Ticket #2070: 5127_file_uploads_no_streaming_fixed.diff

File 5127_file_uploads_no_streaming_fixed.diff, 39.2 KB (added by SmileyChris, 8 years ago)

There was an error uploading large files with streaming turned off

  • django/conf/global_settings.py

     
    240240# isExistingURL validator.
    241241URL_VALIDATOR_USER_AGENT = "Django/0.96pre (http://www.djangoproject.com)"
    242242
     243# The directory to place streamed file uploads. The web server needs write
     244# permissions on this directory.
     245# If this is None, streaming uploads are disabled.
     246FILE_UPLOAD_DIR = None
     247
     248# The minimum size of a POST before file uploads are streamed to disk.
     249# Any less than this number, and the file is uploaded to memory.
     250# Size is in bytes.
     251STREAMING_MIN_POST_SIZE = 512 * (2**10)
     252
    243253##############
    244254# MIDDLEWARE #
    245255##############
  • django/core/handlers/base.py

     
    55
    66class BaseHandler(object):
    77    def __init__(self):
    8         self._request_middleware = self._view_middleware = self._response_middleware = self._exception_middleware = None
     8        self._upload_middleware = self._request_middleware = self._view_middleware = self._response_middleware = self._exception_middleware = None
    99
    1010    def load_middleware(self):
    1111        """
     
    1919        self._view_middleware = []
    2020        self._response_middleware = []
    2121        self._exception_middleware = []
     22        self._upload_middleware = []
    2223        for middleware_path in settings.MIDDLEWARE_CLASSES:
    2324            try:
    2425                dot = middleware_path.rindex('.')
     
    4748                self._response_middleware.insert(0, mw_instance.process_response)
    4849            if hasattr(mw_instance, 'process_exception'):
    4950                self._exception_middleware.insert(0, mw_instance.process_exception)
     51            if hasattr(mw_instance, 'process_upload'):
     52                self._upload_middleware.append(mw_instance.process_upload)
    5053
     54    def file_progress_descriptor(self, request):
     55        """
     56        Returns a descriptor that manages the file_progress
     57        """
     58        for mw_call in self._upload_middleware:
     59            result = mw_call(http.MultiPartParserError)
     60            if result != None:
     61                return result
     62        return http.DefaultFileProgressDescriptor(http.MultiPartParserError)
     63
    5164    def get_response(self, request):
    5265        "Returns an HttpResponse object for the given HttpRequest"
    5366        from django.core import exceptions, urlresolvers
    5467        from django.core.mail import mail_admins
    5568        from django.conf import settings
    5669
     70        # Add file_progress descriptor
     71        request._file_progress = self.file_progress_descriptor(request)
     72
    5773        # Apply request middleware
    5874        for middleware_method in self._request_middleware:
    5975            response = middleware_method(request)
  • django/core/handlers/modpython.py

     
    4747    def _load_post_and_files(self):
    4848        "Populates self._post and self._files"
    4949        if 'content-type' in self._req.headers_in and self._req.headers_in['content-type'].startswith('multipart'):
    50             self._post, self._files = http.parse_file_upload(self._req.headers_in, self.raw_post_data)
     50            self._raw_post_data = None # raw data is not available for streamed multipart messages
     51            try:
     52                self._post, self._files = http.parse_file_upload(self._req.headers_in, self._req, self)
     53            except:
     54                self._post, self._files = {}, {} # make sure we dont read the input stream again
     55                raise
    5156        else:
    5257            self._post, self._files = http.QueryDict(self.raw_post_data), datastructures.MultiValueDict()
    5358
     
