First implementation of open document format

2 files changed, 114 insertions, 8 deletions

diff --git a/lib/mat.py b/lib/mat.py
index 2903ed9..5dcdbc2 100644
--- a/lib/mat.py
+++ b/lib/mat.py
@@ -7,6 +7,7 @@
 import os
 import subprocess
 import logging
+import mimetypes
 
 import hachoir_core.cmd_line
 import hachoir_parser
@@ -14,7 +15,7 @@ import hachoir_editor
 
 import images
 import audio
-import misc
+import office
 import archive
 
 __version__ = "0.1"
@@ -29,7 +30,7 @@ strippers = {
     hachoir_parser.image.PngFile: images.PngStripper,
     hachoir_parser.image.bmp.BmpFile: images.BmpStripper,
     hachoir_parser.audio.MpegAudioFile: audio.MpegAudioStripper,
-    hachoir_parser.misc.PDFDocument: misc.PdfStripper,
+    hachoir_parser.misc.PDFDocument: office.PdfStripper,
     hachoir_parser.archive.TarFile: archive.TarStripper,
     hachoir_parser.archive.gzip_parser.GzipParser: archive.GzipStripper,
     hachoir_parser.archive.bzip2_parser.Bzip2Parser: archive.Bzip2Stripper,
@@ -61,12 +62,14 @@ def create_class_file(name, backup, add2archive):
         corresponding to the filetype of the given file
     '''
     if is_secure(name):
-        print 'a'
         return
 
     filename = ""
     realname = name
-    filename = hachoir_core.cmd_line.unicodeFilename(name)
+    try:
+        filename = hachoir_core.cmd_line.unicodeFilename(name)
+    except TypeError:# get rid of "TypeError: decoding Unicode is not supported"
+        filename = name
     parser = hachoir_parser.createParser(filename)
     if not parser:
         logging.error("Unable to parse %s" % filename)
@@ -82,9 +85,26 @@ def create_class_file(name, backup, add2archive):
         stripper_class = strippers[editor.input.__class__]
     except KeyError:
         #Place for another lib than hachoir
-        logging.error("Don't have stripper for file type %s" % editor.description)
+        logging.error("Don't have stripper for format %s" % editor.description)
         return
-    if editor.input.__class__ == hachoir_parser.misc.PDFDocument:
+
+    if editor.input.__class__ == hachoir_parser.misc.PDFDocument:#pdf
         return stripper_class(filename, realname, backup)
-    return stripper_class(realname, filename, parser, editor, backup,
-        add2archive)
+
+    elif editor.input.__class__ == hachoir_parser.archive.zip.ZipFile:
+        #zip based format
+        mime = mimetypes.guess_type(filename)[0]
+        try:#Ugly workaround, cleaning open document delete mime (wtf?)
+            if mime.startswith(#Open document format
+            'application/vnd.oasis.opendocument'):
+                return office.OpenDocumentStripper(realname, filename, parser,
+                    editor, backup, add2archive)
+            else:#normal zip
+                return stripper_class(realname, filename, parser, editor,
+                    backup, add2archive)
+        except:#normal zip file
+            return stripper_class(realname, filename, parser, editor, backup,
+                add2archive)
+    else:#normal handling
+        return stripper_class(realname, filename, parser, editor, backup,
+            add2archive)
diff --git a/lib/office.py b/lib/office.py
index de38129..5d62732 100644
--- a/lib/office.py
+++ b/lib/office.py
@@ -3,12 +3,98 @@ import mimetypes
 import subprocess
 import tempfile
 import glob
+import logging
+import zipfile
+import shutil
 
 import hachoir_core
 
 import pdfrw
 import mat
 import parser
+import archive
+
+class OpenDocumentStripper(archive.GenericArchiveStripper):
+    '''
+        An open document file is a zip, with xml file into.
+        The one that interest us is meta.xml
+    '''
+
+    def remove_folder(self, folder_list):
+        for folder in folder_list:
+            dirname = folder.split('/')[0]
+            try:
+                shutil.rmtree(dirname)
+            except:#Some folder or open document format are buggies
+                pass
+        self.folder_list = []
+
+    def _remove_all(self, method):
+        '''
+            FIXME ?
+            There is a patch implementing the Zipfile.remove()
+            method here : http://bugs.python.org/issue6818
+        '''
+        zipin = zipfile.ZipFile(self.filename, 'r')
+        zipout = zipfile.ZipFile(self.filename + parser.POSTFIX, 'w',
+            allowZip64=True)
+        folder_list = []
+        for item in zipin.namelist():
+            if os.path.basename(item) is not item:#add folders to folder_list
+                folder_list.insert(0, os.path.dirname(item))
+            if item.endswith('.xml') or item.startswith('manifest'):
+                if item != 'meta.xml':#contains the metadata
+                    zipin.extract(item)
+                    zipout.write(item)
+                    mat.secure_remove(item)
+            elif item == 'mimetype':
+                zipin.extract(item)
+                #remove line meta.xml
+                zipout.write(item)
+                mat.secure_remove(item)
+            else:
+                zipin.extract(item)
+                if os.path.isfile(item):
+                    try:
+                        cfile = mat.create_class_file(item, False,
+                            self.add2archive)
+                        if method == 'normal':
+                            cfile.remove_all()
+                        else:
+                            cfile.remove_all_ugly()
+                        logging.debug('Processing %s from %s' % (item,
+                            self.filename))
+                        zipout.write(item)
+                    except:
+                        logging.info('%s\' fileformat is not supported' %
+                            item)
+                        if self.add2archive:
+                            zipout.write(item)
+                    mat.secure_remove(item)
+        zipout.comment = ''
+        logging.info('%s treated' % self.filename)
+        zipin.close()
+        zipout.close()
+        self.remove_folder(folder_list)
+
+        if self.backup is False:
+            mat.secure_remove(self.filename) #remove the old file
+            os.rename(self.filename + parser.POSTFIX, self.filename)
+
+    def is_clean(self):
+        zipin = zipfile.ZipFile(self.filename, 'r')
+        try:
+            zipin.getinfo('meta.xml')
+        except KeyError:#no meta.xml in the file
+                zipin.close()
+                czf = archive.ZipStripper(self.realname, self.filename,
+                    self.parser, self.editor, self.backup, self.add2archive)
+                if czf.is_clean():
+                    return True
+                else:
+                    return False
+        return False
+
 
 class TorrentStripper(parser.Generic_parser):
     '''