EPUB Output: Remove any margins specified via an Adobe page template in the input document

src/calibre/ebooks/conversion/plumber.py

@@ -1003,8 +1003,10 @@ OptionRecommendation(name='sr3_replace',
         self.opts.insert_blank_line = oibl
         self.opts.remove_paragraph_spacing = orps
 
-        from calibre.ebooks.oeb.transforms.page_margin import RemoveFakeMargins
+        from calibre.ebooks.oeb.transforms.page_margin import \
+            RemoveFakeMargins, RemoveAdobeMargins
         RemoveFakeMargins()(self.oeb, self.log, self.opts)
+        RemoveAdobeMargins()(self.oeb, self.log, self.opts)
 
         pr(0.9)
         self.flush()

src/calibre/ebooks/oeb/transforms/margins.py

@@ -1,56 +0,0 @@
-#!/usr/bin/env python
-# vim:fileencoding=UTF-8:ts=4:sw=4:sta:et:sts=4:ai
-
-__license__   = 'GPL v3'
-__copyright__ = '2010, Kovid Goyal <kovid@kovidgoyal.net>'
-__docformat__ = 'restructuredtext en'
-
-
-class RemoveFakeMargins(object):
-    '''
-    Try to detect and remove fake margins inserted by asinine ebook creation
-    software on each paragraph/wrapper div. Can be used only after CSS
-    flattening.
-    '''
-
-    def __call__(self, oeb, opts, log):
-        self.oeb, self.opts, self.log = oeb, opts, log
-
-        from calibre.ebooks.oeb.base import XPath, OEB_STYLES
-
-        stylesheet = None
-        for item in self.oeb.manifest:
-            if item.media_type.lower() in OEB_STYLES:
-                stylesheet = item.data
-                break
-
-        if stylesheet is None:
-            return
-
-
-        top_level_elements = {}
-        second_level_elements = {}
-
-        for x in self.oeb.spine:
-            root = x.data
-            body = XPath('//h:body')(root)
-            if body:
-                body = body[0]
-
-            if not hasattr(body, 'xpath'):
-                continue
-
-            # Check for margins on top level elements
-            for lb in XPath('./h:div|./h:p|./*/h:div|./*/h:p')(body):
-                cls = lb.get('class', '')
-                level = top_level_elements if lb.getparent() is body else \
-                        second_level_elements
-                if cls not in level:
-                    level[cls] = []
-                    top_level_elements[cls] = []
-                level[cls].append(lb)
-
-
-    def get_margins(self, stylesheet, cls):
-        pass
-

src/calibre/ebooks/oeb/transforms/page_margin.py

@@ -11,6 +11,26 @@ from collections import Counter
 
 from calibre.ebooks.oeb.base import OEB_STYLES, barename, XPath
 
+class RemoveAdobeMargins(object):
+    '''
+    Remove margins specified in Adobe's page templates.
+    '''
+
+    def __call__(self, oeb, log, opts):
+        self.oeb, self.opts, self.log = oeb, opts, log
+
+        for item in self.oeb.manifest:
+            if item.media_type == 'application/vnd.adobe-page-template+xml':
+                self.log('Removing page margins specified in the'
+                        ' Adobe page template')
+                for elem in item.data.xpath(
+                        '//*[@margin-bottom or @margin-top '
+                        'or @margin-left or @margin-right]'):
+                    for margin in ('left', 'right', 'top', 'bottom'):
+                        attr = 'margin-'+margin
+                        elem.attrib.pop(attr, None)
+
+
 class RemoveFakeMargins(object):
 
     '''