    9297                'AUTH_TYPE':         self._req.ap_auth_type,
    9398                'CONTENT_LENGTH':    self._req.clength, # This may be wrong
    9499                'CONTENT_TYPE':      self._req.content_type, # This may be wrong
    95                 'GATEWAY_INTERFACE': 'CGI/1.1',
    96                 'PATH_INFO':         self._req.path_info,
    97                 'PATH_TRANSLATED':   None, # Not supported
    98                 'QUERY_STRING':      self._req.args,
    99                 'REMOTE_ADDR':       self._req.connection.remote_ip,
    100                 'REMOTE_HOST':       None, # DNS lookups not supported
    101                 'REMOTE_IDENT':      self._req.connection.remote_logname,
    102                 'REMOTE_USER':       self._req.user,
    103                 'REQUEST_METHOD':    self._req.method,
    104                 'SCRIPT_NAME':       None, # Not supported
    105                 'SERVER_NAME':       self._req.server.server_hostname,
    106                 'SERVER_PORT':       self._req.server.port,
    107                 'SERVER_PROTOCOL':   self._req.protocol,
    108                 'SERVER_SOFTWARE':   'mod_python'
     100                'GATEWAY_INTERFACE':  'CGI/1.1',
     101                'PATH_INFO':          self._req.path_info,
     102                'PATH_TRANSLATED':    None, # Not supported
     103                'QUERY_STRING':       self._req.args,
     104                'REMOTE_ADDR':        self._req.connection.remote_ip,
     105                'REMOTE_HOST':        None, # DNS lookups not supported
     106                'REMOTE_IDENT':       self._req.connection.remote_logname,
     107                'REMOTE_USER':        self._req.user,
     108                'REQUEST_METHOD':     self._req.method,
     109                'SCRIPT_NAME':        None, # Not supported
     110                'SERVER_NAME':        self._req.server.server_hostname,
     111                'SERVER_PORT':        self._req.server.port,
     112                'SERVER_PROTOCOL':    self._req.protocol,
     113                'UPLOAD_PROGRESS_ID': self._get_file_progress_id(),
     114                'SERVER_SOFTWARE':    'mod_python'
    109115            }
    110116            for key, value in self._req.headers_in.items():
    111117                key = 'HTTP_' + key.upper().replace('-', '_')
     
    122128    def _get_method(self):
    123129        return self.META['REQUEST_METHOD'].upper()
    124130
     131    def _get_file_progress_id(self):
     132        """
     133        Returns the Progress ID of the request,
     134        usually provided if there is a file upload
     135        going on.
     136        Returns ``None`` if no progress ID is specified.
     137        """
     138        return self._get_file_progress_from_args(self._req.headers_in,
     139                                                 self.GET,
     140                                                 self._req.args)
     141
    125142    GET = property(_get_get, _set_get)
    126143    POST = property(_get_post, _set_post)
    127144    COOKIES = property(_get_cookies, _set_cookies)
  • django/core/handlers/wsgi.py

     
    7575        self.environ = environ
    7676        self.path = environ['PATH_INFO']
    7777        self.META = environ
     78        self.META['UPLOAD_PROGRESS_ID'] = self._get_file_progress_id()
    7879        self.method = environ['REQUEST_METHOD'].upper()
    7980
    8081    def __repr__(self):
     
    111112            if self.environ.get('CONTENT_TYPE', '').startswith('multipart'):
    112113                header_dict = dict([(k, v) for k, v in self.environ.items() if k.startswith('HTTP_')])
    113114                header_dict['Content-Type'] = self.environ.get('CONTENT_TYPE', '')
    114                 self._post, self._files = http.parse_file_upload(header_dict, self.raw_post_data)
     115                header_dict['Content-Length'] = self.environ.get('CONTENT_LENGTH', '')
     116                header_dict['X-Progress-ID'] = self.environ.get('HTTP_X_PROGRESS_ID', '')
     117                try:
     118                    self._post, self._files = http.parse_file_upload(header_dict, self.environ['wsgi.input'], self)
     119                except:
     120                    self._post, self._files = {}, {} # make sure we dont read the input stream again
     121                    raise
     122                self._raw_post_data = None # raw data is not available for streamed multipart messages
    115123            else:
    116124                self._post, self._files = http.QueryDict(self.raw_post_data), datastructures.MultiValueDict()
    117125        else:
     
    167175            buf.close()
    168176            return self._raw_post_data
    169177
     178    def _get_file_progress_id(self):
     179        """
     180        Returns the Progress ID of the request,
     181        usually provided if there is a file upload
     182        going on.
     183        Returns ``None`` if no progress ID is specified.
     184        """
     185        return self._get_file_progress_from_args(self.environ,
     186                                                 self.GET,
     187                                                 self.environ.get('QUERY_STRING', ''))
     188
    170189    GET = property(_get_get, _set_get)
    171190    POST = property(_get_post, _set_post)
    172191    COOKIES = property(_get_cookies, _set_cookies)
  • django/db/models/base.py

     
    1212from django.dispatch import dispatcher
    1313from django.utils.datastructures import SortedDict
    1414from django.utils.functional import curry
     15from django.utils.file import file_move_safe
    1516from django.conf import settings
    1617from itertools import izip
    1718import types
    1819import sys
    1920import os
    2021
     22               
    2123class ModelBase(type):
    2224    "Metaclass for all models"
    2325    def __new__(cls, name, bases, attrs):
     
