kopano-core/ECtools/search/kopano_search/magic.py

"""
magic is a wrapper around the libmagic file identification library.

Copyright Adam Hupp, adam@hupp.org

license MIT

Usage:

>>> import magic
>>> magic.from_file("testdata/test.pdf")
'PDF document, version 1.2'
>>> magic.from_file("testdata/test.pdf", mime=True)
'application/pdf'
>>> magic.from_buffer(open("testdata/test.pdf").read(1024))
'PDF document, version 1.2'
>>>


"""

import sys
import glob
import os.path
import ctypes
import ctypes.util
import threading

from ctypes import c_char_p, c_int, c_size_t, c_void_p

class MagicException(Exception): pass

class Magic:
    """
    Magic is a wrapper around the libmagic C library.

    """

    def __init__(self, mime=False, magic_file=None, mime_encoding=False,
                 keep_going=False):
        """
        Create a new libmagic wrapper.

        mime - if True, mimetypes are returned instead of textual descriptions
        mime_encoding - if True, codec is returned
        magic_file - use a mime database other than the system default
        keep_going - don't stop at the first match, keep going
        """
        flags = MAGIC_NONE
        if mime:
            flags |= MAGIC_MIME
        elif mime_encoding:
            flags |= MAGIC_MIME_ENCODING
        if keep_going:
            flags |= MAGIC_CONTINUE

        self.cookie = magic_open(flags)

        magic_load(self.cookie, magic_file)

        self.thread = threading.currentThread()

    def from_buffer(self, buf):
        """
        Identify the contents of `buf`
        """
        self._thread_check()
        return magic_buffer(self.cookie, buf)

    def from_file(self, filename):
        """
        Identify the contents of file `filename`
        raises IOError if the file does not exist
        """
        self._thread_check()
        if not os.path.exists(filename):
            raise IOError("File does not exist: " + filename)

        return magic_file(self.cookie, filename)

    def _thread_check(self):
        if self.thread != threading.currentThread():
            raise Exception('attempting to use libmagic on multiple threads will '
                            'end in SEGV.  Prefer to use the module functions '
                            'from_file or from_buffer, or carefully manage direct '
                            'use of the Magic class')

    def __del__(self):
        # no _thread_check here because there can be no other
        # references to this object at this point.

        # during shutdown magic_close may have been cleared already so
        # make sure it exists before using it.

        # the self.cookie check should be unnessary and was an
        # incorrect fix for a threading problem, however I'm leaving
        # it in because it's harmless and I'm slightly afraid to
        # remove it.
        if self.cookie and magic_close:
            magic_close(self.cookie)
            self.cookie = None


instances = threading.local()

def _get_magic_type(mime):
    i = instances.__dict__.get(mime)
    if i is None:
        i = instances.__dict__[mime] = Magic(mime=mime)
    return i

def from_file(filename, mime=False):
    """"
    Accepts a filename and returns the detected filetype.  Return
    value is the mimetype if mime=True, otherwise a human readable
    name.
    
    >>> magic.from_file("testdata/test.pdf", mime=True)
    'application/pdf'
    """
    m = _get_magic_type(mime)
    return m.from_file(filename)

def from_buffer(buffer, mime=False):
    """
    Accepts a binary string and returns the detected filetype.  Return
    value is the mimetype if mime=True, otherwise a human readable
    name.

    >>> magic.from_buffer(open("testdata/test.pdf").read(1024))
    'PDF document, version 1.2'
    """
    m = _get_magic_type(mime)
    return m.from_buffer(buffer)




libmagic = None
# Let's try to find magic or magic1
dll = ctypes.util.find_library('magic') or ctypes.util.find_library('magic1')

