如何判断文件是否是 gzip 压缩的？

Question

I have a Python program which is going to take text files as input. However, some of these files may be gzip compressed.

Is there a cross-platform, usable from Python way to determine if a file is gzip compressed or not?

Is the following reliable or could an ordinary text file 'accidentally' look gzip-like enough for me to get false positives?

try:
    gzip.GzipFile(filename, 'r')
    # compressed
    # ...
except:
    # not compressed
    # ...

Answer 1

The magic number for gzip compressed files is 1f 8b . Although testing for this is not 100% reliable, it is highly unlikely that "ordinary text files" start with those two bytes—in UTF-8 it's not even legal.

Usually gzip compressed files sport the suffix .gz though. Even gzip(1) itself won't unpack files without it unless you --force it to. You could conceivably use that, but you'd still have to deal with a possible IOError (which you have to in any case).

One problem with your approach is, that gzip.GzipFile() will not throw an exception if you feed it an uncompressed file. Only a later read() will. This means, that you would probably have to implement some of your program logic twice. Ugly.

Answer 2

Is there a cross-platform, usable from Python way to determine if a file is gzip compressed or not?

The accepted answer explains how one can detect a gzip compressed file in general: test if the first two bytes are 1f 8b . However it does not show how to implement it in Python.

Here is one way:

def is_gz_file(filepath):
    with open(filepath, 'rb') as test_f:
        return test_f.read(2) == b'\x1f\x8b'

Answer 3

Testing the magic number of a gzip file is the only reliable way to go. However, as of python3.7 there is no need to mess with comparing bytes yourself anymore. The gzip module will compare the bytes for you and raise an exception if they do not match!

As of python3.7, this works

import gzip
with gzip.open(input_file, 'r') as fh:
    try:
        fh.read(1)
    except OSError:
        print('input_file is not a valid gzip file by OSError')

As of python3.8, this also works:

import gzip
with gzip.open(input_file, 'r') as fh:
    try:
        fh.read(1)
    except gzip.BadGzipFile:
        print('input_file is not a valid gzip file by BadGzipFile')

Answer 4

gzip itself will raise an OSError if it's not a gzipped file.

>>> with gzip.open('README.md', 'rb') as f:
...     f.read()
...
Traceback (most recent call last):
  File "<stdin>", line 2, in <module>
  File "/Users/dennis/.asdf/installs/python/3.6.6/lib/python3.6/gzip.py", line 276, in read
    return self._buffer.read(size)
  File "/Users/dennis/.asdf/installs/python/3.6.6/lib/python3.6/gzip.py", line 463, in read
    if not self._read_gzip_header():
  File "/Users/dennis/.asdf/installs/python/3.6.6/lib/python3.6/gzip.py", line 411, in _read_gzip_header
    raise OSError('Not a gzipped file (%r)' % magic)
OSError: Not a gzipped file (b'# ')

Can combine this approach with some others to increase confidence, such as checking the mimetype or looking for a magic number in the file header (see other answers for an example) and checking the extension.

import pathlib

if '.gz' in pathlib.Path(filepath).suffixes:
   # some more inexpensive checks until confident we can attempt to decompress
   # ...
   try ...
     ...
   except OSError as e:
     ...

Answer 5

Import the mimetypes module. It can automatically guess what kind of file you have, and if it is compressed.

ie

mimetypes.guess_type('blabla.txt.gz')

returns:

('text/plain', 'gzip')

Answer 6

Doesn't seem to work well in python3...

import mimetypes
filename = "./datasets/test"

def file_type(filename):
    type = mimetypes.guess_type(filename)
    return type
print(file_type(filename))

returns (None, None) But from the unix command "File"

:~> file datasets/test datasets/test: gzip compressed data, was "iostat_collection", from Unix, last modified: Thu Jan 29 07:09:34 2015