DOCX: Merge runs with the same styling to reduce amount of generated

markup

src/calibre/ebooks/docx/cleanup.py

@@ -0,0 +1,59 @@
+#!/usr/bin/env python
+# vim:fileencoding=utf-8
+from __future__ import (unicode_literals, division, absolute_import,
+                        print_function)
+
+__license__ = 'GPL v3'
+__copyright__ = '2013, Kovid Goyal <kovid at kovidgoyal.net>'
+
+
+def mergeable(previous, current):
+    if previous.tail or current.tail:
+        return False
+    if previous.get('class', None) != current.get('class', None):
+        return False
+    if current.get('id', False):
+        return False
+    try:
+        return next(previous.itersiblings()) is current
+    except StopIteration:
+        return False
+
+
+def append_text(parent, text):
+    if len(parent) > 0:
+        parent[-1].tail = (parent[-1].tail or '') + text
+    else:
+        parent.text = (parent.text or '') + text
+
+
+def merge(parent, span):
+    if span.text:
+        append_text(parent, span.text)
+    for child in span:
+        parent.append(child)
+    if span.tail:
+        append_text(parent, span.tail)
+    span.getparent().remove(span)
+
+
+def merge_run(run):
+    parent = run[0]
+    for span in run[1:]:
+        merge(parent, span)
+
+
+def cleanup_markup(root, styles):
+    # Merge consecutive spans that have the same styling
+    current_run = []
+    for span in root.xpath('//span'):
+        if not current_run:
+            current_run.append(span)
+        else:
+            last = current_run[-1]
+            if mergeable(last, span):
+                current_run.append(span)
+            else:
+                if len(current_run) > 1:
+                    merge_run(current_run)
+                current_run = [span]

src/calibre/ebooks/docx/to_html.py

@@ -23,6 +23,7 @@ from calibre.ebooks.docx.fonts import Fonts
 from calibre.ebooks.docx.images import Images
 from calibre.ebooks.docx.tables import Tables
 from calibre.ebooks.docx.footnotes import Footnotes
+from calibre.ebooks.docx.cleanup import cleanup_markup
 from calibre.ebooks.metadata.opf2 import OPFCreator
 from calibre.ebooks.metadata.toc import TOC
 from calibre.ebooks.oeb.polish.toc import elem_to_toc_text
@@ -157,6 +158,8 @@ class Convert(object):
                     notes_header.set('class', '%s notes-header' % cls)
                 break
 
+        cleanup_markup(self.html, self.styles)
+
         return self.write()
 
     def read_page_properties(self, doc):