    361363    def _get_FIELD_size(self, field):
    362364        return os.path.getsize(self._get_FIELD_filename(field))
    363365
    364     def _save_FIELD_file(self, field, filename, raw_contents, save=True):
     366    def _save_FIELD_file(self, field, filename, raw_field, save=True):
    365367        directory = field.get_directory_name()
    366368        try: # Create the date-based directory if it doesn't exist.
    367369            os.makedirs(os.path.join(settings.MEDIA_ROOT, directory))
     
    383385        setattr(self, field.attname, filename)
    384386
    385387        full_filename = self._get_FIELD_filename(field)
    386         fp = open(full_filename, 'wb')
    387         fp.write(raw_contents)
    388         fp.close()
     388        if raw_field.has_key('tmpfilename'):
     389            raw_field['tmpfile'].close()
     390            file_move_safe(raw_field['tmpfilename'], full_filename)
     391        else:
     392            fp = open(full_filename, 'wb')
     393            fp.write(raw_field['content'])
     394            fp.close()
    389395
    390396        # Save the width and/or height, if applicable.
    391397        if isinstance(field, ImageField) and (field.width_field or field.height_field):
  • django/db/models/fields/__init__.py

     
    638638        setattr(cls, 'get_%s_filename' % self.name, curry(cls._get_FIELD_filename, field=self))
    639639        setattr(cls, 'get_%s_url' % self.name, curry(cls._get_FIELD_url, field=self))
    640640        setattr(cls, 'get_%s_size' % self.name, curry(cls._get_FIELD_size, field=self))
    641         setattr(cls, 'save_%s_file' % self.name, lambda instance, filename, raw_contents, save=True: instance._save_FIELD_file(self, filename, raw_contents, save))
     641        setattr(cls, 'save_%s_file' % self.name, lambda instance, filename, raw_field, save=True: instance._save_FIELD_file(self, filename, raw_field, save))
    642642        dispatcher.connect(self.delete_file, signal=signals.post_delete, sender=cls)
    643643
    644644    def delete_file(self, instance):
     
    661661        if new_data.get(upload_field_name, False):
    662662            func = getattr(new_object, 'save_%s_file' % self.name)
    663663            if rel:
    664                 func(new_data[upload_field_name][0]["filename"], new_data[upload_field_name][0]["content"], save)
     664                func(new_data[upload_field_name][0]["filename"], new_data[upload_field_name][0], save)
    665665            else:
    666                 func(new_data[upload_field_name]["filename"], new_data[upload_field_name]["content"], save)
     666                func(new_data[upload_field_name]["filename"], new_data[upload_field_name], save)
    667667
    668668    def get_directory_name(self):
    669669        return os.path.normpath(datetime.datetime.now().strftime(self.upload_to))
  • django/http/__init__.py

     
    1 import os
     1import os, pickle
    22from Cookie import SimpleCookie
    33from pprint import pformat
    44from urllib import urlencode, quote
    55from django.utils.datastructures import MultiValueDict
     6from django.http.file_descriptor import DefaultFileProgressDescriptor
     7import re
    68
     9try:
     10    from cStringIO import StringIO
     11except ImportError:
     12    from StringIO import StringIO
     13
    714RESERVED_CHARS="!*'();:@&=+$,/?%#[]"
    815
    916try:
     
    1219except ImportError:
    1320    from cgi import parse_qsl
    1421
     22class MetaFileProgressDescriptor(object):
     23    """
     24    This descriptor allows other descriptors to
     25    be loaded in runtime to a request instance.
     26    """
     27    def __get__(self, request, *args, **kwargs):
     28        return request._file_progress.__get__(request, *args, **kwargs)
     29
     30    def __set__(self, request, *args, **kwargs):
     31        return request._file_progress.__set__(request, *args, **kwargs)
     32
     33    def __delete__(self, request, *args, **kwargs):
     34        return request._file_progress.__delete__(request, *args, **kwargs)
     35
    1536class Http404(Exception):
    1637    pass
    1738
    1839class HttpRequest(object):
    1940    "A basic HTTP request"
     41    upload_id_re = re.compile(r'^[a-fA-F0-9]{32}$')
     42    file_progress = MetaFileProgressDescriptor()
     43
    2044    def __init__(self):
    2145        self.GET, self.POST, self.COOKIES, self.META, self.FILES = {}, {}, {}, {}, {}
    2246        self.path = ''
     
