Lib/cgi.py

   1 #! /usr/local/bin/python
   2
   3 # NOTE: the above "/usr/local/bin/python" is NOT a mistake.  It is
   4 # intentionally NOT "/usr/bin/env python".  On many systems
   5 # (e.g. Solaris), /usr/local/bin is not in $PATH as passed to CGI
   6 # scripts, and /usr/local/bin is the default directory where Python is
   7 # installed, so /usr/bin/env would be unable to find python.  Granted,
   8 # binary installations by Linux vendors often install Python in
   9 # /usr/bin.  So let those vendors patch cgi.py to match their choice
  10 # of installation.
  11
  12 """Support module for CGI (Common Gateway Interface) scripts.
  13
  14 This module defines a number of utilities for use by CGI scripts
  15 written in Python.
  16 """
  17
  18 # XXX Perhaps there should be a slimmed version that doesn't contain
  19 # all those backwards compatible and debugging classes and functions?
  20
  21 # History
  22 # -------
  23 #
  24 # Michael McLay started this module.  Steve Majewski changed the
  25 # interface to SvFormContentDict and FormContentDict.  The multipart
  26 # parsing was inspired by code submitted by Andreas Paepcke.  Guido van
  27 # Rossum rewrote, reformatted and documented the module and is currently
  28 # responsible for its maintenance.
  29 #
  30
  31 __version__ = "2.6"
  32
  33
  34 # Imports
  35 # =======
  36
  37 from operator import attrgetter
  38 import sys
  39 import os
  40 import urllib
  41 import UserDict
  42 import urlparse
  43
  44 from warnings import filterwarnings, catch_warnings, warn
  45 with catch_warnings():
  46     if sys.py3kwarning:
  47         filterwarnings("ignore", ".*mimetools has been removed",
  48                         DeprecationWarning)
  49     import mimetools
  50     if sys.py3kwarning:
  51         filterwarnings("ignore", ".*rfc822 has been removed", DeprecationWarning)
  52     import rfc822
  53
  54 try:
  55     from cStringIO import StringIO
  56 except ImportError:
  57     from StringIO import StringIO
  58
  59 __all__ = ["MiniFieldStorage", "FieldStorage", "FormContentDict",
  60            "SvFormContentDict", "InterpFormContentDict", "FormContent",
  61            "parse", "parse_qs", "parse_qsl", "parse_multipart",
  62            "parse_header", "print_exception", "print_environ",
  63            "print_form", "print_directory", "print_arguments",
  64            "print_environ_usage", "escape"]
  65
  66 # Logging support
  67 # ===============
  68
  69 logfile = ""            # Filename to log to, if not empty
  70 logfp = None            # File object to log to, if not None
  71
  72 def initlog(*allargs):
  73     """Write a log message, if there is a log file.
  74
  75     Even though this function is called initlog(), you should always
  76     use log(); log is a variable that is set either to initlog
  77     (initially), to dolog (once the log file has been opened), or to
  78     nolog (when logging is disabled).
  79
  80     The first argument is a format string; the remaining arguments (if
  81     any) are arguments to the % operator, so e.g.
  82         log("%s: %s", "a", "b")
  83     will write "a: b" to the log file, followed by a newline.
  84
  85     If the global logfp is not None, it should be a file object to
  86     which log data is written.
  87
  88     If the global logfp is None, the global logfile may be a string
  89     giving a filename to open, in append mode.  This file should be
  90     world writable!!!  If the file can't be opened, logging is
  91     silently disabled (since there is no safe place where we could
  92     send an error message).
  93
  94     """
  95     global logfp, log
  96     if logfile and not logfp:
  97         try:
  98             logfp = open(logfile, "a")
  99         except IOError:
 100             pass
 101     if not logfp:
 102         log = nolog
 103     else:
 104         log = dolog
 105     log(*allargs)
 106
 107 def dolog(fmt, *args):
 108     """Write a log message to the log file.  See initlog() for docs."""
 109     logfp.write(fmt%args + "\n")
 110
 111 def nolog(*allargs):
 112     """Dummy function, assigned to log when logging is disabled."""
 113     pass
 114
 115 log = initlog           # The current logging function
 116
 117
 118 # Parsing functions
 119 # =================
 120
 121 # Maximum input we will accept when REQUEST_METHOD is POST
 122 # 0 ==> unlimited input
 123 maxlen = 0
 124
 125 def parse(fp=None, environ=os.environ, keep_blank_values=0, strict_parsing=0):
 126     """Parse a query in the environment or from a file (default stdin)
 127
 128         Arguments, all optional:
 129
 130         fp              : file pointer; default: sys.stdin
 131
 132         environ         : environment dictionary; default: os.environ
 133
 134         keep_blank_values: flag indicating whether blank values in
 135             URL encoded forms should be treated as blank strings.
 136             A true value indicates that blanks should be retained as
 137             blank strings.  The default false value indicates that
 138             blank values are to be ignored and treated as if they were
 139             not included.
