aboutsummaryrefslogtreecommitdiff
path: root/pyaggr3g470r/lib/crawler.py
blob: 339c4b12f8f4164b26bbff169b456b467e6d95d4 (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
"""
Here's a sum up on how it works :

CrawlerScheduler.run
    will retreive a list of feeds to be refreshed and pass result to
CrawlerScheduler.callback
    which will retreive each feed and treat result with
FeedCrawler.callback
    which will interprete the result (status_code, etag) collect ids
    and match them agaisnt pyagg which will cause
PyAggUpdater.callback
    to create the missing entries
"""

import time
import conf
import json
import logging
import feedparser
import dateutil.parser
from hashlib import md5
from functools import wraps
from datetime import datetime
from concurrent.futures import ThreadPoolExecutor
from requests_futures.sessions import FuturesSession
from pyaggr3g470r.lib.utils import default_handler

logger = logging.getLogger(__name__)
logging.captureWarnings(True)
API_ROOT = "api/v2.0/"


def to_hash(text):
    return md5(text.encode('utf8')).hexdigest()


def extract_id(entry, keys=[('link', 'link'),
                            ('published', 'retrieved_date'),
                            ('updated', 'retrieved_date')], force_id=False):
    """For a given entry will return a dict that allows to identify it. The
    dict will be constructed on the uid of the entry. if that identifier is
    absent, the dict will be constructed upon the values of "keys".
    """
    entry_id = entry.get('entry_id') or entry.get('id')
    if entry_id:
        return {'entry_id': entry_id}
    if not entry_id and force_id:
        entry_id = to_hash("".join(entry[entry_key] for _, entry_key in keys
                                   if entry_key in entry).encode('utf8'))
    else:
        ids = {}
        for entry_key, pyagg_key in keys:
            if entry_key in entry and pyagg_key not in ids:
                ids[pyagg_key] = entry[entry_key]
                if 'date' in pyagg_key:
                    ids[pyagg_key] = dateutil.parser.parse(ids[pyagg_key])\
                                                    .isoformat()
        return ids


class AbstractCrawler:
    __session__ = None
    __counter__ = 0

    def __init__(self, auth):
        AbstractCrawler.__counter__ += 1
        self.auth = auth
        self.session = self.get_session()
        self.url = conf.PLATFORM_URL

    @classmethod
    def get_session(cls):
        """methods that allows us to treat session as a singleton"""
        if cls.__session__ is None:
            cls.__session__ = FuturesSession(
                    executor=ThreadPoolExecutor(max_workers=conf.NB_WORKER))
            cls.__session__.verify = False
        return cls.__session__

    @classmethod
    def count_on_me(cls, func):
        """A basic decorator which will count +1 at the begining of a call
        and -1 at the end. It kinda allows us to wait for the __counter__ value
        to be 0, meaning nothing is done anymore."""
        @wraps(func)
        def wrapper(*args, **kwargs):
            cls.__counter__ += 1
            try:
                return func(*args, **kwargs)
            except:
                logger.exception('an error occured while %r', func)
            finally:
                cls.__counter__ -= 1
        return wrapper

    @classmethod
    def get_counter_callback(cls):
        cls.__counter__ += 1

        def debump(*args, **kwargs):
            cls.__counter__ -= 1
        return debump

    def query_pyagg(self, method, urn, data=None):
        """A wrapper for internal call, method should be ones you can find
        on requests (header, post, get, options, ...), urn the distant
        resources you want to access on pyagg, and data, the data you wanna
        transmit."""
        if data is None:
            data = {}
        method = getattr(self.session, method)
        return method("%s%s%s" % (self.url, API_ROOT, urn),
                      auth=self.auth, data=json.dumps(data,
                                                      default=default_handler),
                      headers={'Content-Type': 'application/json',
                               'User-Agent': 'pyaggr3g470r'})

