From 1ae3724038b4f6f7838fec416781ba14495c4e3f Mon Sep 17 00:00:00 2001
From: Kovid Goyal <kovid@kovidgoyal.net>
Date: Sun, 16 Aug 2009 15:41:34 -0600
Subject: [PATCH] New recipe for Cuba Debate by Darko Miletic. Also improve
 effectiveness of the linearize tables transform.

---
 .../ebooks/oeb/transforms/linearize_tables.py |  12 +++--
 src/calibre/gui2/images/news/cubadebate.png   | Bin 0 -> 858 bytes
 src/calibre/web/feeds/recipes/__init__.py     |   2 +-
 .../web/feeds/recipes/recipe_cubadebate.py    |  44 ++++++++++++++++++
 4 files changed, 53 insertions(+), 5 deletions(-)
 create mode 100644 src/calibre/gui2/images/news/cubadebate.png
 create mode 100644 src/calibre/web/feeds/recipes/recipe_cubadebate.py

diff --git a/src/calibre/ebooks/oeb/transforms/linearize_tables.py b/src/calibre/ebooks/oeb/transforms/linearize_tables.py
index 5842db90e5..92c3a00c7b 100644
--- a/src/calibre/ebooks/oeb/transforms/linearize_tables.py
+++ b/src/calibre/ebooks/oeb/transforms/linearize_tables.py
@@ -6,15 +6,19 @@ __license__   = 'GPL v3'
 __copyright__ = '2009, Kovid Goyal <kovid@kovidgoyal.net>'
 __docformat__ = 'restructuredtext en'
 
-from calibre.ebooks.oeb.base import OEB_DOCS, XPNSMAP
+from calibre.ebooks.oeb.base import OEB_DOCS, XPath
 
 class LinearizeTables(object):
 
     def linearize(self, root):
-        for x in root.xpath('//h:table|//h:td|//h:tr|//h:th',
-                namespaces=XPNSMAP):
+        for x in XPath('//h:table|//h:td|//h:tr|//h:th|//h:caption|'
+                '//h:tbody|//h:tfoot|//h:thead|//h:colgroup|//h:col')(root):
             x.tag = 'div'
-            for attr in ('valign', 'colspan', 'rowspan', 'width', 'halign'):
+            for attr in ('style', 'font', 'valign',
+                         'colspan', 'width', 'height',
+                         'rowspan', 'summary', 'align',
+                         'cellspacing', 'cellpadding',
+                         'frames', 'rules', 'border'):
                 if attr in x.attrib:
                     del x.attrib[attr]
 
diff --git a/src/calibre/gui2/images/news/cubadebate.png b/src/calibre/gui2/images/news/cubadebate.png
new file mode 100644
index 0000000000000000000000000000000000000000..4f8fc9e4fcd9e3477beced5472b93fdd72796143
GIT binary patch
literal 858
zcmeAS@N?(olHy`uVBq!ia0vp^0wB!61|;P_|4#%`Y)RhkE)4%caKYZ?lYt_f1s;*b
zK-vS0-A-oPfdtD69Mgd`SU*F|v9*U87?_$pT^vI!PEVb5-uqUe#PR#*zh~Q@(EG>k
z$MH%@WFmW~goyKtOKhx(%Do*4CS5l+-r%cz^PueJf`pP7Iiu@G7ue`1x+qOzZETsQ
zDB=4lJFD)V^$a~W!*2%;KmTN-fA4%Z|A+92TUHiThRe!KcxVtjlc)Kh!seT694{@O
za_K{e;lvNcsjXQB6W+WGjEQ*AUoppT`DDKKW@BSxo>r&pD)EdBL0rp@eYzMZBG=!n
zB*1XNq%lZ=^XrFWj|-0^8J>Rm@}>EtIKfDBMh0{9`-xSz8iP8SI6NNwlhc?U7|PX_
zxFBk+*`!~8>!jK~wx-;1In0orZvWsJo71Ugjs-!CU9KUkLq)h+*&JLL9L&tjzV!9=
zZI}A~?OsCmYS&GXk8T<#WUge}v5V<b<rHnvCq<Sin<IPr`jYbU_*$J9UzSKs^mw8=
z<#z3gEbXm)EGK;0wnTVL?_T(~wu@u>^y!|H%x3p8a5R+do_i?cOW*V7x{EUF7xniu
zcd~5Q)g`-KdU|^Lj9IgsOr%_mX1W9i3o|U(x>ZyyF2<p+pD}j5q}Ei1!z&suWlEHn
zpLklt&@g$j@RFdFZD+P^Gjs6ukuiHT+b6}q<JjXRH+^=+{E_PvVAva{|M{m)dt2Kp
zb&cJ3{d~p5_nS>%W_HkFD=BAgb^4p`biH)V+O>%}IXrv!?%fcf!=T`|{Ii(KY)hM(
z)5n!w|5XUQ{<UpK)$3j7927RJ>-3l@cfNYo*&p*~&7ZH|*V9w8=6KGtZ)E|crGbpE
z;yG*V<URe&GrU$xz4-NO@~k;EpIBC3Jr$u7rt)*qJ9jx}6W?Z~`KO;g&2Br)mzlp^
z?f0*!L)O;&t*ou<zkUDyp6Q2<(#l_d@|{j^=l2x4^97hTRZCnWN|K9G6H8KcQ-B16
zk%57wu7QQFp;?Hbsg<$0m64IIfti(oflk!5Eieta`6-!cm1r7FLktb9j7_YJ3?Le;
Tm+*Q6H86O(`njxgN@xNACqQG<