    4266    def is_secure(self):
    4367        return os.environ.get("HTTPS") == "on"
    4468
    45 def parse_file_upload(header_dict, post_data):
    46     "Returns a tuple of (POST MultiValueDict, FILES MultiValueDict)"
    47     import email, email.Message
    48     from cgi import parse_header
    49     raw_message = '\r\n'.join(['%s:%s' % pair for pair in header_dict.items()])
    50     raw_message += '\r\n\r\n' + post_data
    51     msg = email.message_from_string(raw_message)
    52     POST = MultiValueDict()
    53     FILES = MultiValueDict()
    54     for submessage in msg.get_payload():
    55         if submessage and isinstance(submessage, email.Message.Message):
    56             name_dict = parse_header(submessage['Content-Disposition'])[1]
    57             # name_dict is something like {'name': 'file', 'filename': 'test.txt'} for file uploads
    58             # or {'name': 'blah'} for POST fields
    59             # We assume all uploaded files have a 'filename' set.
    60             if 'filename' in name_dict:
    61                 assert type([]) != type(submessage.get_payload()), "Nested MIME messages are not supported"
    62                 if not name_dict['filename'].strip():
    63                     continue
    64                 # IE submits the full path, so trim everything but the basename.
    65                 # (We can't use os.path.basename because it expects Linux paths.)
    66                 filename = name_dict['filename'][name_dict['filename'].rfind("\\")+1:]
    67                 FILES.appendlist(name_dict['name'], {
    68                     'filename': filename,
    69                     'content-type': 'Content-Type' in submessage and submessage['Content-Type'] or None,
    70                     'content': submessage.get_payload(),
    71                 })
    72             else:
    73                 POST.appendlist(name_dict['name'], submessage.get_payload())
    74     return POST, FILES
     69    def _get_file_progress_from_args(self, headers, get, querystring):
     70        if 'X-Upload-ID' in headers:
     71            progress_id = headers['X-Upload-ID']
     72        elif 'X-Progress-ID' in headers:
     73            progress_id = headers['X-Progress-ID']
     74        elif 'HTTP_X_UPLOAD_ID' in headers:
     75            progress_id = headers['HTTP_X_UPLOAD_ID']
     76        elif 'HTTP_X_PROGRESS_ID' in headers:
     77            progress_id = headers['HTTP_X_PROGRESS_ID']
     78        elif 'upload_id' in get:
     79            progress_id = get['upload_id']
     80        elif 'progress_id' in get:
     81            progress_id = get['progress_id']
     82        elif querystring and len(querystring.strip()) == 32:
     83            progress_id = querystring
     84        else:
     85            return None
    7586
     87        if not self.upload_id_re.match(progress_id):
     88            return None
     89
     90        return progress_id
     91
     92def parse_file_upload(headers, input, request):
     93    from django.conf import settings
     94
     95    # Only stream files to disk if FILE_STREAMING_DIR is set
     96    file_upload_dir = settings.FILE_UPLOAD_DIR
     97    streaming_min_post_size = settings.STREAMING_MIN_POST_SIZE
     98
     99    try:
     100        parser = MultiPartParser(headers, input, request, file_upload_dir, streaming_min_post_size)
     101        return parser.parse()
     102    except MultiPartParserError, e:
     103        return MultiValueDict({ '_file_upload_error': [e.message] }), {}
     104
     105class MultiPartParserError(Exception):
     106    def __init__(self, message):
     107        self.message = message
     108    def __str__(self):
     109        return repr(self.message)
     110
     111class MultiPartParser(object):
     112    """
     113    A rfc2388 multipart/form-data parser.
     114   
     115    parse() reads the input stream in chunk_size chunks and returns a
     116    tuple of (POST MultiValueDict, FILES MultiValueDict). If
     117    file_upload_dir is defined files will be streamed to temporary
     118    files in the specified directory.
     119
     120    The FILES dictionary will have 'filename', 'content-type',
     121    'content' and 'content-length' entries. For streamed files it will
     122    also have 'tmpfilename' and 'tmpfile'. The 'content' entry will
     123    only be read from disk when referenced for streamed files.
     124
     125    If the header X-Progress-ID is sent with a 32 character hex string
     126    a temporary file with the same name will be created in
     127    `file_upload_dir`` with a pickled { 'received', 'size' }
     128    dictionary with the number of bytes received and the size expected
     129    respectively. The file will be unlinked when the parser finishes.
     130    """
     131    def __init__(self, headers, input, request, file_upload_dir=None, streaming_min_post_size=None, chunk_size=1024*64):
     132        try:
     133            content_length = int(headers['Content-Length'])
     134        except:
     135            raise MultiPartParserError('Invalid Content-Length: %s' % headers.get('Content-Length'))
     136
     137        content_type = headers.get('Content-Type')
     138
     139        if not content_type or not content_type.startswith('multipart/'):
     140            raise MultiPartParserError('Invalid Content-Type: %s' % content_type)
     141           
     142        ctype, opts = self.parse_header(content_type)
     143        boundary = opts.get('boundary')
     144        from cgi import valid_boundary
     145        if not boundary or not valid_boundary(boundary):
     146            raise MultiPartParserError('Invalid boundary in multipart form: %s' % boundary)
     147
     148        progress_id = request.META['UPLOAD_PROGRESS_ID']
     149
     150        if file_upload_dir and progress_id:
     151            self._progress_filename = os.path.join(file_upload_dir, progress_id)
     152        else:
     153            self._progress_filename = None
     154        self._boundary = '--' + boundary
     155        self._input = input
     156        self._size = content_length
     157        self._received = 0
     158        self._file_upload_dir = file_upload_dir
     159        self._chunk_size = chunk_size
     160        self._state = 'PREAMBLE'
     161        self._partial = ''
     162        self._post = MultiValueDict()
     163        self._files = MultiValueDict()
     164        self._request = request
     165
     166        if streaming_min_post_size is not None and content_length < streaming_min_post_size:
     167            self._file_upload_dir = None # disable file streaming for small request
     168        elif self._progress_filename:
     169            request.file_progress = {'state': 'starting'}
     170
     171        try:
     172            # Use mx fast string search if available.
     173            from mx.TextTools import FS
     174            self._fs = FS(self._boundary)
     175        except ImportError:
     176            self._fs = None
     177
     178    def parse(self):
     179        try:
     180            self._parse()
     181        finally:
     182            if self._progress_filename:
     183                self._request.file_progress = {'state': 'done'}
     184        return self._post, self._files
     185
     186    def _parse(self):
     187        size = self._size
     188
     189        try:
     190            while size > 0:
     191                n = self._read(self._input, min(self._chunk_size, size))
     192                if not n:
     193                    break
     194                size -= n
     195        except:
     196            # consume any remaining data so we dont generate a "Connection Reset" error
     197            size = self._size - self._received
     198            while size > 0:
     199                data = self._input.read(min(self._chunk_size, size))
     200                size -= len(data)
     201            raise
     202
     203    def _find_boundary(self, data, start, stop):
     204        """
     205        Find the next boundary and return the end of current part
     206        and start of next part.
     207        """
     208        if self._fs:
     209            boundary = self._fs.find(data, start, stop)
     210        else:
