Lib/mimetools.py

   1 """Various tools used by MIME-reading or MIME-writing programs."""
   2
   3
   4 import os
   5 import rfc822
   6 import tempfile
   7
   8 __all__ = ["Message","choose_boundary","encode","decode","copyliteral",
   9            "copybinary"]
  10
  11 class Message(rfc822.Message):
  12     """A derived class of rfc822.Message that knows about MIME headers and
  13     contains some hooks for decoding encoded and multipart messages."""
  14
  15     def __init__(self, fp, seekable = 1):
  16         rfc822.Message.__init__(self, fp, seekable)
  17         self.encodingheader = \
  18                 self.getheader('content-transfer-encoding')
  19         self.typeheader = \
  20                 self.getheader('content-type')
  21         self.parsetype()
  22         self.parseplist()
  23
  24     def parsetype(self):
  25         str = self.typeheader
  26         if str is None:
  27             str = 'text/plain'
  28         if ';' in str:
  29             i = str.index(';')
  30             self.plisttext = str[i:]
  31             str = str[:i]
  32         else:
  33             self.plisttext = ''
  34         fields = str.split('/')
  35         for i in range(len(fields)):
  36             fields[i] = fields[i].strip().lower()
  37         self.type = '/'.join(fields)
  38         self.maintype = fields[0]
  39         self.subtype = '/'.join(fields[1:])
  40
  41     def parseplist(self):
  42         str = self.plisttext
  43         self.plist = []
  44         while str[:1] == ';':
  45             str = str[1:]
  46             if ';' in str:
  47                 # XXX Should parse quotes!
  48                 end = str.index(';')
  49             else:
  50                 end = len(str)
  51             f = str[:end]
  52             if '=' in f:
  53                 i = f.index('=')
  54                 f = f[:i].strip().lower() + \
  55                         '=' + f[i+1:].strip()
  56             self.plist.append(f.strip())
  57             str = str[end:]
  58
  59     def getplist(self):
  60         return self.plist
  61
  62     def getparam(self, name):
  63         name = name.lower() + '='
  64         n = len(name)
  65         for p in self.plist:
  66             if p[:n] == name:
  67                 return rfc822.unquote(p[n:])
  68         return None
  69
  70     def getparamnames(self):
  71         result = []
  72         for p in self.plist:
  73             i = p.find('=')
  74             if i >= 0:
  75                 result.append(p[:i].lower())
  76         return result
  77
  78     def getencoding(self):
  79         if self.encodingheader is None:
  80             return '7bit'
  81         return self.encodingheader.lower()
  82
  83     def gettype(self):
  84         return self.type
  85
  86     def getmaintype(self):
  87         return self.maintype
  88
  89     def getsubtype(self):
  90         return self.subtype
  91
  92
  93
  94
  95 # Utility functions
  96 # -----------------
  97
  98 try:
  99     import thread
 100 except ImportError:
 101     import dummy_thread as thread
 102 _counter_lock = thread.allocate_lock()
 103 del thread
 104
 105 _counter = 0
 106 def _get_next_counter():
 107     global _counter
 108     _counter_lock.acquire()
 109     _counter += 1
 110     result = _counter
 111     _counter_lock.release()
 112     return result
 113
 114 _prefix = None
 115
 116 def choose_boundary():
 117     """Return a string usable as a multipart boundary.
 118
 119     The string chosen is unique within a single program run, and
 120     incorporates the user id (if available), process id (if available),
 121     and current time.  So it's very unlikely the returned string appears
 122     in message text, but there's no guarantee.
