Question

I’m using http://docs.python.org/library/zipfile.html to compress file into a zip. It works well but when extracting that file via my ubuntu (not in python, just right clicking file and then ‘Extract here’) file is truncated. My original file has ~167kb, extracted via ubuntu ~164kb (it’s csv file, so I miss few lines from the end of file).

Any suggestions?

Edit:

My code looks as follows:

try:
    fp2 = zipfile.ZipFile(el.replace('.csv', '.zip'), 'w',
            zipfile.ZIP_DEFLATED)
    fp2.testzip()
    fp2.write(el)
finally:
    fp2.close()

Where el is some csv file.

Edit2:
It appears that while writing to a zipfile some data is lost.

Answer 1

Some suggestions:

(1) You call testzip but ignore the returned value. It may mess things up if the archive is empty (as it is in this case). Remove the call to testzip.

(2) Test the integrity of the created file by another means e.g. (a) linux_zipfile_executable -t your_created_archive.zip (b) short Python script using testzip. Try some other extraction means.

(3) Ignore the mode='wb' pushers; that is relevant only to files opened with built-in open. As the zipfile docs say, the only permissible modes are r, w, and a. Else: RuntimeError: ZipFile() requires mode "r", "w", or "a". The zipfile code actually does use the b flag when it opens the archive file using built-in open.

(4) Check carefully that in the flurry of trying to find your problem you are in fact comparing the correct pair of csv files.

(5) Post the 3 files (input, output, and .zip) on the web somewhere.