     211            boundary = data.find(self._boundary, start, stop)
     212        if boundary >= 0:
     213            end = boundary
     214            next = boundary + len(self._boundary)
     215
     216            # backup over CRLF
     217            if end > 0 and data[end-1] == '\n': end -= 1
     218            if end > 0 and data[end-1] == '\r': end -= 1
     219            # skip over --CRLF
     220            if next < stop and data[next] == '-': next += 1
     221            if next < stop and data[next] == '-': next += 1
     222            if next < stop and data[next] == '\r': next += 1
     223            if next < stop and data[next] == '\n': next += 1
     224
     225            return True, end, next
     226        else:
     227            return False, stop, stop
     228
     229    class TemporaryFile(object):
     230        "A temporary file that tries to delete itself when garbage collected."
     231        def __init__(self, dir):
     232            import tempfile
     233            (fd, name) = tempfile.mkstemp(suffix='.upload', dir=dir)
     234            self.file = os.fdopen(fd, 'w+b')
     235            self.name = name
     236
     237        def __getattr__(self, name):
     238            a = getattr(self.__dict__['file'], name)
     239            if type(a) != type(0):
     240                setattr(self, name, a)
     241            return a
     242
     243        def __del__(self):
     244            try:
     245                os.unlink(self.name)
     246            except OSError:
     247                pass
     248
     249    class LazyContent(dict):
     250        """
     251        A lazy FILES dictionary entry that reads the contents from
     252        tmpfile only when referenced.
     253        """
     254        def __init__(self, data):
     255            dict.__init__(self, data)
     256
     257        def __getitem__(self, key):
     258            if key == 'content' and not self.has_key(key):
     259                self['tmpfile'].seek(0)
     260                self['content'] = self['tmpfile'].read()
     261            return dict.__getitem__(self, key)
     262
     263    def _read(self, input, size):
     264        data = input.read(size)
     265
     266        if not data:
     267            return 0
     268
     269        read_size = len(data)
     270        self._received += read_size
     271
     272        if self._partial:
     273            data = self._partial + data
     274
     275        start = 0
     276        stop = len(data)
     277
     278        while start < stop:
     279            boundary, end, next = self._find_boundary(data, start, stop)
     280
     281            if not boundary and read_size:
     282                # make sure we dont treat a partial boundary (and its separators) as data
     283                stop -= len(self._boundary) + 16
     284                end = next = stop
     285                if end <= start:
     286                    break # need more data
     287
     288            if self._state == 'PREAMBLE':
     289                # Preamble, just ignore it
     290                self._state = 'HEADER'
     291
     292            elif self._state == 'HEADER':
     293                # Beginning of header, look for end of header and parse it if found.
     294
     295                header_end = data.find('\r\n\r\n', start, stop)
     296                if header_end == -1:
     297                    break # need more data
     298
     299                header = data[start:header_end]
     300
     301                self._fieldname = None
     302                self._filename = None
     303                self._content_type = None
     304
     305                for line in header.split('\r\n'):
     306                    ctype, opts = self.parse_header(line)
     307                    if ctype == 'content-disposition: form-data':
     308                        self._fieldname = opts.get('name')
     309                        self._filename = opts.get('filename')
     310                    elif ctype.startswith('content-type: '):
     311                        self._content_type = ctype[14:]
     312
     313                if self._filename is not None:
     314                    # cleanup filename from IE full paths:
     315                    self._filename = self._filename[self._filename.rfind("\\")+1:].strip()
     316
     317                    if self._filename: # ignore files without filenames
     318                        if self._file_upload_dir:
     319                            try:
     320                                self._file = self.TemporaryFile(dir=self._file_upload_dir)
     321                            except:
     322                                raise MultiPartParserError("Failed to create temporary file.")
     323                        else:
     324                            self._file = StringIO()
     325                    else:
     326                        self._file = None
     327                    self._filesize = 0
     328                    self._state = 'FILE'
     329                else:
     330                    self._field = StringIO()
     331                    self._state = 'FIELD'
     332                next = header_end + 4
     333
     334            elif self._state == 'FIELD':
     335                # In a field, collect data until a boundary is found.
     336
     337                self._field.write(data[start:end])
     338                if boundary:
     339                    if self._fieldname:
     340                        self._post.appendlist(self._fieldname, self._field.getvalue())
     341                    self._field.close()
     342                    self._state = 'HEADER'
     343
     344            elif self._state == 'FILE':
     345                # In a file, collect data until a boundary is found.
     346
     347                if self._file:
     348                    try:
     349                        self._file.write(data[start:end])
     350                    except IOError, e:
     351                        raise MultiPartParserError("Failed to write to temporary file.")
     352                    self._filesize += end-start
     353
     354                    if self._progress_filename:
     355                        self._request.file_progress = {'received': self._received,
     356                                                       'size':     self._size,
     357                                                       'state':    'uploading'}
     358
     359                if boundary:
     360                    if self._file:
     361                        if self._file_upload_dir:
     362                            self._file.seek(0)
     363                            file = self.LazyContent({
     364                                'filename': self._filename,
     365                                'content-type':  self._content_type,
     366                                # 'content': is read on demand
     367                                'content-length': self._filesize,
     368                                'tmpfilename': self._file.name,
     369                                'tmpfile': self._file
     370                            })
     371                        else:
     372                            file = {
     373                                'filename': self._filename,
     374                                'content-type':  self._content_type,
     375                                'content': self._file.getvalue(),
     376                                'content-length': self._filesize
     377                            }
     378                            self._file.close()
     379
     380                        self._files.appendlist(self._fieldname, file)
     381
     382                    self._state = 'HEADER'
     383
     384            start = next
     385
     386        self._partial = data[start:]
     387
     388        return read_size
     389
     390    def parse_header(self, line):
     391        from cgi import parse_header
     392        return parse_header(line)
     393
    76394class QueryDict(MultiValueDict):
    77395    """A specialized MultiValueDict that takes a query string when initialized.
    78396    This is immutable unless you create a copy of it."""
  • django/http/file_descriptor.py

