From 536e629356f990053e84bcb853820adad1d2dc51 Mon Sep 17 00:00:00 2001 From: Kovid Goyal Date: Mon, 4 Jan 2010 22:34:33 -0700 Subject: [PATCH] Improved recipe for CNN --- resources/recipes/cnn.recipe | 54 +++++++++++++++++------------------- src/calibre/gui2/__init__.py | 19 +++++++------ 2 files changed, 36 insertions(+), 37 deletions(-) diff --git a/resources/recipes/cnn.recipe b/resources/recipes/cnn.recipe index 21c4139d6e..8c3cfa6de8 100644 --- a/resources/recipes/cnn.recipe +++ b/resources/recipes/cnn.recipe @@ -4,13 +4,14 @@ __copyright__ = '2008, Kovid Goyal ' Profile to download CNN ''' from calibre.web.feeds.news import BasicNewsRecipe +from calibre.ebooks.BeautifulSoup import BeautifulSoup class CNN(BasicNewsRecipe): title = 'CNN' description = 'Global news' timefmt = ' [%d %b %Y]' - __author__ = 'Kovid Goyal and Sujata Raman' + __author__ = 'Krittika Goyal and Sujata Raman' language = 'en' no_stylesheets = True @@ -18,7 +19,8 @@ class CNN(BasicNewsRecipe): oldest_article = 15 recursions = 1 match_regexps = [r'http://sportsillustrated.cnn.com/.*/[1-9].html'] - + max_articles_per_feed = 25 + extra_css = ''' .cnn_strycntntlft{font-family :Arial,Helvetica,sans-serif;} h2{font-family :Arial,Helvetica,sans-serif; font-size:x-small} @@ -44,7 +46,7 @@ class CNN(BasicNewsRecipe): .cnnInlineT1Caption{font-family :Arial,Helvetica,sans-serif; font-size:x-small;font-weight:bold;} .cnnInlineT1Credit{font-family :Arial,Helvetica,sans-serif; font-size:x-small;color:#333333;} .col10{color:#5A637E;} - .cnnInlineRailBulletList{color:black;} + .cnnInlineRailBulletList{color:black;} .cnnLine0{font-family :Arial,Helvetica,sans-serif; color:#666666;font-weight:bold;} .cnnTimeStamp{font-family :Arial,Helvetica,sans-serif; font-size:x-small;color:#333333;} .galleryhedDek{font-family :Arial,Helvetica,sans-serif; font-size:x-small;color:#575757;} @@ -55,31 +57,22 @@ class CNN(BasicNewsRecipe): .captionname{font-family :Arial,Helvetica,sans-serif; font-size:x-small;color:#575757;} inStoryIE{{font-family :Arial,Helvetica,sans-serif; font-size:x-small;} ''' - keep_only_tags = [ - dict(name='div', attrs={'class':["cnnWCBoxContent","cnnContent","cnnMainBodySecs","cnn_storyarea","cnn_strycntntlft","cnn_strytmstmp"]}), - dict(name='div', attrs={'id':["contentBody","content"]}), - dict(name='td', attrs={'id':["cnnRecapStory"]}), - dict(name='td', attrs={'class':["cnnBigSideLeft"]}), - ] - remove_tags = [ - dict(name='div', attrs={'class':["cnnInlineRailSelectList","cnn_strycntntrgt","cnn_strybtntools","cnn_strylctcntr cnn_strylctcqrelt","cnn_strylceclbtn","cnn_stryftsbttm","cnn_strybtmmorebx","storyLink","article-tools clearfix","widget video related-video