File type identification using libmagic
libmagic is the library that commonly supports the file command on Unix system, other than Max OSX which has its own implementation. The library handles the loading of database files that describe the magic numbers used to identify various file types, as well as the associated mime types. The library also handles character set detections.
Before installing filemagic, the libmagic library will need to be availabile. To test this is the check for the presence of the file command and/or the libmagic man page.
$ which file $ man libmagic
On Mac OSX, Apple has implemented their own version of the file command. However, libmagic can be installed using homebrew
$ brew install libmagic
After brew finished installing, the test for the libmagic man page should pass.
Now that the presence of libmagic has been confirmed, use pip to install filemagic.
$ pip install filemagic
The magic module should now be availabe from the Python shell.
>>> import magic
The next section will describe how to use the magic.Magic class to identify file types.
The magic module uses ctypes <http://docs.python.org/dev/library/ctypes.html> to wrap the primitives from libmagic in the more user friendly magic.Magic class. This class handles initialization, loading databases and the release of resources.
>>> import magic
To ensure that resources are correctly released by magic.Magic, it’s necessary to either explicitly call close on instances, or use with statement.
>>> with magic.Magic as m: >>> pass
magic.Magic supports context managers which ensures resources are correctly released at the end of the with statements irrespective of any exceptions.
To identify a file from it’s filename, use the id_filename method.
>>> with magic.Magic as m: >>> m.id_filename('setup.py') 'Python script, ASCII text executable'
Similarily to identify a file from a str or buffer that has already been read, use the id_filename method.
>>> with magic.Magic as m: >>> m.desc_buffer('#!/usr/bin/python\n') 'Python script, ASCII text executable'
To identify with mime type, rather than a textual description, pass the magic.MAGIC_MIME_TYPE flag when creating the magic.Magic instance.
>>> with magic.Magic(flags=magic.MAGIC_MIME_TYPE) as m: >>> m.desc_file('setup.py') 'text/x-python'
Similarily, magic.MAGIC_MIME_ENCODING can be passed to return the encoding type.
>>> with magic.Magic(flags=magic.MAGIC_MIME_ENCODING) as m: >>> m.desc_file('setup.py') 'us-ascii'
Unicode and Python3
On both Python2 and Python3, magic.Magic will encode any unicode objects (the default string type for Python3) to byte strings before being passed to libmagic. On Python3, returned strings will be decoded to unicode using the default encoding type. The user should not be concerned whether unicode or bytes are passed to magic.Magic methods. However, the use will need to be aware that return strings are always unicode on Python3 and byte strings on Python2.