 140
 141         strict_parsing: flag indicating what to do with parsing errors.
 142             If false (the default), errors are silently ignored.
 143             If true, errors raise a ValueError exception.
 144     """
 145     if fp is None:
 146         fp = sys.stdin
 147     if not 'REQUEST_METHOD' in environ:
 148         environ['REQUEST_METHOD'] = 'GET'       # For testing stand-alone
 149     if environ['REQUEST_METHOD'] == 'POST':
 150         ctype, pdict = parse_header(environ['CONTENT_TYPE'])
 151         if ctype == 'multipart/form-data':
 152             return parse_multipart(fp, pdict)
 153         elif ctype == 'application/x-www-form-urlencoded':
 154             clength = int(environ['CONTENT_LENGTH'])
 155             if maxlen and clength > maxlen:
 156                 raise ValueError, 'Maximum content length exceeded'
 157             qs = fp.read(clength)
 158         else:
 159             qs = ''                     # Unknown content-type
 160         if 'QUERY_STRING' in environ:
 161             if qs: qs = qs + '&'
 162             qs = qs + environ['QUERY_STRING']
 163         elif sys.argv[1:]:
 164             if qs: qs = qs + '&'
 165             qs = qs + sys.argv[1]
 166         environ['QUERY_STRING'] = qs    # XXX Shouldn't, really
 167     elif 'QUERY_STRING' in environ:
 168         qs = environ['QUERY_STRING']
 169     else:
 170         if sys.argv[1:]:
 171             qs = sys.argv[1]
 172         else:
 173             qs = ""
 174         environ['QUERY_STRING'] = qs    # XXX Shouldn't, really
 175     return parse_qs(qs, keep_blank_values, strict_parsing)
 176
 177
 178 # parse query string function called from urlparse,
 179 # this is done in order to maintain backward compatiblity.
 180
 181 def parse_qs(qs, keep_blank_values=0, strict_parsing=0):
 182     """Parse a query given as a string argument."""
 183     warn("cgi.parse_qs is deprecated, use urlparse.parse_qs \
 184             instead", PendingDeprecationWarning, 2)
 185     return urlparse.parse_qs(qs, keep_blank_values, strict_parsing)
 186
 187
 188 def parse_qsl(qs, keep_blank_values=0, strict_parsing=0):
 189     """Parse a query given as a string argument."""
 190     warn("cgi.parse_qsl is deprecated, use urlparse.parse_qsl instead",
 191          PendingDeprecationWarning, 2)
 192     return urlparse.parse_qsl(qs, keep_blank_values, strict_parsing)
 193
 194 def parse_multipart(fp, pdict):
 195     """Parse multipart input.
 196
 197     Arguments:
 198     fp   : input file
 199     pdict: dictionary containing other parameters of content-type header
 200
 201     Returns a dictionary just like parse_qs(): keys are the field names, each
 202     value is a list of values for that field.  This is easy to use but not
 203     much good if you are expecting megabytes to be uploaded -- in that case,
 204     use the FieldStorage class instead which is much more flexible.  Note
 205     that content-type is the raw, unparsed contents of the content-type
 206     header.
 207
 208     XXX This does not parse nested multipart parts -- use FieldStorage for
 209     that.
 210
 211     XXX This should really be subsumed by FieldStorage altogether -- no
 212     point in having two implementations of the same parsing algorithm.
 213     Also, FieldStorage protects itself better against certain DoS attacks
 214     by limiting the size of the data read in one chunk.  The API here
 215     does not support that kind of protection.  This also affects parse()
 216     since it can call parse_multipart().
 217
 218     """
 219     boundary = ""
 220     if 'boundary' in pdict:
 221         boundary = pdict['boundary']
 222     if not valid_boundary(boundary):
 223         raise ValueError,  ('Invalid boundary in multipart form: %r'
 224                             % (boundary,))
 225
 226     nextpart = "--" + boundary
 227     lastpart = "--" + boundary + "--"
 228     partdict = {}
 229     terminator = ""
 230
 231     while terminator != lastpart:
 232         bytes = -1
 233         data = None
 234         if terminator:
 235             # At start of next part.  Read headers first.