     
     1"""
     2This file contains a fallback FileProgressDescriptor
     3for file upload progress.
     4"""
     5import pickle
     6import os
     7
     8class DefaultFileProgressDescriptor(object):
     9    def __init__(self, FileException):
     10        from django.conf import settings
     11
     12        self.FileException = FileException
     13        self.file_upload_dir = settings.FILE_UPLOAD_DIR
     14
     15    def __get__(self, request, HttpRequest):
     16        """
     17        Returns the file progress for this request or an empty dictionary if
     18        the file progress is not known. The result is cached.
     19        """
     20        progress_id = request.META['UPLOAD_PROGRESS_ID']
     21
     22        if not progress_id or not self.file_upload_dir:
     23            return {}
     24
     25        if getattr(self, '_file_progress', None) is not None:
     26            return self._file_progress
     27
     28        try:
     29            f = open(os.path.join(self.file_upload_dir, progress_id), 'rb')
     30            progress = pickle.load(f)
     31            f.close()
     32        except:
     33            progress = {}
     34
     35        self._file_progress = progress
     36        return progress
     37
     38    def __set__(self, request, new_progress):
     39        """
     40        Sets the value of the file progress for this request.
     41        If no file progress is underway, raises an error.
     42        """
     43        progress_id = request.META['UPLOAD_PROGRESS_ID']
     44
     45        if not progress_id or not self.file_upload_dir:
     46            raise self.FileException('There is no upload in progress.')
     47
     48        self._file_progress = new_progress
     49        f = open(os.path.join(self.file_upload_dir, progress_id), 'wb')
     50        pickle.dump(new_progress, f)
     51        f.close()
     52
     53    def __delete__(self, request):
     54        """
     55        Removes the file if there is an upload in process.
     56        """
     57        progress_id = request.META.get('UPLOAD_PROGRESS_ID')
     58
     59        if not progress_id or not self.file_upload_dir:
     60            raise self.FileException('There is no upload in progress.')
     61
     62        try:
     63            os.remove(os.path.join(self.file_upload_dir, progress_id))
     64        except:
     65            pass
  • django/oldforms/__init__.py

