From ac4110fbfc60b112836d0b279ae07ea9a779b16d Mon Sep 17 00:00:00 2001
From: jvoisin
Date: Sun, 20 Jan 2013 03:57:17 +0100
Subject: preliminary support of gtk3 (one test fails for PDF)

---
 MAT/office.py | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

(limited to 'MAT/office.py')

diff --git a/MAT/office.py b/MAT/office.py
index 20664d2..b74b49c 100644
--- a/MAT/office.py
+++ b/MAT/office.py
@@ -10,7 +10,7 @@ import xml.dom.minidom as minidom
 
 try:
     import cairo
-    import poppler
+    from gi.repository import Poppler
 except ImportError:
     pass
 
@@ -125,7 +125,7 @@ class PdfStripper(parser.GenericParser):
         uri = 'file://' + os.path.abspath(self.filename)
         self.password = None
         self.pdf_quality = kwargs['low_pdf_quality']
-        self.document = poppler.document_new_from_file(uri, self.password)
+        self.document = Poppler.Document.new_from_file(uri, self.password)
         self.meta_list = frozenset(['title', 'author', 'subject', 'keywords', 'creator',
             'producer', 'metadata'])
 
-- 
cgit v1.3


From 3d8e11ce644833106f22778f3171c52a51ff69fe Mon Sep 17 00:00:00 2001
From: jvoisin
Date: Fri, 5 Apr 2013 10:55:34 +0200
Subject: Pdf support is back

---
 MAT/office.py | 10 +++++++++-
 test/test.py  |  2 +-
 2 files changed, 10 insertions(+), 2 deletions(-)

(limited to 'MAT/office.py')

diff --git a/MAT/office.py b/MAT/office.py
index b74b49c..a861919 100644
--- a/MAT/office.py
+++ b/MAT/office.py
@@ -6,6 +6,8 @@ import os
 import logging
 import zipfile
 import fileinput
+import tempfile
+import shutil
 import xml.dom.minidom as minidom
 
 try:
@@ -145,11 +147,16 @@ class PdfStripper(parser.GenericParser):
 
             http://cairographics.org/documentation/pycairo/2/
             python-poppler is not documented at all : have fun ;)
+
+            The use of an intermediate tempfile is necessary because
+            python-cairo segfaults on unicode.
+            See http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=699457
         '''
+        output = tempfile.mkstemp()[1]
         page = self.document.get_page(0)
         # assume that every pages are the same size
         page_width, page_height = page.get_size()
-        surface = cairo.PDFSurface(self.output, page_width, page_height)
+        surface = cairo.PDFSurface(output, page_width, page_height)
         context = cairo.Context(surface)  # context draws on the surface
         logging.debug('PDF rendering of %s' % self.filename)
         for pagenum in xrange(self.document.get_n_pages()):
@@ -161,6 +168,7 @@ class PdfStripper(parser.GenericParser):
                 page.render_for_printing(context)  # render the page on context
             context.show_page()  # draw context on surface
         surface.finish()
+        shutil.move(output, self.output)
 
         try:
             import pdfrw  # For now, poppler cannot write meta, so we must use pdfrw
diff --git a/test/test.py b/test/test.py
index 7becd5b..ef1c768 100644
--- a/test/test.py
+++ b/test/test.py
@@ -43,7 +43,7 @@ try:  # file format exclusively managed by exiftool
 except OSError:
     pass  # None for now
 
-FILE_LIST.remove(('clean é.pdf', 'dirty é.pdf'))
+#FILE_LIST.remove(('clean é.pdf', 'dirty é.pdf'))
 #FILE_LIST.remove(('clean é.tar', 'dirty é.tar'))
 FILE_LIST.remove(('clean é.tar.gz', 'dirty é.tar.gz'))
 #FILE_LIST.remove(('clean é.tar.bz2', 'dirty é.tar.bz2'))
-- 
cgit v1.3