handler.pdf: use pyPdf in setMetada

cloudooo/handler/pdf/handler.py

@@ -25,6 +25,7 @@
 # Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA  02111-1307, USA.
 #
 ##############################################################################
+import io
 
 from zope.interface import implements
 from cloudooo.interfaces.handler import IHandler
@@ -33,6 +34,8 @@ from cloudooo.util import logger, parseContentType
 from subprocess import Popen, PIPE
 from tempfile import mktemp
 
+from pyPdf import PdfFileWriter, PdfFileReader
+from pyPdf.generic import NameObject, createStringObject
 
 class Handler(object):
   """PDF Handler is used to handler inputed pdf document."""
@@ -47,6 +50,7 @@ class Handler(object):
 
   def convert(self, destination_format=None, **kw):
     """ Convert a pdf document """
+    # TODO: use pyPdf
     logger.debug("PDFConvert: %s > %s" % (self.document.source_format, destination_format))
     output_url = mktemp(suffix=".%s" % destination_format,
                         dir=self.document.directory_name)
@@ -66,6 +70,7 @@ class Handler(object):
     """Returns a dictionary with all metadata of document.
     along with the metadata.
     """
+    # TODO: use pyPdf and not use lower()
     command = ["pdfinfo", self.document.getUrl()]
     stdout, stderr = Popen(command,
                            stdout=PIPE,
@@ -75,13 +80,10 @@ class Handler(object):
     info_list = filter(None, stdout.split("\n"))
     metadata = {}
     for info in iter(info_list):
-      if info.count(":") == 1:
-        info_name, info_value = info.split(":")
-      else:
-        info_name, info_value = info.split("  ")
-        info_name = info_name.replace(":", "")
-      info_value = info_value.strip()
-      metadata[info_name.lower()] = info_value
+      info = info.split(":")
+      info_name = info[0].lower()
+      info_value = ":".join(info[1:]).strip()
+      metadata[info_name] = info_value
     self.document.trash()
     return metadata
 
@@ -90,31 +92,27 @@ class Handler(object):
     Keyword arguments:
     metadata -- expected an dictionary with metadata.
     """
-    text_template = "InfoKey: %s\nInfoValue: %s\n"
-    text_list = [text_template % (key.capitalize(), value) \
-                                 for key, value in metadata.iteritems()]
-    metadata_file = File(self.document.directory_name,
-                         "".join(text_list),
-                         "txt")
-    output_url = mktemp(suffix=".pdf",
-                        dir=self.document.directory_name)
-    command = ["pdftk",
-               self.document.getUrl(),
-               "update_info",
-               metadata_file.getUrl(),
-               "output",
-               output_url
-               ]
-    stdout, stderr = Popen(command,
-                           stdout=PIPE,
-                           stderr=PIPE,
-                           close_fds=True,
-                           env=self.environment).communicate()
-    self.document.reload(output_url)
-    try:
-      return self.document.getContent()
-    finally:
-      self.document.trash()
+    # TODO: date as "D:20090401124817-04'00'" ASN.1 for ModDate and CreationDate
+    input_pdf = PdfFileReader(open(self.document.getUrl(), "rb"))
+    output_pdf = PdfFileWriter()
+
+    modification_date = metadata.pop("ModificationDate", None)
+    if modification_date:
+      metadata['ModDate'] = modification_date
+    if type(metadata.get('Keywords', None)) is list:
+      metadata['Keywords'] = metadata['Keywords'].join(' ')
+    args = {}
+    for key, value in list(metadata.items()):
+      args[NameObject('/' + key.capitalize())] = createStringObject(value)
+
+    output_pdf._info.getObject().update(args)
+
+    for page_num in range(input_pdf.getNumPages()):
+      output_pdf.addPage(input_pdf.getPage(page_num))
+
+    output_stream = io.BytesIO()
+    output_pdf.write(output_stream)
+    return output_stream.getvalue()
 
   @staticmethod
   def getAllowedConversionFormatList(source_mimetype):

setup.py

 from setuptools import setup, find_packages
 import sys
 
-version = '1.2.5-dev'
+version = '1.2.6-dev'
 
 def read(name):
     return open(name).read()
@@ -13,6 +13,7 @@ install_requires = [
           'zope.interface',
           'PasteDeploy',
           'PasteScript',
+          'pyPdf',
           'WSGIUtils',
           'psutil>=3.0.0',
           'lxml',