     
    666666        self.validator_list = [self.isNonEmptyFile] + validator_list
    667667
    668668    def isNonEmptyFile(self, field_data, all_data):
    669         try:
    670             content = field_data['content']
    671         except TypeError:
     669        if field_data.has_key('_file_upload_error'):
     670            raise validators.CriticalValidationError, field_data['_file_upload_error']
     671        if not field_data.has_key('filename'):
    672672            raise validators.CriticalValidationError, gettext("No file was submitted. Check the encoding type on the form.")
    673         if not content:
     673        if not field_data['content-length']:
    674674            raise validators.CriticalValidationError, gettext("The submitted file is empty.")
    675675
    676676    def render(self, data):
    677677        return '<input type="file" id="%s" class="v%s" name="%s" />' % \
    678678            (self.get_id(), self.__class__.__name__, self.field_name)
    679679
     680    def prepare(self, new_data):
     681        if new_data.has_key('_file_upload_error'):
     682            # pretend we got something in the field to raise a validation error later
     683            new_data[self.field_name] = { '_file_upload_error': new_data['_file_upload_error'] }
     684
    680685    def html2python(data):
    681686        if data is None:
    682687            raise EmptyValue
  • django/utils/file.py

     
     1import os
     2
     3try:
     4    import shutils
     5    file_move = shutils.move
     6except:
     7    file_move = os.rename
     8
     9def file_move_safe(old_file_name, new_file_name, chunk_size = 1024*64):
     10    """
     11    Moves a file from one location to another in the safest way possible.
     12   
     13    First, it tries using shutils.move, which is OS-dependent but doesn't
     14    break with change of filesystems. Then it tries os.rename, which will
     15    break if it encounters a change in filesystems. Lastly, it streams
     16    it manually from one file to another in python.
     17    """
     18   
     19    try:
     20        file_move(old_file_name, new_file_name)
     21        return
     22    except:
     23        pass
     24   
     25    new_file = open(new_file_name, 'wb')
     26    old_file = open(old_file_name, 'rb')
     27    current_chunk = None
     28   
     29    while current_chunk != '':
     30        current_chunk = old_file.read(chunk_size)
     31        new_file.write(current_chunk)
     32       
     33    new_file.close()
     34    old_file.close()
     35
     36    os.remove(old_file_name)
  • docs/forms.txt

     
    475475   new_data = request.POST.copy()
    476476   new_data.update(request.FILES)
    477477
     478Streaming file uploads.
     479-----------------------
     480
     481File uploads will be read into memory by default. This works fine for
     482small to medium sized uploads (from 1MB to 100MB depending on your
     483setup and usage). If you want to support larger uploads you can enable
     484upload streaming where only a small part of the file will be in memory
     485at any time. To do this you need to specify the ``FILE_UPLOAD_DIR``
     486setting (see the settings_ document for more details).
     487
     488See `request object`_ for more details about ``request.FILES`` objects
     489with streaming file uploads enabled.
     490
    478491Validators
    479492==========
    480493
     
    697710.. _`generic views`: ../generic_views/
    698711.. _`models API`: ../model-api/
    699712.. _settings: ../settings/
     713.. _request object: ../request_response/#httprequest-objects
  • docs/request_response.txt