 123
 124     The boundary contains dots so you have to quote it in the header."""
 125
 126     global _prefix
 127     import time
 128     if _prefix is None:
 129         import socket
 130         hostid = socket.gethostbyname(socket.gethostname())
 131         try:
 132             uid = repr(os.getuid())
 133         except AttributeError:
 134             uid = '1'
 135         try:
 136             pid = repr(os.getpid())
 137         except AttributeError:
 138             pid = '1'
 139         _prefix = hostid + '.' + uid + '.' + pid
 140     return "%s.%.3f.%d" % (_prefix, time.time(), _get_next_counter())
 141
 142
 143 # Subroutines for decoding some common content-transfer-types
 144
 145 def decode(input, output, encoding):
 146     """Decode common content-transfer-encodings (base64, quopri, uuencode)."""
 147     if encoding == 'base64':
 148         import base64
 149         return base64.decode(input, output)
 150     if encoding == 'quoted-printable':
 151         import quopri
 152         return quopri.decode(input, output)
 153     if encoding in ('uuencode', 'x-uuencode', 'uue', 'x-uue'):
 154         import uu
 155         return uu.decode(input, output)
 156     if encoding in ('7bit', '8bit'):
 157         return output.write(input.read())
 158     if encoding in decodetab:
 159         pipethrough(input, decodetab[encoding], output)
 160     else:
 161         raise ValueError, \
 162               'unknown Content-Transfer-Encoding: %s' % encoding
 163
 164 def encode(input, output, encoding):
 165     """Encode common content-transfer-encodings (base64, quopri, uuencode)."""
 166     if encoding == 'base64':
 167         import base64
 168         return base64.encode(input, output)
 169     if encoding == 'quoted-printable':
 170         import quopri
 171         return quopri.encode(input, output, 0)
 172     if encoding in ('uuencode', 'x-uuencode', 'uue', 'x-uue'):
 173         import uu
 174         return uu.encode(input, output)
 175     if encoding in ('7bit', '8bit'):
 176         return output.write(input.read())
 177     if encoding in encodetab:
 178         pipethrough(input, encodetab[encoding], output)
 179     else:
 180         raise ValueError, \
 181               'unknown Content-Transfer-Encoding: %s' % encoding
 182
 183 # The following is no longer used for standard encodings
 184
 185 # XXX This requires that uudecode and mmencode are in $PATH
 186
 187 uudecode_pipe = '''(
 188 TEMP=/tmp/@uu.$$
 189 sed "s%^begin [0-7][0-7]* .*%begin 600 $TEMP%" | uudecode
 190 cat $TEMP
 191 rm $TEMP
 192 )'''
 193
 194 decodetab = {
 195         'uuencode':             uudecode_pipe,
 196         'x-uuencode':           uudecode_pipe,
 197         'uue':                  uudecode_pipe,
 198         'x-uue':                uudecode_pipe,
 199         'quoted-printable':     'mmencode -u -q',
 200         'base64':               'mmencode -u -b',
 201 }
 202
 203 encodetab = {
 204         'x-uuencode':           'uuencode tempfile',
 205         'uuencode':             'uuencode tempfile',
 206         'x-uue':                'uuencode tempfile',
 207         'uue':                  'uuencode tempfile',
 208         'quoted-printable':     'mmencode -q',
 209         'base64':               'mmencode -b',
 210 }
 211
 212 def pipeto(input, command):
 213     pipe = os.popen(command, 'w')
 214     copyliteral(input, pipe)
 215     pipe.close()
 216
 217 def pipethrough(input, command, output):
 218     (fd, tempname) = tempfile.mkstemp()
 219     temp = os.fdopen(fd, 'w')
 220     copyliteral(input, temp)
 221     temp.close()
 222     pipe = os.popen(command + ' <' + tempname, 'r')
 223     copybinary(pipe, output)
 224     pipe.close()
 225     os.unlink(tempname)
 226
 227 def copyliteral(input, output):
 228     while 1:
 229         line = input.readline()
 230         if not line: break
 231         output.write(line)
 232
 233 def copybinary(input, output):
 234     BUFSIZE = 8192
 235     while 1:
 236         line = input.read(BUFSIZE)
 237         if not line: break
 238         output.write(line)