    @classmethod
    def wait(cls, max_wait=600):
        "See count_on_me, that method will just wait for the counter to be 0"
        time.sleep(1)
        second_waited = 1
        while cls.__counter__:
            if second_waited > max_wait:
                logger.warn('Exiting after %d seconds, counter at %d',
                            max_wait, cls.__counter__)
                break
            time.sleep(1)
            second_waited += 1


class PyAggUpdater(AbstractCrawler):

    def __init__(self, feed, entries, headers, parsed_feed, auth):
        self.feed = feed
        self.entries = entries
        self.headers = headers
        self.parsed_feed = parsed_feed.get('feed', {})
        super(PyAggUpdater, self).__init__(auth)

    def to_article(self, entry):
        "Safe method to transorm a feedparser entry into an article"
        date = datetime.now()

        for date_key in ('published', 'updated'):
            if entry.get(date_key):
                try:
                    date = dateutil.parser.parse(entry[date_key])
                except Exception:
                    pass
                else:
                    break
        content = ''
        if entry.get('content'):
            content = entry['content'][0]['value']
        elif entry.get('summary'):
            content = entry['summary']

        return {'feed_id': self.feed['id'],
                'user_id': self.feed['user_id'],
                'entry_id': extract_id(entry).get('entry_id', None),
                'link': entry.get('link', self.feed['site_link']),
                'title': entry.get('title', 'No title'),
                'readed': False, 'like': False,
                'content': content,
                'retrieved_date': date.isoformat(),
                'date': date.isoformat()}

    @AbstractCrawler.count_on_me
    def callback(self, response):
        """Will process the result from the challenge, creating missing article
        and updating the feed"""
        AbstractCrawler.__counter__ -= 1
        results = response.result().json()
        logger.debug('%r %r - %d entries were not matched and will be created',
                     self.feed['id'], self.feed['title'], len(results))
        for id_to_create in results:
            entry = self.to_article(
                    self.entries[tuple(sorted(id_to_create.items()))])
            logger.warn('%r %r - creating %r for %r - %r', self.feed['id'],
                        self.feed['title'], entry['title'], entry['user_id'],
                        id_to_create)
            self.query_pyagg('post', 'article', entry)

        logger.debug('%r %r - updating feed etag %r last_mod %r',
                     self.feed['id'], self.feed['title'],
                     self.headers.get('etag', ''),
                     self.headers.get('last-modified', ''))

        dico = {'error_count': 0, 'last_error': None,
                'etag': self.headers.get('etag', ''),
                'last_modified': self.headers.get('last-modified', ''),
                'site_link': self.parsed_feed.get('link')}
        if not self.feed.get('title'):
            dico['title'] = self.parsed_feed.get('title', '')
        logger.info('%r %r - pushing feed attrs %r',
                    self.feed['id'], self.feed['title'],
                    {key: "%s -> %s" % (dico[key], self.feed.get(key))
                     for key in dico if dico[key] != self.feed.get(key)})
        if any([dico[key] != self.feed.get(key) for key in dico]):
            future = self.query_pyagg('put', 'feed/%d' % self.feed['id'], dico)
            future.add_done_callback(self.get_counter_callback())


class FeedCrawler(AbstractCrawler):

    def __init__(self, feed, auth):
        self.feed = feed
        super(FeedCrawler, self).__init__(auth)

    def clean_feed(self):
        """Will reset the errors counters on a feed that have known errors"""
        if self.feed.get('error_count') or self.feed.get('last_error'):
            future = self.query_pyagg('put', 'feed/%d' % self.feed['id'],
                                      {'error_count': 0, 'last_error': ''})
            future.add_done_callback(self.get_counter_callback())