 236             headers = mimetools.Message(fp)
 237             clength = headers.getheader('content-length')
 238             if clength:
 239                 try:
 240                     bytes = int(clength)
 241                 except ValueError:
 242                     pass
 243             if bytes > 0:
 244                 if maxlen and bytes > maxlen:
 245                     raise ValueError, 'Maximum content length exceeded'
 246                 data = fp.read(bytes)
 247             else:
 248                 data = ""
 249         # Read lines until end of part.
 250         lines = []
 251         while 1:
 252             line = fp.readline()
 253             if not line:
 254                 terminator = lastpart # End outer loop
 255                 break
 256             if line[:2] == "--":
 257                 terminator = line.strip()
 258                 if terminator in (nextpart, lastpart):
 259                     break
 260             lines.append(line)
 261         # Done with part.
 262         if data is None:
 263             continue
 264         if bytes < 0:
 265             if lines:
 266                 # Strip final line terminator
 267                 line = lines[-1]
 268                 if line[-2:] == "\r\n":
 269                     line = line[:-2]
 270                 elif line[-1:] == "\n":
 271                     line = line[:-1]
 272                 lines[-1] = line
 273                 data = "".join(lines)
 274         line = headers['content-disposition']
 275         if not line:
 276             continue
 277         key, params = parse_header(line)
 278         if key != 'form-data':
 279             continue
 280         if 'name' in params:
 281             name = params['name']
 282         else:
 283             continue
 284         if name in partdict:
 285             partdict[name].append(data)
 286         else:
 287             partdict[name] = [data]
 288
 289     return partdict
 290
 291
 292 def _parseparam(s):
 293     while s[:1] == ';':
 294         s = s[1:]
 295         end = s.find(';')
 296         while end > 0 and s.count('"', 0, end) % 2:
 297             end = s.find(';', end + 1)
 298         if end < 0:
 299             end = len(s)
 300         f = s[:end]
 301         yield f.strip()
 302         s = s[end:]
 303
 304 def parse_header(line):
 305     """Parse a Content-type like header.
 306
 307     Return the main content-type and a dictionary of options.
 308
 309     """
 310     parts = _parseparam(';' + line)
 311     key = parts.next()
 312     pdict = {}
 313     for p in parts:
 314         i = p.find('=')
 315         if i >= 0:
 316             name = p[:i].strip().lower()
 317             value = p[i+1:].strip()
 318             if len(value) >= 2 and value[0] == value[-1] == '"':
 319                 value = value[1:-1]
 320                 value = value.replace('\\\\', '\\').replace('\\"', '"')
 321             pdict[name] = value
 322     return key, pdict
 323
 324
 325 # Classes for field storage
 326 # =========================
 327
 328 class MiniFieldStorage:
 329
 330     """Like FieldStorage, for use when no file uploads are possible."""
 331
 332     # Dummy attributes
 333     filename = None
 334     list = None
 335     type = None
 336     file = None
 337     type_options = {}
 338     disposition = None
 339     disposition_options = {}
 340     headers = {}
 341
 342     def __init__(self, name, value):
 343         """Constructor from field name and value."""
 344         self.name = name
 345         self.value = value
 346         # self.file = StringIO(value)
 347
 348     def __repr__(self):
 349         """Return printable representation."""
 350         return "MiniFieldStorage(%r, %r)" % (self.name, self.value)
 351
 352
 353 class FieldStorage:
 354
 355     """Store a sequence of fields, reading multipart/form-data.
 356
 357     This class provides naming, typing, files stored on disk, and
 358     more.  At the top level, it is accessible like a dictionary, whose
 359     keys are the field names.  (Note: None can occur as a field name.)
 360     The items are either a Python list (if there's multiple values) or
 361     another FieldStorage or MiniFieldStorage object.  If it's a single
 362     object, it has the following attributes:
 363
 364     name: the field name, if specified; otherwise None
 365
 366     filename: the filename, if specified; otherwise None; this is the
 367         client side filename, *not* the file name on which it is
 368         stored (that's a temporary file you don't deal with)
 369
 370     value: the value as a *string*; for file uploads, this
 371         transparently reads the file every time you request the value
 372
 373     file: the file(-like) object from which you can read the data;
 374         None if the data is stored a simple string
 375
 376     type: the content-type, or None if not specified
 377
 378     type_options: dictionary of options specified on the content-type
 379         line
 380
 381     disposition: content-disposition, or None if not specified
 382
 383     disposition_options: dictionary of corresponding options
 384
 385     headers: a dictionary(-like) object (sometimes rfc822.Message or a
 386         subclass thereof) containing *all* headers
 387
 388     The class is subclassable, mostly for the purpose of overriding
 389     the make_file() method, which is called internally to come up with
 390     a file open for reading and writing.  This makes it possible to
 391     override the default choice of storing all files in a temporary
 392     directory and unlinking them as soon as they have been opened.
