Cutlery/calibre
1
0
Fork 0
mirror of https://github.com/kovidgoyal/calibre.git synced 2025-07-09 03:04:10 -04:00
Code Issues Packages Projects Releases Wiki Activity

...

Browse Source
This commit is contained in:
Kovid Goyal 2011-01-18 13:00:20 -07:00
parent 0b5003ab10
commit 54fb874621
1 changed files with 18 additions and 9 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

27
resources/recipes/nytimes_sub.recipe
View File

@ -1,4 +1,5 @@
#!/usr/bin/env python #!/usr/bin/env python
# -*- coding: utf-8 -*-
__license__ = 'GPL v3' __license__ = 'GPL v3'
__copyright__ = '2008, Kovid Goyal <kovid at kovidgoyal.net>' __copyright__ = '2008, Kovid Goyal <kovid at kovidgoyal.net>'
@ -23,6 +24,10 @@ class NYTimes(BasicNewsRecipe):
webEdition = False webEdition = False
oldest_article = 7 oldest_article = 7
# replace paid Kindle Version: the name will be changed to "The New York Times" to cause
# previous paid versions of the new york times to best sent to the back issues folder on the kindle
replaceKindleVersion = False
# includeSections: List of sections to include. If empty, all sections found will be included. # includeSections: List of sections to include. If empty, all sections found will be included.
# Otherwise, only the sections named will be included. For example, # Otherwise, only the sections named will be included. For example,
# #
@ -94,6 +99,10 @@ class NYTimes(BasicNewsRecipe):
title='New York Times (Web)' title='New York Times (Web)'
description = 'New York Times on the Web' description = 'New York Times on the Web'
needs_subscription = True needs_subscription = True
elif replaceKindleVersion:
title='The New York Times'
description = 'Today\'s New York Times'
needs_subscription = True
else: else:
title='New York Times' title='New York Times'
description = 'Today\'s New York Times' description = 'Today\'s New York Times'
@ -623,7 +632,7 @@ class NYTimes(BasicNewsRecipe):
self.log(">>> No class:'columnGroup first' found <<<") self.log(">>> No class:'columnGroup first' found <<<")
except: except:
self.log("ERROR: One picture per article in postprocess_html") self.log("ERROR: One picture per article in postprocess_html")
try: try:
# Change captions to italic # Change captions to italic
for caption in soup.findAll(True, {'class':'caption'}) : for caption in soup.findAll(True, {'class':'caption'}) :
@ -637,7 +646,7 @@ class NYTimes(BasicNewsRecipe):
caption.replaceWith(cTag) caption.replaceWith(cTag)
except: except:
self.log("ERROR: Problem in change captions to italic") self.log("ERROR: Problem in change captions to italic")
try: try:
# Change <nyt_headline> to <h2> # Change <nyt_headline> to <h2>
h1 = soup.find('h1') h1 = soup.find('h1')
@ -675,7 +684,7 @@ class NYTimes(BasicNewsRecipe):
except: except:
self.log("ERROR: Problem in Change <h1> to <h3> - used in editorial blogs") self.log("ERROR: Problem in Change <h1> to <h3> - used in editorial blogs")
try: try:
# Change <span class="bold"> to <b> # Change <span class="bold"> to <b>
for subhead in soup.findAll(True, {'class':'bold'}) : for subhead in soup.findAll(True, {'class':'bold'}) :
if subhead.contents: if subhead.contents:
@ -684,15 +693,15 @@ class NYTimes(BasicNewsRecipe):
subhead.replaceWith(bTag) subhead.replaceWith(bTag)
except: except:
self.log("ERROR: Problem in Change <h1> to <h3> - used in editorial blogs") self.log("ERROR: Problem in Change <h1> to <h3> - used in editorial blogs")
try: try:
divTag = soup.find('div',attrs={'id':'articleBody'}) divTag = soup.find('div',attrs={'id':'articleBody'})
if divTag: if divTag:
divTag['class'] = divTag['id'] divTag['class'] = divTag['id']
except: except:
self.log("ERROR: Problem in soup.find(div,attrs={id:articleBody})") self.log("ERROR: Problem in soup.find(div,attrs={id:articleBody})")
try: try:
# Add class="authorId" to <div> so we can format with CSS # Add class="authorId" to <div> so we can format with CSS
divTag = soup.find('div',attrs={'id':'authorId'}) divTag = soup.find('div',attrs={'id':'authorId'})
if divTag and divTag.contents[0]: if divTag and divTag.contents[0]:
@ -700,10 +709,10 @@ class NYTimes(BasicNewsRecipe):
tag['class'] = "authorId" tag['class'] = "authorId"
tag.insert(0, self.fixChars(self.tag_to_string(divTag.contents[0], tag.insert(0, self.fixChars(self.tag_to_string(divTag.contents[0],
use_alt=False))) use_alt=False)))
divTag.replaceWith(tag) divTag.replaceWith(tag)
except: except:
self.log("ERROR: Problem in Add class=authorId to <div> so we can format with CSS") self.log("ERROR: Problem in Add class=authorId to <div> so we can format with CSS")
return soup return soup
def populate_article_metadata(self, article, soup, first): def populate_article_metadata(self, article, soup, first):
shortparagraph = "" shortparagraph = ""