    @AbstractCrawler.count_on_me
    def callback(self, response):
        """will fetch the feed and interprete results (304, etag) or will
        challenge pyagg to compare gotten entries with existing ones"""
        AbstractCrawler.__counter__ -= 1
        try:
            response = response.result()
            response.raise_for_status()
        except Exception as error:
            error_count = self.feed['error_count'] + 1
            logger.warn('%r %r - an error occured while fetching feed; bumping'
                        ' error count to %r', self.feed['id'],
                        self.feed['title'], error_count)
            future = self.query_pyagg('put', 'feed/%d' % self.feed['id'],
                                      {'error_count': error_count,
                                       'last_error': str(error)})
            future.add_done_callback(self.get_counter_callback())
            return

        if response.status_code == 304:
            logger.info("%r %r - feed responded with 304",
                        self.feed['id'], self.feed['title'])
            self.clean_feed()
            return
        if 'etag' not in response.headers:
            logger.debug('%r %r - manually generating etag',
                         self.feed['id'], self.feed['title'])
            response.headers['etag'] = 'pyagg/"%s"' % to_hash(response.text)
        if response.headers['etag'] and self.feed['etag'] \
                and response.headers['etag'] == self.feed['etag']:
            if 'pyagg' in self.feed['etag']:
                logger.info("%r %r - calculated hash matches (%d)",
                            self.feed['id'], self.feed['title'],
                            response.status_code)
            else:
                logger.info("%r %r - feed responded with same etag (%d)",
                            self.feed['id'], self.feed['title'],
                            response.status_code)
            self.clean_feed()
            return
        else:
            logger.debug('%r %r - etag mismatch %r != %r',
                         self.feed['id'], self.feed['title'],
                         response.headers['etag'], self.feed['etag'])
        logger.info('%r %r - cache validation failed, challenging entries',
                    self.feed['id'], self.feed['title'])

        ids, entries = [], {}
        parsed_response = feedparser.parse(response.text)
        for entry in parsed_response['entries']:
            entry_ids = extract_id(entry)
            entry_ids['feed_id'] = self.feed['id']
            entries[tuple(sorted(entry_ids.items()))] = entry
            ids.append(entry_ids)
        logger.debug('%r %r - found %d entries %r',
                     self.feed['id'], self.feed['title'], len(ids), ids)
        future = self.query_pyagg('get', 'articles/challenge', {'ids': ids})
        updater = PyAggUpdater(self.feed, entries, response.headers,
                               parsed_response, self.auth)
        future.add_done_callback(updater.callback)


class CrawlerScheduler(AbstractCrawler):

    def __init__(self, username, password):
        self.auth = (username, password)
        super(CrawlerScheduler, self).__init__(self.auth)
        AbstractCrawler.__counter__ = 0

    def prepare_headers(self, feed):
        """For a known feed, will construct some header dictionnary"""
        headers = {'User-Agent': 'pyaggr3g470r/crawler'}
        if feed.get('last_modified'):
            headers['If-Modified-Since'] = feed['last_modified']
        if feed.get('etag') and 'pyagg' not in feed['etag']:
            headers['If-None-Match'] = feed['etag']
        logger.debug('%r %r - calculated headers %r',
                     feed['id'], feed['title'], headers)
        return headers

    @AbstractCrawler.count_on_me
    def callback(self, response):
        """processes feeds that need to be fetched"""
        AbstractCrawler.__counter__ -= 1
        response = response.result()
        response.raise_for_status()
        feeds = response.json()
        logger.debug('%d to fetch %r', len(feeds), feeds)
        for feed in feeds:
            logger.debug('%r %r - fetching resources',
                         feed['id'], feed['title'])
            future = self.session.get(feed['link'],
                                      headers=self.prepare_headers(feed))
            future.add_done_callback(FeedCrawler(feed, self.auth).callback)

    @AbstractCrawler.count_on_me
    def run(self, **kwargs):
        """entry point, will retreive feeds to be fetch
        and launch the whole thing"""
        logger.debug('retreving fetchable feed')
        future = self.query_pyagg('get', 'feeds/fetchable', kwargs)
        AbstractCrawler.__counter__ += 1
        future.add_done_callback(self.callback)
bgstack15