Initial commit of the new office converter class

apps/converter/api.py

@@ -4,8 +4,6 @@ import hashlib
 
 from common.conf.settings import TEMPORARY_DIRECTORY
 
-from converter.conf.settings import UNOCONV_PATH
-from converter.exceptions import OfficeConversionError
 from converter.literals import DEFAULT_PAGE_NUMBER, \
     DEFAULT_ZOOM_LEVEL, DEFAULT_ROTATION, DEFAULT_FILE_FORMAT
 
@@ -16,28 +14,12 @@ from converter.literals import TRANSFORMATION_RESIZE, \
 from converter.literals import DIMENSION_SEPARATOR
 from converter.literals import FILE_FORMATS
 from converter.utils import cleanup
+from converter.office_converter import OfficeConverter
+
 
 HASH_FUNCTION = lambda x: hashlib.sha256(x).hexdigest()
-    
-CONVERTER_OFFICE_FILE_EXTENSIONS = [
-    u'ods', u'docx', u'doc'
-]
-
-
-def execute_unoconv(input_filepath, arguments=''):
-    """
-    Executes the program unoconv using subprocess's Popen
-    """
-    command = []
-    command.append(UNOCONV_PATH)
-    command.extend(unicode(arguments).split())
-    command.append(input_filepath)
-    proc = subprocess.Popen(command, close_fds=True, stderr=subprocess.PIPE)
-    return_code = proc.wait()
-    if return_code != 0:
-        raise OfficeConversionError(proc.stderr.readline())
-
 
+            
 def cache_cleanup(input_filepath, *args, **kwargs):
     try:
         os.remove(create_image_cache_filename(input_filepath, *args, **kwargs))
@@ -53,13 +35,6 @@ def create_image_cache_filename(input_filepath, *args, **kwargs):
         return None
 
 
-def convert_office_document(input_filepath):
-    if os.path.exists(UNOCONV_PATH):
-        execute_unoconv(input_filepath, arguments='-f pdf')
-        return input_filepath + u'.pdf'
-    return None
-
-
 def convert(input_filepath, output_filepath=None, cleanup_files=False, *args, **kwargs):
     size = kwargs.get('size')
     file_format = kwargs.get('file_format', DEFAULT_FILE_FORMAT)
@@ -70,20 +45,23 @@ def convert(input_filepath, output_filepath=None, cleanup_files=False, *args, **
     if transformations is None:
         transformations = []
 
-    unoconv_output = None
-
     if output_filepath is None:
         output_filepath = create_image_cache_filename(input_filepath, *args, **kwargs)
+    print 'cache image', output_filepath
         
     if os.path.exists(output_filepath):
         return output_filepath
+    
+    print 'cleanup_files', cleanup_files
 
-    path, extension = os.path.splitext(input_filepath)
-    if extension[1:].lower() in CONVERTER_OFFICE_FILE_EXTENSIONS:
-        result = convert_office_document(input_filepath)
-        if result:
-            unoconv_output = result
-            input_filepath = result
+    office_converter = OfficeConverter(input_filepath)
+    if office_converter:
+        try:
+            #cleanup_files =False.
+            input_filepath = office_converter.output_filepath
+        except OfficeConverter:
+            print 'office converter exception'
+            raise UnknownFileFormat('office converter exception')
 
     if size:
         transformations.append(
@@ -114,8 +92,6 @@ def convert(input_filepath, output_filepath=None, cleanup_files=False, *args, **
     finally:
         if cleanup_files:
             cleanup(input_filepath)
-        if unoconv_output:
-            cleanup(unoconv_output)
 
     return output_filepath
 

apps/converter/exceptions.py

@@ -29,3 +29,7 @@ class UnkownConvertError(ConvertError):
 
 class OfficeConversionError(ConvertError):
     pass
+
+
+class OfficeBackendError(OfficeConversionError):
+    pass