vList","cnnFooterBox","scrollArrows","boxHeading","cnnInlineMailbag","mainCol_lastBlock","cnn_bookmarks","cnnFooterBox","cnnEndOfStory","cnnInlineSL","cnnStoryHighlights","cnnFooterClick","cnnSnapShotHeader","cnnStoryToolsFooter","cnnWsnr","cnnUGCBox","cnnTopNewsModule","cnnStoryElementBox","cnnStoryPhotoBoxNavigation"]}), - dict(name='span', attrs={'class':["cnnEmbeddedMosLnk"]}), - dict(name='div', attrs={'id':["cnnIncldHlder","articleCommentsContainer","featuredContent","superstarsWidget","shareMenuContainer","rssMenuContainer","storyBrandingBanner","cnnRightCol","siteFeatures","quigo628","rightColumn","clickIncludeBox","cnnHeaderRightCol","cnnSCFontLabel","cnnSnapShotBottomRight","cnnSCFontButtons","rightColumn"]}), - dict(name='p', attrs={'class':["cnnTopics"]}), - dict(name='td', attrs={'class':["cnnRightRail"]}), - dict(name='table', attrs={'class':["cnnTMbox"]}), - dict(name='ul', attrs={'id':["cnnTopNav","cnnBotNav","cnnSBNav"]}), - dict(name='div', attrs={'id':["cnn_ftrcntnt"]}) - ] - # def print_version(self, url): - # return 'http://www.printthis.clickability.com/pt/printThis?clickMap=printThis&fb=Y&url=' + url + #remove_tags_before = dict(name='h1', attrs={'class':'heading'}) + #remove_tags_after = dict(name='td', attrs={'class':'newptool1'}) + remove_tags = [ + dict(name='iframe'), + dict(name='div', attrs={'class':['cnnEndOfStory', 'cnnShareThisItem', 'cnn_strylctcntr cnn_strylctcqrelt', 'cnnShareBoxContent', 'cnn_strybtmcntnt', 'cnn_strycntntrgt']}), + dict(name='div', attrs={'id':['IEContainer', 'clickIncludeBox']}), + #dict(name='ul', attrs={'class':'article-tools'}), + #dict(name='ul', attrs={'class':'articleTools'}), + ] feeds = [ ('Top News', 'http://rss.cnn.com/rss/cnn_topstories.rss'), ('World', 'http://rss.cnn.com/rss/cnn_world.rss'), ('U.S.', 'http://rss.cnn.com/rss/cnn_us.rss'), - ('Sports', 'http://rss.cnn.com/rss/si_topstories.rss'), + #('Sports', 'http://rss.cnn.com/rss/si_topstories.rss'), ('Business', 'http://rss.cnn.com/rss/money_latest.rss'), ('Politics', 'http://rss.cnn.com/rss/cnn_allpolitics.rss'), ('Law', 'http://rss.cnn.com/rss/cnn_law.rss'), @@ -91,10 +84,15 @@ class CNN(BasicNewsRecipe): ('Offbeat', 'http://rss.cnn.com/rss/cnn_offbeat.rss'), ('Most Popular', 'http://rss.cnn.com/rss/cnn_mostpopular.rss') ] - def preprocess_html(self, soup): - - for tag in soup.findAll(name=['ul','li']): - tag.name = 'div' - - return soup + story = soup.find(name='div', attrs={'class':'cnnBody_Left'}) + if story is None: + story = soup.find(name='div', attrs={'id':'cnnContentContainer'}) + soup = BeautifulSoup('t') + body = soup.find(name='body') + body.insert(0, story) + else: + soup = BeautifulSoup('t') + body = soup.find(name='body') + body.insert(0, story) + return soup diff --git a/src/calibre/gui2/__init__.py b/src/calibre/gui2/__init__.py index 5d67e3a1b1..82f65f8c42 100644 --- a/src/calibre/gui2/__init__.py +++ b/src/calibre/gui2/__init__.py @@ -529,15 +529,16 @@ class Application(QApplication): self.load_translations() qt_app = self - self.setStyleSheet(''' - QToolTip { - border: 2px solid black; - padding: 5px; - border-radius: 10px; - opacity: 200; - background-color: white; - } - ''') + if islinux: + self.setStyleSheet(''' + QToolTip { + border: 2px solid black; + padding: 5px; + border-radius: 10px; + opacity: 200; + background-color: #e1e1ff; + } + ''') def load_translations(self):