe.write('* {ingredient}\n'.format(ingredient=ingredient))

            recipe.write('\n')
            recipe.write('## Instructions\n')
            for instruction in scraper.instructions().split('\n'):
                if instruction.strip()[0].isdigit():
                    recipe.write('{instruction}\n'.format(instruction=instruction))
                else:
                    recipe.write('1. {instruction}\n'.format(instruction=instruction))

            recipe.write('\n')
            recipe.write('[{url}]({url})\n'.format(url=url))