From 47159b1e3818d1a5f1f4eeca15356e99a12c9922 Mon Sep 17 00:00:00 2001
From: Cédric Bonhomme <cedric@cedricbonhomme.org>
Date: Tue, 9 Feb 2016 23:05:50 +0100
Subject: The classic crawler is now able to update an article. Only content or
 title are updatable, as discussed in #27. Need to implement the notification
 system with the boolean when the content of an article has been updated.

---
 src/crawler/classic_crawler.py | 20 +++++++++++++++++---
 src/web/lib/article_utils.py   | 14 +++++++++-----
 2 files changed, 26 insertions(+), 8 deletions(-)

(limited to 'src')

diff --git a/src/crawler/classic_crawler.py b/src/crawler/classic_crawler.py
index 1bc37a47..b21a7859 100644
--- a/src/crawler/classic_crawler.py
+++ b/src/crawler/classic_crawler.py
@@ -38,7 +38,8 @@ from bootstrap import db
 from web.models import User
 from web.controllers import FeedController, ArticleController
 from web.lib.feed_utils import construct_feed_from, is_parsing_ok
-from web.lib.article_utils import construct_article, extract_id
+from web.lib.article_utils import construct_article, extract_id, \
+                                    get_article_content
 
 logger = logging.getLogger(__name__)
 
@@ -117,11 +118,24 @@ async def insert_database(user, feed):
     new_articles = []
     art_contr = ArticleController(user.id)
     for article in articles:
-        exist = art_contr.read(feed_id=feed.id,
-                        **extract_id(article)).count() != 0
+        existing_article_req = art_contr.read(feed_id=feed.id,
+                        **extract_id(article))
+        exist = existing_article_req.count() != 0
         if exist:
+            existing_article = existing_article_req.first()
+            is_updated = False
             logger.debug("Article %r (%r) already in the database.",
                          article['title'], article['link'])
+            content = get_article_content(article)
+            if existing_article.title != article['title']:
+                existing_article.title = article['title']
+                is_updated = True
+            if existing_article.content != content:
+                existing_article.content = content
+                existing_article.readed = False
+                is_updated = True
+            if is_updated:
+                art_contr.update({'entry_id': existing_article.entry_id}, existing_article.dump())
             continue
         article = construct_article(article, feed)
         try:
diff --git a/src/web/lib/article_utils.py b/src/web/lib/article_utils.py
index 02ca2cd1..328c27ab 100644
--- a/src/web/lib/article_utils.py
+++ b/src/web/lib/article_utils.py
@@ -46,11 +46,7 @@ def construct_article(entry, feed):
                 pass
             else:
                 break
-    content = ''
-    if entry.get('content'):
-        content = entry['content'][0]['value']
-    elif entry.get('summary'):
-        content = entry['summary']
+    content = get_article_content(entry)
 
     article_link = entry.get('link')
     if conf.RESOLVE_ARTICLE_URL and article_link:
@@ -72,3 +68,11 @@ def construct_article(entry, feed):
             'content': content,
             'retrieved_date': now.isoformat(),
             'date': (date or now).isoformat()}
+
+def get_article_content(entry):
+    content = ''
+    if entry.get('content'):
+        content = entry['content'][0]['value']
+    elif entry.get('summary'):
+        content = entry['summary']
+    return content
-- 
cgit