various improvements to the crawler (better use of coroutines, test if an article should be updated). tags are now retrieved for the k-means clustering (previously achived with the content of articles)

4 files changed, 192 insertions, 67 deletions

diff --git a/src/web/lib/article_utils.py b/src/web/lib/article_utils.py
index f79c1b30..81a0b145 100644
--- a/src/web/lib/article_utils.py
+++ b/src/web/lib/article_utils.py
@@ -1,74 +1,60 @@
+import html
 import logging
-import dateutil.parser
+import re
 from datetime import datetime, timezone
+from enum import Enum
+from urllib.parse import SplitResult, urlsplit, urlunsplit
+
+import dateutil.parser
+from bs4 import BeautifulSoup, SoupStrainer
+from requests.exceptions import MissingSchema
 
 import conf
-from web.lib.utils import to_hash
+from web.lib.utils import jarr_get
 
 logger = logging.getLogger(__name__)
+PROCESSED_DATE_KEYS = {'published', 'created', 'updated'}
 
 
-def extract_id(entry, keys=[('link', 'link'), ('published', 'date'),
-                            ('updated', '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:
-        return 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:
-                    try:
-                        ids[pyagg_key] = dateutil.parser.parse(ids[pyagg_key])\
-                                                    .isoformat()
-                    except ValueError as e:
-                        logger.exception("extract_id: " + str(e))
-                        ids[pyagg_key] =  datetime.now().isoformat()
-        return ids
-
-
-def construct_article(entry, feed):
-    if hasattr(feed, 'dump'):  # this way can be a sqlalchemy obj or a dict
-        feed = feed.dump()
-    "Safe method to transorm a feedparser entry into an article"
-    now = datetime.now()
-    date = None
-    for date_key in ('published', 'created', 'date'):
-        if entry.get(date_key):
-            try:
-                date = dateutil.parser.parse(entry[date_key])\
-                        .astimezone(timezone.utc)
-            except Exception as e:
-                logger.exception(str(e))
-            else:
-                break
+def extract_id(entry):
+    """ extract a value from an entry that will identify it among the other of
+    that feed"""
+    return entry.get('entry_id') or entry.get('id') or entry['link']
 
-    updated_date = None
-    try:
-        updated_date = dateutil.parser.parse(entry['updated'])
-    except Exception:
-        pass
-    content = get_article_content(entry)
-    article_link = entry.get('link')
 
-    return {'feed_id': feed['id'],
-            'user_id': feed['user_id'],
-            'entry_id': extract_id(entry).get('entry_id', None),
-            'link': entry.get('link', feed['site_link']),
-            'title': entry.get('title', 'No title'),
-            'readed': False, 'like': False,
-            'content': content,
-            'retrieved_date': now,
-            'date': date or now,
-            'updated_date': updated_date or date or now}
+async def construct_article(entry, feed, fields=None, fetch=True):
+    "Safe method to transorm a feedparser entry into an article"
+    now = datetime.utcnow()
+    article = {}
+    def push_in_article(key, value):
+        if not fields or key in fields:
+            article[key] = value
+    push_in_article('feed_id', feed.id)
+    push_in_article('user_id', feed.user_id)
+    push_in_article('entry_id', extract_id(entry))
+    push_in_article('retrieved_date', now)
+    if not fields or 'date' in fields:
+        for date_key in PROCESSED_DATE_KEYS:
+            if entry.get(date_key):
+                try:
+                    article['date'] = dateutil.parser.parse(entry[date_key])\
+                            .astimezone(timezone.utc)
+                except Exception as e:
+                    logger.exception(e)
+                else:
+                    break
+    push_in_article('content', get_article_content(entry))
+    if fields is None or {'link', 'title'}.intersection(fields):
+        link, title = await get_article_details(entry, fetch)
+        push_in_article('link', link)
+        push_in_article('title', title)
+        if 'content' in article:
+            #push_in_article('content', clean_urls(article['content'], link))
+            push_in_article('content', article['content'])
+    push_in_article('tags', {tag.get('term').strip()
+                             for tag in entry.get('tags', [])})
+    return article
+
 
 def get_article_content(entry):
     content = ''
@@ -77,3 +63,123 @@ def get_article_content(entry):
     elif entry.get('summary'):
         content = entry['summary']
     return content
+
+
+async def get_article_details(entry, fetch=True):
+    article_link = entry.get('link')
+    article_title = html.unescape(entry.get('title', ''))
+    if fetch and conf.CRAWLER_RESOLV and article_link or not article_title:
+        try:
+            # resolves URL behind proxies (like feedproxy.google.com)
+            response = await jarr_get(article_link, timeout=5)
+        except MissingSchema:
+            split, failed = urlsplit(article_link), False
+            for scheme in 'https', 'http':
+                new_link = urlunsplit(SplitResult(scheme, *split[1:]))
+                try:
+                    response = await jarr_get(new_link, timeout=5)
+                except Exception as error:
+                    failed = True
+                    continue
+                failed = False
+                article_link = new_link
+                break
+            if failed:
+                return article_link, article_title or 'No title'
+        except Exception as error:
+            logger.info("Unable to get the real URL of %s. Won't fix "
+                        "link or title. Error: %s", article_link, error)
+            return article_link, article_title or 'No title'
+        article_link = response.url
+        if not article_title:
+            bs_parsed = BeautifulSoup(response.content, 'html.parser',
+                                      parse_only=SoupStrainer('head'))
+            try:
+                article_title = bs_parsed.find_all('title')[0].text
+            except IndexError:  # no title
+                pass
+    return article_link, article_title or 'No title'
+
+
+class FiltersAction(Enum):
+    READ = 'mark as read'
+    LIKED = 'mark as favorite'
+    SKIP = 'skipped'
+
+
+class FiltersType(Enum):
+    REGEX = 'regex'
+    MATCH = 'simple match'
+    EXACT_MATCH = 'exact match'
+    TAG_MATCH = 'tag match'
+    TAG_CONTAINS = 'tag contains'
+
+
+class FiltersTrigger(Enum):
+    MATCH = 'match'
+    NO_MATCH = 'no match'
+
+
+def process_filters(filters, article, only_actions=None):
+    skipped, read, liked = False, None, False
+    filters = filters or []
+    if only_actions is None:
+        only_actions = set(FiltersAction)
+    for filter_ in filters:
+        match = False
+        try:
+            pattern = filter_.get('pattern', '')
+            filter_type = FiltersType(filter_.get('type'))
+            filter_action = FiltersAction(filter_.get('action'))
+            filter_trigger = FiltersTrigger(filter_.get('action on'))
+            if filter_type is not FiltersType.REGEX:
+                pattern = pattern.lower()
+        except ValueError:
+            continue
+        if filter_action not in only_actions:
+            logger.debug('ignoring filter %r' % filter_)
+            continue
+        if filter_action in {FiltersType.REGEX, FiltersType.MATCH,
+                FiltersType.EXACT_MATCH} and 'title' not in article:
+            continue
+        if filter_action in {FiltersType.TAG_MATCH, FiltersType.TAG_CONTAINS} \
+                and 'tags' not in article:
+            continue
+        title = article.get('title', '').lower()
+        tags = [tag.lower() for tag in article.get('tags', [])]
+        if filter_type is FiltersType.REGEX:
+            match = re.match(pattern, title)
+        elif filter_type is FiltersType.MATCH:
+            match = pattern in title
+        elif filter_type is FiltersType.EXACT_MATCH:
+            match = pattern == title
+        elif filter_type is FiltersType.TAG_MATCH:
+            match = pattern in tags
+        elif filter_type is FiltersType.TAG_CONTAINS:
+            match = any(pattern in tag for tag in tags)
+        take_action = match and filter_trigger is FiltersTrigger.MATCH \
+                or not match and filter_trigger is FiltersTrigger.NO_MATCH
+
+        if not take_action:
+            continue
+
+        if filter_action is FiltersAction.READ:
+            read = True
+        elif filter_action is FiltersAction.LIKED:
+            liked = True
+        elif filter_action is FiltersAction.SKIP:
+            skipped = True
+
+    if skipped or read or liked:
+        logger.info("%r applied on %r", filter_action.value,
+                    article.get('link') or article.get('title'))
+    return skipped, read, liked
+
+
+def get_skip_and_ids(entry, feed):
+    entry_ids = construct_article(entry, feed,
+                {'entry_id', 'feed_id', 'user_id'}, fetch=False)
+    skipped, _, _ = process_filters(feed.filters,
+            construct_article(entry, feed, {'title', 'tags'}, fetch=False),
+            {FiltersAction.SKIP})
+    return skipped, entry_ids
diff --git a/src/web/lib/feed_utils.py b/src/web/lib/feed_utils.py
index 9925613f..94ae6e53 100644
--- a/src/web/lib/feed_utils.py
+++ b/src/web/lib/feed_utils.py
@@ -3,7 +3,7 @@ import urllib
 import logging
 import requests
 import feedparser
-from conf import USER_AGENT
+from conf import CRAWLER_USER_AGENT
 from bs4 import BeautifulSoup, SoupStrainer
 
