Source code for revproxy.utils


import re

import logging

from wsgiref.util import is_hop_by_hop

try:
    from http.cookies import SimpleCookie
    COOKIE_PREFIX = ''
except ImportError:
    from Cookie import SimpleCookie
    COOKIE_PREFIX = 'Set-Cookie: '


#: List containing string constant that are used to represent headers that can
#: be ignored in the required_header function
IGNORE_HEADERS = (
    'HTTP_ACCEPT_ENCODING',  # We want content to be uncompressed so
                             # we remove the Accept-Encoding from
                             # original request
    'HTTP_HOST',
    'HTTP_REMOTE_USER',
)


# Default from HTTP RFC 2616
#   See: http://www.w3.org/Protocols/rfc2616/rfc2616-sec3.html#sec3.7.1
#: Variable that represent the default charset used
DEFAULT_CHARSET = 'latin-1'

#: List containing string constants that represents possible html content type
HTML_CONTENT_TYPES = (
    'text/html',
    'application/xhtml+xml'
)

#: Variable used to represent a minimal content size required for response
#: to be turned into stream
MIN_STREAMING_LENGTH = 4 * 1024  # 4KB

#: Regex used to find charset in a html content type
_get_charset_re = re.compile(r';\s*charset=(?P<charset>[^\s;]+)', re.I)
#: Regex used to clean extra HTTP prefixes in headers
_get_header_name_re = re.compile(
    r'((http[-|_])*)(?P<header_name>(http[-|_]).*)',
    re.I,
)




[docs]
def is_html_content_type(content_type):
    """Function used to verify if the parameter is a proper html content type

    :param content_type: String variable that represent a content-type
    :returns:  A boolean value stating if the content_type is a valid html
               content type
    """
    for html_content_type in HTML_CONTENT_TYPES:
        if content_type.startswith(html_content_type):
            return True

    return False






[docs]
def should_stream(proxy_response):
    """Function to verify if the proxy_response must be converted into
    a stream.This will be done by checking the proxy_response content-length
    and verify if its length is bigger than one stipulated
    by MIN_STREAMING_LENGTH.

    :param proxy_response: An Instance of urllib3.response.HTTPResponse
    :returns: A boolean stating if the proxy_response should
              be treated as a stream
    """
    content_type = proxy_response.headers.get('Content-Type')

    if is_html_content_type(content_type):
        return False

    try:
        content_length = int(proxy_response.headers.get('Content-Length', 0))
    except ValueError:
        content_length = 0

    if not content_length or content_length > MIN_STREAMING_LENGTH:
        return True

    return False






[docs]
def get_charset(content_type):
    """Function used to retrieve the charset from a content-type.If there is no
    charset in the content type then the charset defined on DEFAULT_CHARSET
    will be returned

    :param  content_type:   A string containing a Content-Type header
    :returns:               A string containing the charset
    """
    if not content_type:
        return DEFAULT_CHARSET

    matched = _get_charset_re.search(content_type)
    if matched:
        # Extract the charset and strip its double quotes
        return matched.group('charset').replace('"', '')
    return DEFAULT_CHARSET






[docs]
def required_header(header):
    """Function that verify if the header parameter is an essential header

    :param header:  A string represented a header
    :returns:       A boolean value that represent if the header is required
    """
    matched = _get_header_name_re.search(header)

    # Ensure there is only one HTTP prefix in the header
    header_name = matched.group('header_name') if matched else header

    header_name_upper = header_name.upper().replace('-', '_')

    if header_name_upper in IGNORE_HEADERS:
        return False

    if header_name_upper.startswith('HTTP_') or header == 'CONTENT_TYPE':
        return True

    return False






[docs]
def set_response_headers(response, response_headers):
    # check for Django 3.2 headers interface
    # https://code.djangoproject.com/ticket/31789
    # check and set pointer before loop to improve efficiency
    if hasattr(response, 'headers'):
        headers = response.headers
    else:
        headers = response

    for header, value in response_headers.items():
        if is_hop_by_hop(header) or header.lower() == 'set-cookie':
            continue

        headers[header] = value

    if hasattr(response, 'headers'):
        logger.debug('Response headers: %s', response.headers)
    else:
        logger.debug('Response headers: %s',
                     getattr(response, '_headers', None))






[docs]
def normalize_request_headers(request):
    r"""Function used to transform header, replacing 'HTTP\_' to ''
    and replace '_' to '-'

    :param request:  A HttpRequest that will be transformed
    :returns:        A dictionary with the normalized headers
    """
    norm_headers = {}
    for header, value in request.META.items():
        if required_header(header):
            norm_header = header.replace('HTTP_', '').title().replace('_', '-')
            norm_headers[norm_header] = value

    return norm_headers






[docs]
def encode_items(items):
    """Function that encode all elements in the list of items passed as
    a parameter

    :param items:  A list of tuple
    :returns:      A list of tuple with all items encoded in 'utf-8'
    """
    encoded = []
    for key, values in items:
        for value in values:
            encoded.append((key.encode('utf-8'), value.encode('utf-8')))
    return encoded




logger = logging.getLogger('revproxy.cookies')








[docs]
def unquote(value):
    """Remove wrapping quotes from a string.

    :param value: A string that might be wrapped in double quotes, such
                  as a HTTP cookie value.
    :returns: Beginning and ending quotes removed and escaped quotes (``\"``)
              unescaped
    """
    if len(value) > 1 and value[0] == '"' and value[-1] == '"':
        value = value[1:-1].replace(r'\"', '"')
    return value