From df6f0f8dc73f50fafb3e5d44dfdce2e23d0cc10e Mon Sep 17 00:00:00 2001
From: Kovid Goyal <kovid@kovidgoyal.net>
Date: Sat, 4 May 2013 11:06:59 +0530
Subject: [PATCH] Docx metadata: Read the language of the file, if present

---
 src/calibre/ebooks/metadata/docx.py | 10 ++++++++++
 1 file changed, 10 insertions(+)

diff --git a/src/calibre/ebooks/metadata/docx.py b/src/calibre/ebooks/metadata/docx.py
index 1505d397f3..cb265424cc 100644
--- a/src/calibre/ebooks/metadata/docx.py
+++ b/src/calibre/ebooks/metadata/docx.py
@@ -10,6 +10,7 @@ __docformat__ = 'restructuredtext en'
 from lxml import etree
 
 from calibre.ebooks.metadata.book.base import Metadata
+from calibre.utils.localization import canonicalize_lang
 from calibre.utils.zipfile import ZipFile
 from calibre.utils.magick.draw import identify_data
 from calibre.ebooks.oeb.base import DC11_NS
@@ -52,6 +53,15 @@ def _read_doc_props(raw, mi):
         raw = etree.tostring(desc[0], method='text', encoding=unicode)
         mi.comments = raw
 
+    langs = []
+    for lang in XPath('//dc:language')(root):
+        if lang.text and lang.text.strip():
+            l = canonicalize_lang(lang.text)
+            if l:
+                langs.append(l)
+    if langs:
+        mi.languages = langs
+
 def _read_app_props(raw, mi):
     root = etree.fromstring(raw, parser=RECOVER_PARSER)
     company = root.xpath('//*[local-name()="Company"]')