apps/converter/office_converter.py

@@ -0,0 +1,106 @@
+import os
+import subprocess
+import hashlib
+
+from mimetype.api import get_mimetype
+from common.conf.settings import TEMPORARY_DIRECTORY
+
+from converter.conf.settings import UNOCONV_PATH
+from converter.exceptions import (OfficeConversionError,
+    OfficeBackendError, UnknownFileFormat)
+
+HASH_FUNCTION = lambda x: hashlib.sha256(x).hexdigest()
+
+CACHED_FILE_SUFFIX = u'_office_converter'
+    
+CONVERTER_OFFICE_FILE_MIMETYPES = [
+    'application/msword',
+    'application/mswrite',
+    'application/mspowerpoint',
+    'application/msexcel',
+    'application/vnd.ms-excel',
+    'application/vnd.ms-powerpoint',    
+    'text/plain',
+    'application/vnd.oasis.opendocument.presentation',
+]
+#    'application/vnd.oasis.opendocument.text': 'ODF_textdocument_32x32.png',
+#    'application/vnd.oasis.opendocument.spreadsheet': 'ODF_spreadsheet_32x32.png',
+#    'application/vnd.oasis.opendocument.presentation': 'ODF_presentation_32x32.png',
+#    'application/vnd.oasis.opendocument.graphics': 'ODF_drawing_32x32.png',
+#    'application/vnd.openxmlformats-officedocument.spreadsheetml.sheet': 'file_extension_xls.png',
+#    'application/vnd.openxmlformats-officedocument.wordprocessingml.document': 'file_extension_doc.png',
+#    'application/vnd.oasis.opendocument.text': 'ODF_textdocument_32x32.png',
+
+class OfficeConverter(object):
+    def __init__(self, input_filepath):
+        self.backend = OfficeConverterBackendUnoconv(unoconv_path=UNOCONV_PATH)
+        self.input_filepath = input_filepath
+        self.exists = False
+
+        # Make sure file is of a known office format  
+        descriptor = open(self.input_filepath)
+        mimetype, encoding = get_mimetype(descriptor, self.input_filepath)
+
+        if mimetype in CONVERTER_OFFICE_FILE_MIMETYPES:
+            # Hash file to cache results of conversion
+            #descriptor = open(self.input_filepath)
+            #file_hash = HASH_FUNCTION(descriptor.read())
+            #descriptor.close()
+            
+            #self.output_filepath = os.path.join(TEMPORARY_DIRECTORY, u''.join([file_hash, CACHED_FILE_SUFFIX]))
+            self.output_filepath = os.path.join(TEMPORARY_DIRECTORY, u''.join([self.input_filepath, CACHED_FILE_SUFFIX]))
+            self.exists = os.path.exists(self.output_filepath)
+            print 'self.input_filepath',self.input_filepath
+            print 'self.output_filepath',self.output_filepath
+            print 'self.exists', self.exists
+            if not self.exists:
+                try:
+                    self.backend.convert(self.input_filepath, self.output_filepath)
+                except OfficeBackendError, msg:
+                    print 'OFFICE EXCEPTION'
+                    # convert exception so that atleas the mime type icon is displayed
+                    raise UnknownFileFormat(msg)
+        
+
+        
+    def __unicode__(self):
+        return getattr(self, 'output_filepath', None)
+        
+    def __str__(self):
+        return str(self.__unicode__())
+                
+    def __nonzero__(self):
+        return self.exists
+
+    __bool__ = __nonzero__
+    
+
+class OfficeConverterBackendUnoconv(object):
+    def __init__(self, unoconv_path=None):
+        self.unoconv_path = unoconv_path if unoconv_path else u'/usr/bin/unoconv'
+        if not os.path.exists(self.unoconv_path):
+            raise OfficeBackendError('cannot find unoconv executable')
+
+    def convert(self, input_filepath, output_filepath):
+        """
+        Executes the program unoconv using subprocess's Popen
+        """
+        self.input_filepath = input_filepath
+        self.output_filepath = output_filepath
+        
+        command = []
+        command.append(self.unoconv_path)
+        #command.append(u'-v')
+        command.append(u'--pipe')
+        command.append(u'--format="pdf"')
+        command.append(u'--output=%s' % self.output_filepath)
+        command.append(self.input_filepath)
+        print 'convert'
+        try:
+            proc = subprocess.Popen(command, close_fds=True, stderr=subprocess.PIPE, stdout=subprocess.PIPE)
+            return_code = proc.wait()
+            readline = proc.stderr.readline()
+            if return_code != 0:
+                raise OfficeBackendError(proc.stderr.readline())
+        except OSError, msg:
+            raise OfficeBackendError(msg)