literal 0
HcmV?d00001

diff --git a/src/calibre/web/feeds/recipes/__init__.py b/src/calibre/web/feeds/recipes/__init__.py
index 1588a4c8ab..d6ba724256 100644
--- a/src/calibre/web/feeds/recipes/__init__.py
+++ b/src/calibre/web/feeds/recipes/__init__.py
@@ -48,7 +48,7 @@ recipe_modules = ['recipe_' + r for r in (
            'the_budget_fashionista', 'elperiodico_catalan',
            'elperiodico_spanish', 'expansion_spanish', 'lavanguardia',
            'marca', 'kellog_faculty', 'kellog_insight', 'noaa',
-           '7dias', 'buenosaireseconomico', 'huntechnet',
+           '7dias', 'buenosaireseconomico', 'huntechnet', 'cubadebate',
            'diagonales', 'miradasalsur', 'newsweek_argentina', 'veintitres',
            'gva_be', 'hln', 'tijd', 'degentenaar', 'inquirer_net', 'uncrate',
            'fastcompany', 'accountancyage', 'laprensa_hn', 'latribuna',
diff --git a/src/calibre/web/feeds/recipes/recipe_cubadebate.py b/src/calibre/web/feeds/recipes/recipe_cubadebate.py
new file mode 100644
index 0000000000..13747348fd
--- /dev/null
+++ b/src/calibre/web/feeds/recipes/recipe_cubadebate.py
@@ -0,0 +1,44 @@
+#!/usr/bin/env  python
+
+__license__   = 'GPL v3'
+__copyright__ = '2009, Darko Miletic <darko.miletic at gmail.com>'
+'''
+newyorker.com
+'''
+
+from calibre.web.feeds.news import BasicNewsRecipe
+from calibre.ebooks.BeautifulSoup import Tag
+
+class CubaDebate(BasicNewsRecipe):
+    title                 = 'CubaDebate'
+    __author__            = 'Darko Miletic'
+    description           = 'Contra el Terorismo Mediatico'
+    oldest_article        = 15
+    language              = _('Spanish')
+    max_articles_per_feed = 100
+    no_stylesheets        = True
+    use_embedded_content  = False
+    publisher             = 'Cubadebate'
+    category              = 'news, politics, Cuba'
+    encoding              = 'utf-8'
+    extra_css             = ' #BlogTitle{font-size: x-large; font-weight: bold} '
+    
+    conversion_options = {  
+                             'comments'    : description
+                            ,'tags'        : category
+                            ,'language'    : 'es'
+                            ,'publisher'   : publisher
+                            ,'pretty_print': True
+                         }
+                         
+    keep_only_tags = [dict(name='div', attrs={'id':'Outline'})]
+    remove_tags_after = dict(name='div',attrs={'id':'BlogContent'})
+    remove_tags = [dict(name='link')]
+
+    feeds          = [(u'Articulos', u'http://www.cubadebate.cu/feed/')]
+
+    def print_version(self, url):
+        return url + 'print/'
+
+    def preprocess_html(self, soup):
+        return self.adeify_images(soup)