 from web.lib.utils import try_keys, try_get_icon_url, rebuild_url
@@ -32,7 +32,8 @@ def escape_keys(*keys):
 
 @escape_keys('title', 'description')
 def construct_feed_from(url=None, fp_parsed=None, feed=None, query_site=True):
-    requests_kwargs = {'headers': {'User-Agent': USER_AGENT}, 'verify': False}
+    requests_kwargs = {'headers': {'User-Agent': CRAWLER_USER_AGENT},
+                        'verify': False}
     if url is None and fp_parsed is not None:
         url = fp_parsed.get('url')
     if url is not None and fp_parsed is None:
diff --git a/src/web/lib/misc_utils.py b/src/web/lib/misc_utils.py
index 22d52b70..1359c798 100755
--- a/src/web/lib/misc_utils.py
+++ b/src/web/lib/misc_utils.py
@@ -109,6 +109,8 @@ def fetch(id, feed_id=None):
     """
     cmd = [sys.executable, conf.BASE_DIR + '/manager.py', 'fetch_asyncio',
            '--user_id='+str(id)]
+    if feed_id:
+        cmd.append('--feed_id='+str(feed_id))
     return subprocess.Popen(cmd, stdout=subprocess.PIPE)
 
 def history(user_id, year=None, month=None):
diff --git a/src/web/lib/utils.py b/src/web/lib/utils.py
index f2bed3ff..d206b769 100644
--- a/src/web/lib/utils.py
+++ b/src/web/lib/utils.py
@@ -6,6 +6,8 @@ import requests
 from hashlib import md5
 from flask import request, url_for
 
+import conf
+
 logger = logging.getLogger(__name__)
 
 
@@ -46,11 +48,17 @@ def try_get_icon_url(url, *splits):
         if split is None:
             continue
         rb_url = rebuild_url(url, split)
-        response = requests.get(rb_url, verify=False, timeout=10)
+        response = None
         # if html in content-type, we assume it's a fancy 404 page
-        content_type = response.headers.get('content-type', '')
-        if response.ok and 'html' not in content_type and response.content:
-            return response.url
+        try:
+            response = jarr_get(rb_url)
+            content_type = response.headers.get('content-type', '')
+        except Exception:
+            pass
+        else:
+            if response is not None and response.ok \
+                    and 'html' not in content_type and response.content:
+                return response.url
     return None
 
 
@@ -71,3 +79,11 @@ def clear_string(data):
 
 def redirect_url(default='home'):
     return request.args.get('next') or request.referrer or url_for(default)
+
+
+async def jarr_get(url, **kwargs):
+    request_kwargs = {'verify': False, 'allow_redirects': True,
+                      'timeout': conf.CRAWLER_TIMEOUT,
+                      'headers': {'User-Agent': conf.CRAWLER_USER_AGENT}}
+    request_kwargs.update(kwargs)
+    return requests.get(url, **request_kwargs)