     
    7272``FILES``
    7373    A dictionary-like object containing all uploaded files. Each key in
    7474    ``FILES`` is the ``name`` from the ``<input type="file" name="" />``. Each
    75     value in ``FILES`` is a standard Python dictionary with the following three
     75    value in ``FILES`` is a standard Python dictionary with the following four
    7676    keys:
    7777
    7878        * ``filename`` -- The name of the uploaded file, as a Python string.
    7979        * ``content-type`` -- The content type of the uploaded file.
    8080        * ``content`` -- The raw content of the uploaded file.
     81        * ``content-length`` -- The length of the content in bytes.
    8182
     83    If streaming file uploads are enabled two additional keys
     84    describing the uploaded file will be present:
     85
     86        * ``tmpfilename`` -- The filename for the temporary file.
     87        * ``tmpfile`` -- An open file object for the temporary file.
     88
     89    The temporary file will be removed when the request finishes.
     90
     91    Note that accessing ``content`` when streaming uploads are enabled
     92    will read the whole file into memory which may not be what you want.
     93
    8294    Note that ``FILES`` will only contain data if the request method was POST
    8395    and the ``<form>`` that posted to the request had
    8496    ``enctype="multipart/form-data"``. Otherwise, ``FILES`` will be a blank
  • docs/settings.txt

     
    439439
    440440.. _Testing Django Applications: ../testing/
    441441
     442FILE_UPLOAD_DIR
     443---------------
     444
     445Default: ``None``
     446
     447Path to a directory where temporary files should be written during
     448file uploads. Leaving this as ``None`` will disable streaming file uploads,
     449and cause all uploaded files to be stored (temporarily) in memory.
     450
    442451IGNORABLE_404_ENDS
    443452------------------
    444453
     
    782791
    783792.. _site framework docs: ../sites/
    784793
     794STREAMING_MIN_POST_SIZE
     795-----------------------
     796
     797Default: 524288 (``512*1024``)
     798
     799An integer specifying the minimum number of bytes that has to be
     800received (in a POST) for file upload streaming to take place. Any
     801request smaller than this will be handled in memory.
     802Note: ``FILE_UPLOAD_DIR`` has to be defined to enable streaming.
     803
    785804TEMPLATE_CONTEXT_PROCESSORS
    786805---------------------------
    787806
  • tests/modeltests/test_client/models.py

     
    33
    44The test client is a class that can act like a simple
    55browser for testing purposes.
    6  
     6
    77It allows the user to compose GET and POST requests, and
    88obtain the response that the server gave to those requests.
    99The server Response objects are annotated with the details
     
    7575        self.assertEqual(response.template.name, "Book template")
    7676        self.assertEqual(response.content, "Blink - Malcolm Gladwell")
    7777
     78    def test_post_file_view(self):
     79        "POST this python file to a view"
     80        import os, tempfile
     81        from django.conf import settings
     82        file = __file__.replace('.pyc', '.py')
     83        for upload_dir in [None, tempfile.gettempdir()]:
     84            settings.FILE_UPLOAD_DIR = upload_dir
     85            post_data = { 'name': file, 'file': open(file) }
     86            response = self.client.post('/test_client/post_file_view/', post_data)
     87            self.failUnless('models.py' in response.context['file']['filename'])
     88            self.failUnless(len(response.context['file']['content']) == os.path.getsize(file))
     89            if upload_dir:
     90                self.failUnless(response.context['file']['tmpfilename'])
     91
    7892    def test_redirect(self):
    7993        "GET a URL that redirects elsewhere"
    8094        response = self.client.get('/test_client/redirect_view/')
  • tests/modeltests/test_client/urls.py

     
    44urlpatterns = patterns('',
    55    (r'^get_view/$', views.get_view),
    66    (r'^post_view/$', views.post_view),
     7    (r'^post_file_view/$', views.post_file_view),
    78    (r'^raw_post_view/$', views.raw_post_view),
    89    (r'^redirect_view/$', views.redirect_view),
    910    (r'^form_view/$', views.form_view),
  • tests/modeltests/test_client/views.py

     
    4444
    4545    return HttpResponse(t.render(c))
    4646
     47def post_file_view(request):
     48    "A view that expects a multipart post and returns a file in the context"
     49    t = Template('File {{ file.filename }} received', name='POST Template')
     50    c = Context({'file': request.FILES['file_file']})
     51    return HttpResponse(t.render(c))
     52
    4753def redirect_view(request):
    4854    "A view that redirects all requests to the GET view"
    4955    return HttpResponseRedirect('/test_client/get_view/')
Back to Top