from __future__ import unicode_literals

import re

from .common import InfoExtractor


class InstagramIE(InfoExtractor):
    _VALID_URL = r'http://instagram\.com/p/(?P<id>.*?)/'
    _TEST = {
        'url': 'http://instagram.com/p/aye83DjauH/?foo=bar#abc',
        'md5': '0d2da106a9d2631273e192b372806516',
        'info_dict': {
            'id': 'aye83DjauH',
            'ext': 'mp4',
            'uploader_id': 'naomipq',
            'title': 'Video by naomipq',
            'description': 'md5:1f17f0ab29bd6fe2bfad705f58de3cb8',
        }
    }

    def _real_extract(self, url):
        mobj = re.match(self._VALID_URL, url)
        video_id = mobj.group('id')
        webpage = self._download_webpage(url, video_id)
        uploader_id = self._search_regex(r'"owner":{"username":"(.+?)"',
            webpage, 'uploader id', fatal=False)
        desc = self._search_regex(r'"caption":"(.*?)"', webpage, 'description',
            fatal=False)

        return {
            'id': video_id,
            'url': self._og_search_video_url(webpage, secure=False),
            'ext': 'mp4',
            'title': 'Video by %s' % uploader_id,
            'thumbnail': self._og_search_thumbnail(webpage),
            'uploader_id': uploader_id,
            'description': desc,
        }