 393
 394     """
 395
 396     def __init__(self, fp=None, headers=None, outerboundary="",
 397                  environ=os.environ, keep_blank_values=0, strict_parsing=0):
 398         """Constructor.  Read multipart/* until last part.
 399
 400         Arguments, all optional:
 401
 402         fp              : file pointer; default: sys.stdin
 403             (not used when the request method is GET)
 404
 405         headers         : header dictionary-like object; default:
 406             taken from environ as per CGI spec
 407
 408         outerboundary   : terminating multipart boundary
 409             (for internal use only)
 410
 411         environ         : environment dictionary; default: os.environ
 412
 413         keep_blank_values: flag indicating whether blank values in
 414             URL encoded forms should be treated as blank strings.
 415             A true value indicates that blanks should be retained as
 416             blank strings.  The default false value indicates that
 417             blank values are to be ignored and treated as if they were
 418             not included.
 419
 420         strict_parsing: flag indicating what to do with parsing errors.
 421             If false (the default), errors are silently ignored.
 422             If true, errors raise a ValueError exception.
 423
 424         """
 425         method = 'GET'
 426         self.keep_blank_values = keep_blank_values
 427         self.strict_parsing = strict_parsing
 428         if 'REQUEST_METHOD' in environ:
 429             method = environ['REQUEST_METHOD'].upper()
 430         self.qs_on_post = None
 431         if method == 'GET' or method == 'HEAD':
 432             if 'QUERY_STRING' in environ:
 433                 qs = environ['QUERY_STRING']
 434             elif sys.argv[1:]:
 435                 qs = sys.argv[1]
 436             else:
 437                 qs = ""
 438             fp = StringIO(qs)
 439             if headers is None:
 440                 headers = {'content-type':
 441                            "application/x-www-form-urlencoded"}
 442         if headers is None:
 443             headers = {}
 444             if method == 'POST':
 445                 # Set default content-type for POST to what's traditional
 446                 headers['content-type'] = "application/x-www-form-urlencoded"
 447             if 'CONTENT_TYPE' in environ:
 448                 headers['content-type'] = environ['CONTENT_TYPE']
 449             if 'QUERY_STRING' in environ:
 450                 self.qs_on_post = environ['QUERY_STRING']
 451             if 'CONTENT_LENGTH' in environ:
 452                 headers['content-length'] = environ['CONTENT_LENGTH']
 453         self.fp = fp or sys.stdin
 454         self.headers = headers
 455         self.outerboundary = outerboundary
 456
 457         # Process content-disposition header
 458         cdisp, pdict = "", {}
 459         if 'content-disposition' in self.headers:
 460             cdisp, pdict = parse_header(self.headers['content-disposition'])
 461         self.disposition = cdisp
 462         self.disposition_options = pdict
 463         self.name = None
 464         if 'name' in pdict:
 465             self.name = pdict['name']
 466         self.filename = None
 467         if 'filename' in pdict:
 468             self.filename = pdict['filename']
 469
 470         # Process content-type header
 471         #
 472         # Honor any existing content-type header.  But if there is no
 473         # content-type header, use some sensible defaults.  Assume
 474         # outerboundary is "" at the outer level, but something non-false
 475         # inside a multi-part.  The default for an inner part is text/plain,
 476         # but for an outer part it should be urlencoded.  This should catch
 477         # bogus clients which erroneously forget to include a content-type
 478         # header.
 479         #
 480         # See below for what we do if there does exist a content-type header,
 481         # but it happens to be something we don't understand.
 482         if 'content-type' in self.headers:
 483             ctype, pdict = parse_header(self.headers['content-type'])
 484         elif self.outerboundary or method != 'POST':
 485             ctype, pdict = "text/plain", {}
 486         else:
 487             ctype, pdict = 'application/x-www-form-urlencoded', {}
 488         self.type = ctype
 489         self.type_options = pdict
 490         self.innerboundary = ""
 491         if 'boundary' in pdict:
 492             self.innerboundary = pdict['boundary']
 493         clen = -1
 494         if 'content-length' in self.headers:
 495             try:
 496                 clen = int(self.headers['content-length'])
 497             except ValueError:
 498                 pass
 499             if maxlen and clen > maxlen:
 500                 raise ValueError, 'Maximum content length exceeded'
 501         self.length = clen
 502
 503         self.list = self.file = None
 504         self.done = 0
 505         if ctype == 'application/x-www-form-urlencoded':
 506             self.read_urlencoded()
 507         elif ctype[:10] == 'multipart/':
 508             self.read_multi(environ, keep_blank_values, strict_parsing)
 509         else:
 510             self.read_single()
 511
 512     def __repr__(self):
 513         """Return a printable representation."""
 514         return "FieldStorage(%r, %r, %r)" % (
 515                 self.name, self.filename, self.value)
 516
 517     def __iter__(self):
 518         return iter(self.keys())
 519
 520     def __getattr__(self, name):
 521         if name != 'value':
 522             raise AttributeError, name
 523         if self.file:
 524             self.file.seek(0)
 525             value = self.file.read()
 526             self.file.seek(0)
 527         elif self.list is not None:
 528             value = self.list
 529         else:
 530             value = None
 531         return value
 532
 533     def __getitem__(self, key):
 534         """Dictionary style indexing."""
 535         if self.list is None:
 536             raise TypeError, "not indexable"
 537         found = []
 538         for item in self.list:
 539             if item.name == key: found.append(item)
 540         if not found:
 541             raise KeyError, key
 542         if len(found) == 1:
 543             return found[0]
 544         else:
 545             return found
 546
 547     def getvalue(self, key, default=None):
 548         """Dictionary style get() method, including 'value' lookup."""
 549         if key in self:
 550             value = self[key]
 551             if type(value) is type([]):
 552                 return map(attrgetter('value'), value)
 553             else:
 554                 return value.value
 555         else:
 556             return default
 557
 558     def getfirst(self, key, default=None):
 559         """ Return the first value received."""
 560         if key in self:
 561             value = self[key]
 562             if type(value) is type([]):
 563                 return value[0].value
 564             else:
 565                 return value.value
 566         else:
 567             return default
 568
 569     def getlist(self, key):
 570         """ Return list of received values."""
 571         if key in self:
 572             value = self[key]
 573             if type(value) is type([]):
 574                 return map(attrgetter('value'), value)
 575             else:
 576                 return [value.value]
 577         else:
 578             return []
 579
 580     def keys(self):
 581         """Dictionary style keys() method."""
 582         if self.list is None:
 583             raise TypeError, "not indexable"
 584         return list(set(item.name for item in self.list))
 585
 586     def has_key(self, key):
 587         """Dictionary style has_key() method."""
 588         if self.list is None:
 589             raise TypeError, "not indexable"
 590         return any(item.name == key for item in self.list)
 591
 592     def __contains__(self, key):
 593         """Dictionary style __contains__ method."""
 594         if self.list is None:
 595             raise TypeError, "not indexable"
 596         return any(item.name == key for item in self.list)
 597
 598     def __len__(self):
 599         """Dictionary style len(x) support."""
 600         return len(self.keys())
 601
 602     def __nonzero__(self):
 603         return bool(self.list)
 604
 605     def read_urlencoded(self):
 606         """Internal: read data in query string format."""
 607         qs = self.fp.read(self.length)
 608         if self.qs_on_post:
 609             qs += '&' + self.qs_on_post
 610         self.list = list = []
 611         for key, value in urlparse.parse_qsl(qs, self.keep_blank_values,
 612                                             self.strict_parsing):
 613             list.append(MiniFieldStorage(key, value))
 614         self.skip_lines()
 615
 616     FieldStorageClass = None
 617
 618     def read_multi(self, environ, keep_blank_values, strict_parsing):
 619         """Internal: read a part that is itself multipart."""
 620         ib = self.innerboundary
 621         if not valid_boundary(ib):
 622             raise ValueError, 'Invalid boundary in multipart form: %r' % (ib,)
 623         self.list = []
 624         if self.qs_on_post:
 625             for key, value in urlparse.parse_qsl(self.qs_on_post,
 626                                 self.keep_blank_values, self.strict_parsing):
 627                 self.list.append(MiniFieldStorage(key, value))
 628             FieldStorageClass = None
 629
 630         klass = self.FieldStorageClass or self.__class__
 631         part = klass(self.fp, {}, ib,
 632                      environ, keep_blank_values, strict_parsing)
 633         # Throw first part away
 634         while not part.done:
 635             headers = rfc822.Message(self.fp)
 636             part = klass(self.fp, headers, ib,
 637                          environ, keep_blank_values, strict_parsing)
 638             self.list.append(part)
 639         self.skip_lines()
 640
 641     def read_single(self):
 642         """Internal: read an atomic part."""
 643         if self.length >= 0:
 644             self.read_binary()
 645             self.skip_lines()
 646         else:
 647             self.read_lines()
 648         self.file.seek(0)
 649
 650     bufsize = 8*1024            # I/O buffering size for copy to file
 651
 652     def read_binary(self):
 653         """Internal: read binary data."""
 654         self.file = self.make_file('b')
 655         todo = self.length
 656         if todo >= 0:
 657             while todo > 0:
 658                 data = self.fp.read(min(todo, self.bufsize))
 659                 if not data:
 660                     self.done = -1
 661                     break
 662                 self.file.write(data)
 663                 todo = todo - len(data)
 664
 665     def read_lines(self):
 666         """Internal: read lines until EOF or outerboundary."""
 667         self.file = self.__file = StringIO()
 668         if self.outerboundary:
 669             self.read_lines_to_outerboundary()
 670         else:
 671             self.read_lines_to_eof()
 672
 673     def __write(self, line):
 674         if self.__file is not None:
 675             if self.__file.tell() + len(line) > 1000:
 676                 self.file = self.make_file('')
 677                 self.file.write(self.__file.getvalue())
 678                 self.__file = None
 679         self.file.write(line)
 680
 681     def read_lines_to_eof(self):
 682         """Internal: read lines until EOF."""
 683         while 1:
 684             line = self.fp.readline(1<<16)
 685             if not line:
 686                 self.done = -1
 687                 break
 688             self.__write(line)
 689
 690     def read_lines_to_outerboundary(self):
 691         """Internal: read lines until outerboundary."""
 692         next = "--" + self.outerboundary
 693         last = next + "--"
 694         delim = ""
 695         last_line_lfend = True
 696         while 1:
 697             line = self.fp.readline(1<<16)
 698             if not line:
 699                 self.done = -1
 700                 break
 701             if line[:2] == "--" and last_line_lfend:
 702                 strippedline = line.strip()
 703                 if strippedline == next:
 704                     break
 705                 if strippedline == last:
 706                     self.done = 1
 707                     break
 708             odelim = delim
 709             if line[-2:] == "\r\n":
 710                 delim = "\r\n"
 711                 line = line[:-2]
 712                 last_line_lfend = True
 713             elif line[-1] == "\n":
 714                 delim = "\n"
 715                 line = line[:-1]
 716                 last_line_lfend = True
 717             else:
 718                 delim = ""
 719                 last_line_lfend = False
 720             self.__write(odelim + line)
 721
 722     def skip_lines(self):
 723         """Internal: skip lines until outer boundary if defined."""
 724         if not self.outerboundary or self.done:
 725             return
 726         next = "--" + self.outerboundary
 727         last = next + "--"
 728         last_line_lfend = True
 729         while 1:
 730             line = self.fp.readline(1<<16)
 731             if not line:
 732                 self.done = -1
 733                 break
 734             if line[:2] == "--" and last_line_lfend:
 735                 strippedline = line.strip()
 736                 if strippedline == next:
 737                     break
 738                 if strippedline == last:
 739                     self.done = 1
 740                     break
 741             last_line_lfend = line.endswith('\n')
 742
 743     def make_file(self, binary=None):
 744         """Overridable: return a readable & writable file.
 745
 746         The file will be used as follows:
 747         - data is written to it
 748         - seek(0)
 749         - data is read from it
 750
 751         The 'binary' argument is unused -- the file is always opened
 752         in binary mode.
 753
 754         This version opens a temporary file for reading and writing,
 755         and immediately deletes (unlinks) it.  The trick (on Unix!) is
 756         that the file can still be used, but it can't be opened by
 757         another process, and it will automatically be deleted when it
 758         is closed or when the current process terminates.
 759
 760         If you want a more permanent file, you derive a class which
 761         overrides this method.  If you want a visible temporary file
 762         that is nevertheless automatically deleted when the script
 763         terminates, try defining a __del__ method in a derived class
 764         which unlinks the temporary files you have created.
 765
 766         """
 767         import tempfile
 768         return tempfile.TemporaryFile("w+b")
 769
 770
 771
 772 # Backwards Compatibility Classes
 773 # ===============================
 774
 775 class FormContentDict(UserDict.UserDict):
 776     """Form content as dictionary with a list of values per field.
 777
 778     form = FormContentDict()
 779
 780     form[key] -> [value, value, ...]
 781     key in form -> Boolean
 782     form.keys() -> [key, key, ...]
 783     form.values() -> [[val, val, ...], [val, val, ...], ...]
 784     form.items() ->  [(key, [val, val, ...]), (key, [val, val, ...]), ...]
 785     form.dict == {key: [val, val, ...], ...}
 786
 787     """
 788     def __init__(self, environ=os.environ, keep_blank_values=0, strict_parsing=0):
 789         self.dict = self.data = parse(environ=environ,
 790                                       keep_blank_values=keep_blank_values,
 791                                       strict_parsing=strict_parsing)
 792         self.query_string = environ['QUERY_STRING']
 793
 794
 795 class SvFormContentDict(FormContentDict):
 796     """Form content as dictionary expecting a single value per field.
 797
 798     If you only expect a single value for each field, then form[key]
 799     will return that single value.  It will raise an IndexError if
 800     that expectation is not true.  If you expect a field to have
 801     possible multiple values, than you can use form.getlist(key) to
 802     get all of the values.  values() and items() are a compromise:
 803     they return single strings where there is a single value, and
 804     lists of strings otherwise.
 805
 806     """
 807     def __getitem__(self, key):
 808         if len(self.dict[key]) > 1:
 809             raise IndexError, 'expecting a single value'
 810         return self.dict[key][0]
 811     def getlist(self, key):
 812         return self.dict[key]
 813     def values(self):
 814         result = []
 815         for value in self.dict.values():
 816             if len(value) == 1:
 817                 result.append(value[0])
 818             else: result.append(value)
 819         return result
 820     def items(self):
 821         result = []
 822         for key, value in self.dict.items():
 823             if len(value) == 1:
 824                 result.append((key, value[0]))
 825             else: result.append((key, value))
 826         return result
 827
 828
 829 class InterpFormContentDict(SvFormContentDict):
 830     """This class is present for backwards compatibility only."""
 831     def __getitem__(self, key):
 832         v = SvFormContentDict.__getitem__(self, key)
 833         if v[0] in '0123456789+-.':
 834             try: return int(v)
 835             except ValueError:
 836                 try: return float(v)
 837                 except ValueError: pass
 838         return v.strip()
 839     def values(self):
 840         result = []
 841         for key in self.keys():
 842             try:
 843                 result.append(self[key])
 844             except IndexError:
 845                 result.append(self.dict[key])
 846         return result
 847     def items(self):
 848         result = []
 849         for key in self.keys():
 850             try:
 851                 result.append((key, self[key]))
 852             except IndexError:
 853                 result.append((key, self.dict[key]))
 854         return result
 855
 856
 857 class FormContent(FormContentDict):
 858     """This class is present for backwards compatibility only."""
 859     def values(self, key):
 860         if key in self.dict :return self.dict[key]
 861         else: return None
 862     def indexed_value(self, key, location):
 863         if key in self.dict:
 864             if len(self.dict[key]) > location:
 865                 return self.dict[key][location]
 866             else: return None
 867         else: return None
 868     def value(self, key):
 869         if key in self.dict: return self.dict[key][0]
 870         else: return None
 871     def length(self, key):
 872         return len(self.dict[key])
 873     def stripped(self, key):
 874         if key in self.dict: return self.dict[key][0].strip()
 875         else: return None
 876     def pars(self):
 877         return self.dict
 878
 879
 880 # Test/debug code
 881 # ===============
 882
 883 def test(environ=os.environ):
 884     """Robust test CGI script, usable as main program.
 885
 886     Write minimal HTTP headers and dump all information provided to
 887     the script in HTML form.
 888
 889     """
 890     print "Content-type: text/html"
 891     print
 892     sys.stderr = sys.stdout
 893     try:
 894         form = FieldStorage()   # Replace with other classes to test those
 895         print_directory()
 896         print_arguments()
 897         print_form(form)
 898         print_environ(environ)
 899         print_environ_usage()
 900         def f():
 901             exec "testing print_exception() -- <I>italics?</I>"
 902         def g(f=f):
 903             f()
 904         print "<H3>What follows is a test, not an actual exception:</H3>"
 905         g()
 906     except:
 907         print_exception()
 908
 909     print "<H1>Second try with a small maxlen...</H1>"
 910
 911     global maxlen
 912     maxlen = 50
 913     try:
 914         form = FieldStorage()   # Replace with other classes to test those
 915         print_directory()
 916         print_arguments()
 917         print_form(form)
 918         print_environ(environ)
 919     except:
 920         print_exception()
 921
 922 def print_exception(type=None, value=None, tb=None, limit=None):
 923     if type is None:
 924         type, value, tb = sys.exc_info()
 925     import traceback
 926     print
 927     print "<H3>Traceback (most recent call last):</H3>"
 928     list = traceback.format_tb(tb, limit) + \
 929            traceback.format_exception_only(type, value)
 930     print "<PRE>%s<B>%s</B></PRE>" % (
 931         escape("".join(list[:-1])),
 932         escape(list[-1]),
 933         )
 934     del tb
 935
 936 def print_environ(environ=os.environ):
 937     """Dump the shell environment as HTML."""
 938     keys = environ.keys()
 939     keys.sort()
 940     print
 941     print "<H3>Shell Environment:</H3>"
 942     print "<DL>"
 943     for key in keys:
 944         print "<DT>", escape(key), "<DD>", escape(environ[key])
 945     print "</DL>"
 946     print
 947
 948 def print_form(form):
 949     """Dump the contents of a form as HTML."""
 950     keys = form.keys()
 951     keys.sort()
 952     print
 953     print "<H3>Form Contents:</H3>"
 954     if not keys:
 955         print "<P>No form fields."
 956     print "<DL>"
 957     for key in keys:
 958         print "<DT>" + escape(key) + ":",
 959         value = form[key]
 960         print "<i>" + escape(repr(type(value))) + "</i>"
 961         print "<DD>" + escape(repr(value))
 962     print "</DL>"
 963     print
 964
 965 def print_directory():
 966     """Dump the current directory as HTML."""
 967     print
 968     print "<H3>Current Working Directory:</H3>"
 969     try:
 970         pwd = os.getcwd()
 971     except os.error, msg:
 972         print "os.error:", escape(str(msg))
 973     else:
 974         print escape(pwd)
 975     print
 976
 977 def print_arguments():
 978     print
 979     print "<H3>Command Line Arguments:</H3>"
 980     print
 981     print sys.argv
 982     print
 983
 984 def print_environ_usage():
 985     """Dump a list of environment variables used by CGI as HTML."""
 986     print """
 987 <H3>These environment variables could have been set:</H3>
 988 <UL>
 989 <LI>AUTH_TYPE
 990 <LI>CONTENT_LENGTH
 991 <LI>CONTENT_TYPE
 992 <LI>DATE_GMT
 993 <LI>DATE_LOCAL
 994 <LI>DOCUMENT_NAME
 995 <LI>DOCUMENT_ROOT
 996 <LI>DOCUMENT_URI
 997 <LI>GATEWAY_INTERFACE
 998 <LI>LAST_MODIFIED
 999 <LI>PATH
1000 <LI>PATH_INFO
1001 <LI>PATH_TRANSLATED
1002 <LI>QUERY_STRING
1003 <LI>REMOTE_ADDR
1004 <LI>REMOTE_HOST
1005 <LI>REMOTE_IDENT
1006 <LI>REMOTE_USER
1007 <LI>REQUEST_METHOD
1008 <LI>SCRIPT_NAME
1009 <LI>SERVER_NAME
1010 <LI>SERVER_PORT
1011 <LI>SERVER_PROTOCOL
1012 <LI>SERVER_ROOT
1013 <LI>SERVER_SOFTWARE
1014 </UL>
1015 In addition, HTTP headers sent by the server may be passed in the
1016 environment as well.  Here are some common variable names:
1017 <UL>
1018 <LI>HTTP_ACCEPT
1019 <LI>HTTP_CONNECTION
1020 <LI>HTTP_HOST
1021 <LI>HTTP_PRAGMA
1022 <LI>HTTP_REFERER
1023 <LI>HTTP_USER_AGENT
1024 </UL>
1025 """
1026
1027
1028 # Utilities
1029 # =========
1030
1031 def escape(s, quote=None):
1032     '''Replace special characters "&", "<" and ">" to HTML-safe sequences.
1033     If the optional flag quote is true, the quotation mark character (")
1034     is also translated.'''
1035     s = s.replace("&", "&amp;") # Must be done first!
1036     s = s.replace("<", "&lt;")
1037     s = s.replace(">", "&gt;")
1038     if quote:
1039         s = s.replace('"', "&quot;")
1040     return s
1041
1042 def valid_boundary(s, _vb_pattern="^[ -~]{0,200}[!-~]$"):
1043     import re
1044     return re.match(_vb_pattern, s)
1045
1046 # Invoke mainline
1047 # ===============
1048
1049 # Call test() when this file is run as a script (not imported as a module)
1050 if __name__ == '__main__':
1051     test()