DOCX: Add support for theme fonts

src/calibre/ebooks/docx/char_styles.py

@@ -117,8 +117,12 @@ def read_vert_align(parent, dest):
 
 def read_font_family(parent, dest):
     ans = inherit
-    for col in XPath('./w:rFonts[@w:ascii]')(parent):
-        val = get(col, 'w:ascii')
+    for col in XPath('./w:rFonts')(parent):
+        val = get(col, 'w:asciiTheme')
+        if val:
+            val = '|%s|' % val
+        else:
+            val = get(col, 'w:ascii')
         if val:
             ans = val
     setattr(dest, 'font_family', ans)

src/calibre/ebooks/docx/names.py

@@ -22,6 +22,7 @@ IMAGES    = 'http://schemas.openxmlformats.org/officeDocument/2006/relationships
 LINKS     = 'http://schemas.openxmlformats.org/officeDocument/2006/relationships/hyperlink'
 FOOTNOTES = 'http://schemas.openxmlformats.org/officeDocument/2006/relationships/footnotes'
 ENDNOTES  = 'http://schemas.openxmlformats.org/officeDocument/2006/relationships/endnotes'
+THEMES    = 'http://schemas.openxmlformats.org/officeDocument/2006/relationships/theme'
 
 namespaces = {
     'mo': 'http://schemas.microsoft.com/office/mac/office/2008/main',

src/calibre/ebooks/docx/styles.py

@@ -142,8 +142,8 @@ class Styles(object):
     def get(self, key, default=None):
         return self.id_map.get(key, default)
 
-    def __call__(self, root, fonts):
-        self.fonts = fonts
+    def __call__(self, root, fonts, theme):
+        self.fonts, self.theme = fonts, theme
         for s in XPath('//w:style')(root):
             s = Style(s)
             if s.style_id:
@@ -304,7 +304,8 @@ class Styles(object):
                 setattr(ans, attr, self.run_val(parent_styles, direct_formatting, attr))
 
             if ans.font_family is not inherit:
-                ans.font_family = self.fonts.family_for(ans.font_family, ans.b, ans.i)
+                ff = self.theme.resolve_font_family(ans.font_family)
+                ans.font_family = self.fonts.family_for(ff, ans.b, ans.i)
 
         return ans
 

src/calibre/ebooks/docx/theme.py

@@ -0,0 +1,31 @@
+#!/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>'
+
+from calibre.ebooks.docx.names import XPath
+
+
+class Theme(object):
+
+    def __init__(self):
+        self.major_latin_font = 'Cambria'
+        self.minor_latin_font = 'Calibri'
+
+    def __call__(self, root):
+        for fs in XPath('//a:fontScheme')(root):
+            for mj in XPath('./a:majorFont')(fs):
+                for l in XPath('./a:latin[@typeface]')(mj):
+                    self.major_latin_font = l.get('typeface')
+            for mj in XPath('./a:minorFont')(fs):
+                for l in XPath('./a:latin[@typeface]')(mj):
+                    self.minor_latin_font = l.get('typeface')
+
+    def resolve_font_family(self, ff):
+        if ff.startswith('|'):
+            ff = ff[1:-1]
+            ff = self.major_latin_font if ff.startswith('major') else self.minor_latin_font
+        return ff

src/calibre/ebooks/docx/to_html.py

@@ -16,7 +16,7 @@ from lxml.html.builder import (
 from calibre.ebooks.docx.container import DOCX, fromstring
 from calibre.ebooks.docx.names import (
     XPath, is_tag, XML, STYLES, NUMBERING, FONTS, get, generate_anchor,
-    descendants, FOOTNOTES, ENDNOTES, children)
+    descendants, FOOTNOTES, ENDNOTES, children, THEMES)
 from calibre.ebooks.docx.styles import Styles, inherit, PageProperties
 from calibre.ebooks.docx.numbering import Numbering
 from calibre.ebooks.docx.fonts import Fonts
@@ -24,6 +24,7 @@ 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.docx.theme import Theme
 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
@@ -49,6 +50,7 @@ class Convert(object):
         self.dest_dir = dest_dir or os.getcwdu()
         self.mi = self.docx.metadata
         self.body = BODY()
+        self.theme = Theme()
         self.tables = Tables()
         self.styles = Styles(self.tables)
         self.images = Images()
@@ -203,6 +205,7 @@ class Convert(object):
         nname = get_name(NUMBERING, 'numbering.xml')
         sname = get_name(STYLES, 'styles.xml')
         fname = get_name(FONTS, 'fontTable.xml')
+        tname = get_name(THEMES, 'theme1.xml')
         foname = get_name(FOOTNOTES, 'footnotes.xml')
         enname = get_name(ENDNOTES, 'endnotes.xml')
         numbering = self.numbering = Numbering()
@@ -231,13 +234,21 @@ class Convert(object):
             else:
                 fonts(fromstring(raw), embed_relationships, self.docx, self.dest_dir)
 
+        if tname is not None:
+            try:
+                raw = self.docx.read(tname)
+            except KeyError:
+                self.log.warn('Styles %s do not exist' % sname)
+            else:
+                self.theme(fromstring(raw))
+
         if sname is not None:
             try:
                 raw = self.docx.read(sname)
             except KeyError:
                 self.log.warn('Styles %s do not exist' % sname)
             else:
-                self.styles(fromstring(raw), fonts)
+                self.styles(fromstring(raw), fonts, self.theme)
 
         if nname is not None:
             try: