ZipFile error

[vidar]

I am getting an error on python2.6 on the following script that works fine in python2.7, is this something that is fixable?

I can mitigate it with the fix in https://github.com/vidar/python-pptx/commit/3c87a0fd311c4f22a428877719638409cf4edd97 but that is obviously not a long term approach.

from pptx import Presentation

if __name__ == '__main__':
    prs = Presentation()
    slidelayout = prs.slidemasters[0].slidelayouts[5]
    img_slide = prs.slides.add_slide(slidelayout)
    prs.save('test.pptx')

gives the following stacktrace:

(DM)vidar@berzerker:~$ python a.py 
Traceback (most recent call last):
  File "a.py", line 7, in <module>
    prs.save('test.pptx')
  File "/Users/vidar/.virtualenvs/DM/src/pptx/pptx/api.py", line 64, in save
    return self.__package.save(path)
  File "/Users/vidar/.virtualenvs/DM/src/pptx/pptx/presentation.py", line 125, in save
    pkgng_pkg.save(path)
  File "/Users/vidar/.virtualenvs/DM/src/pptx/pptx/packaging.py", line 134, in save
    zipfs.write_element(cti.element, '/[Content_Types].xml')
  File "/Users/vidar/.virtualenvs/DM/src/pptx/pptx/packaging.py", line 849, in write_element
    if itemURI in self:
  File "/Users/vidar/.virtualenvs/DM/src/pptx/pptx/packaging.py", line 687, in __contains__
    return itemURI in self.itemURIs
  File "/Users/vidar/.virtualenvs/DM/src/pptx/pptx/packaging.py", line 811, in itemURIs
    zip = zipfile.ZipFile(self.path)
  File "/System/Library/Frameworks/Python.framework/Versions/2.6/lib/python2.6/zipfile.py", line 696, in __init__
    self._GetContents()
  File "/System/Library/Frameworks/Python.framework/Versions/2.6/lib/python2.6/zipfile.py", line 716, in _GetContents
    self._RealGetContents()
  File "/System/Library/Frameworks/Python.framework/Versions/2.6/lib/python2.6/zipfile.py", line 728, in _RealGetContents
    raise BadZipfile, "File is not a zip file"
zipfile.BadZipfile: File is not a zip file

[scanny]

Hi Vidar, I also noticed this error when I tested on Python 2.6. It appears to be a bug in the zipfile module in that version that is not likely to ever be fixed.

As I was analyzing the issue back then, the root cause appeared to be that zipfile in 2.6 did not behave well when trying to add additional members to an existing Zip archive. The way the code is structured in python-pptx, the archive is actually closed and reopened between each write of a new member (part). This keeps the interface to ZipFileSystem stateless, which I thought was worth the modest performance hit.

I believe the final solution to this is to change ZipFileSystem so it writes the .pptx file in a single open()/close() session. I'm inclined to see this done by "caching" writes of individual parts until a final "write()" or perhaps "commit()" call is received, at which time ZipFileSystem opens a new zip file for writing and writes all members at once.

I'll take a look and see if I can attend to this, but was wondering if you wanted to take a crack at it first?

Btw, would you mind joining the mailing list at python.pptx@librelist.com? Some of these feature discussions are probably best held there :) Just send an email to that address and it will reply with a confirmation email to get you joined.

[vidar]

• Joined the mailing list.
• This sounds like a reasonable approach, I can take a crack at it, at least explore it. Not sure where to get the time but I would really like to contribute to your awesome work..

[scanny]

Hi vidar, on closer inspection I see the reason for this bug and it's not what I thought above. Good news is it should be quite a bit easier to fix :)

Stay tuned, I should know here in the next hour or two ... I'm pretty sure it has to do with a behavior added to zipfile in Python 2.7.1.

[scanny]

Yep, it was as I suspected. Might be a day or two before I can commit and sync, I have a couple other items in-progress I need to clean up before the tests all pass again, but I'll post the fix as soon as I'm done, probably tomorrow.

[scanny]

Ok, the good news is this latest commit fixes this exception. Also in the good news category is the fix improves performance quite a bit, maybe by a factor of three! I hadn't realized the overhead that reopening the zip file on each part write was introducing. Barely noticeable for most folks I'm sure since it would typically be the difference between 500 and 175 milliseconds for your average presentation, but sure makes the test suite run faster :)

The bad news is there's another one exception right behind it for Python 2.6 that I'll need to track down, but I'll see if I can get it to test clean on 2.6 since we've started down that road here :)

[scanny]

Ok, fixed that exception, now test suite runs clean on Python 2.6, woo hoo!

[ndj91]

Hi guys,

I'm currently getting the same error using python 2.7.0.
Do you know what the cause of this would be?

Thanks.

[scanny]

what version of python-pptx are you running?

python -c "import pptx; print pptx.__version__"

[scanny]

Start with an update to the latest version 0.2.6. Then if it's still a problem send me the stack trace you're seeing when the error is raised and ten or so lines of the code you're using to call it for context. Also would be good to know the filename and path you're using, you can change it to a simple one if it doesn't seem to matter. And also whether you're trying to overwrite an existing file or create a new one. Operating system and major version is good to know too.

[ndj91]

Hi, the issue was that the version of the library was not updated to the latest version. After an update had been completed the script written functioned correctly. Thank you for spending the time to reply and write such a useful python library. I suggested the use of this library to work as I was so impressed by it functionality and thought it would aid in a project which we are currently working on. Hopefully if my python knowledge and understanding of pptx file formats increases, I might even be able to contribute to the project. Keep up the great work.

[scanny]

Glad to hear it KingNDj247 :)

And thanks for your kind words :)