# This is necessary because find_library returns None if it doesn't find the library
if dll:
    libmagic = ctypes.CDLL(dll)

if not libmagic or not libmagic._name:
    import sys
    platform_to_lib = {'darwin': ['/opt/local/lib/libmagic.dylib',
                                  '/usr/local/lib/libmagic.dylib'] +
                       # Assumes there will only be one version installed
                       glob.glob('/usr/local/Cellar/libmagic/*/lib/libmagic.dylib'),
                       'win32':  ['magic1.dll']}
    for dll in platform_to_lib.get(sys.platform, []):
        try:
            libmagic = ctypes.CDLL(dll)
            break
        except OSError:
            pass

if not libmagic or not libmagic._name:
    # It is better to raise an ImportError since we are importing magic module
    raise ImportError('failed to find libmagic.  Check your installation')

magic_t = ctypes.c_void_p

def errorcheck_null(result, func, args):
    if result is None:
        err = magic_error(args[0])
        raise MagicException(err)
    else:
        return result

def errorcheck_negative_one(result, func, args):
    if result is -1:
        err = magic_error(args[0])
        raise MagicException(err)
    else:
        return result
    

def coerce_filename(filename):
    if filename is None:
        return None
    return filename.encode(sys.getfilesystemencoding())

magic_open = libmagic.magic_open
magic_open.restype = magic_t
magic_open.argtypes = [c_int]

magic_close = libmagic.magic_close
magic_close.restype = None
magic_close.argtypes = [magic_t]

magic_error = libmagic.magic_error
magic_error.restype = c_char_p
magic_error.argtypes = [magic_t]

magic_errno = libmagic.magic_errno
magic_errno.restype = c_int
magic_errno.argtypes = [magic_t]

_magic_file = libmagic.magic_file
_magic_file.restype = c_char_p
_magic_file.argtypes = [magic_t, c_char_p]
_magic_file.errcheck = errorcheck_null

def magic_file(cookie, filename):
    return _magic_file(cookie, coerce_filename(filename))

_magic_buffer = libmagic.magic_buffer
_magic_buffer.restype = c_char_p
_magic_buffer.argtypes = [magic_t, c_void_p, c_size_t]
_magic_buffer.errcheck = errorcheck_null

def magic_buffer(cookie, buf):
    return _magic_buffer(cookie, buf, len(buf))


_magic_load = libmagic.magic_load
_magic_load.restype = c_int
_magic_load.argtypes = [magic_t, c_char_p]
_magic_load.errcheck = errorcheck_negative_one

def magic_load(cookie, filename):
    return _magic_load(cookie, coerce_filename(filename))

magic_setflags = libmagic.magic_setflags
magic_setflags.restype = c_int
magic_setflags.argtypes = [magic_t, c_int]

magic_check = libmagic.magic_check
magic_check.restype = c_int
magic_check.argtypes = [magic_t, c_char_p]

magic_compile = libmagic.magic_compile
magic_compile.restype = c_int
magic_compile.argtypes = [magic_t, c_char_p]



MAGIC_NONE = 0x000000 # No flags

MAGIC_DEBUG = 0x000001 # Turn on debugging

MAGIC_SYMLINK = 0x000002 # Follow symlinks

MAGIC_COMPRESS = 0x000004 # Check inside compressed files

MAGIC_DEVICES = 0x000008 # Look at the contents of devices

MAGIC_MIME = 0x000010 # Return a mime string

MAGIC_MIME_ENCODING = 0x000400 # Return the MIME encoding

MAGIC_CONTINUE = 0x000020 # Return all matches

MAGIC_CHECK = 0x000040 # Print warnings to stderr

MAGIC_PRESERVE_ATIME = 0x000080 # Restore access time on exit

MAGIC_RAW = 0x000100 # Don't translate unprintable chars

MAGIC_ERROR = 0x000200 # Handle ENOENT etc as real errors

MAGIC_NO_CHECK_COMPRESS = 0x001000 # Don't check for compressed files

MAGIC_NO_CHECK_TAR = 0x002000 # Don't check for tar files

MAGIC_NO_CHECK_SOFT = 0x004000 # Don't check magic entries

MAGIC_NO_CHECK_APPTYPE = 0x008000 # Don't check application type

MAGIC_NO_CHECK_ELF = 0x010000 # Don't check for elf details

MAGIC_NO_CHECK_ASCII = 0x020000 # Don't check for ascii files

MAGIC_NO_CHECK_TROFF = 0x040000 # Don't check ascii/troff

MAGIC_NO_CHECK_FORTRAN = 0x080000 # Don't check ascii/fortran

MAGIC_NO_CHECK_TOKENS = 0x100000 # Don't check ascii/tokens