root/docs/topics/request-response.rst @ 1936:a32f3fe4fe6d

Revision 1936:a32f3fe4fe6d, 21.3 kB (checked in by Pablo Hoffman <pablo@…>, 7 months ago)

Fixed encoding issue (reported in #135) when the encoding declared in the HTTP header is unknown. This is the patch proposed by Rolando, with an update to the Request/Response documentation.

Requests and Responses

System Message: ERROR/3 (<string>, line 7)

Unknown directive type "module".

.. module:: scrapy.http
   :synopsis: Request and Response classes

Scrapy uses :class:`Request` and :class:`Response` objects for crawling web sites.

System Message: ERROR/3 (<string>, line 10); backlink

Unknown interpreted text role "class".

System Message: ERROR/3 (<string>, line 10); backlink

Unknown interpreted text role "class".

Typically, :class:`Request` objects are generated in the spiders and pass across the system until they reach the Downloader, which executes the request and returns a :class:`Response` object which travels back to the spider that issued the request.

System Message: ERROR/3 (<string>, line 13); backlink

Unknown interpreted text role "class".

System Message: ERROR/3 (<string>, line 13); backlink

Unknown interpreted text role "class".

Both :class:`Request` and :class:`Response` classes have subclasses which adds additional functionality not required in the base classes. These are described below in :ref:`topics-request-response-ref-request-subclasses` and :ref:`topics-request-response-ref-response-subclasses`.

System Message: ERROR/3 (<string>, line 18); backlink

Unknown interpreted text role "class".

System Message: ERROR/3 (<string>, line 18); backlink

Unknown interpreted text role "class".

System Message: ERROR/3 (<string>, line 18); backlink

Unknown interpreted text role "ref".

System Message: ERROR/3 (<string>, line 18); backlink

Unknown interpreted text role "ref".

Request objects

A :class:`Request` object represents an HTTP request, which is usually generated in the Spider and executed by the Downloader, and thus generating a :class:`Response`.

System Message: ERROR/3 (<string>, line 29); backlink

Unknown interpreted text role "class".

System Message: ERROR/3 (<string>, line 29); backlink

Unknown interpreted text role "class".
param url:

the URL of this request

type url:

string

param callback:

the function that will be called with the response of this request (once its downloaded) as its first parameter. For more information see :ref:`topics-request-response-ref-request-callback-arguments` below.

System Message: ERROR/3 (<string>, line 36); backlink

Unknown interpreted text role "ref".

type callback:

callable

param method:

the HTTP method of this request. Defaults to 'GET'.

type method:

string

param meta:

the initial values for the :attr:`Request.meta` attribute. If given, the dict passed in this parameter will be shallow copied.

System Message: ERROR/3 (<string>, line 44); backlink

Unknown interpreted text role "attr".

type meta:

dict

param body:

the request body. If a unicode is passed, then it's encoded to str using the encoding passed (which defaults to utf-8). If body is not given,, an empty string is stored. Regardless of the type of this argument, the final value stored will be a str` (never unicode or None).

type body:

str or unicode

param headers:

the headers of this request. The dict values can be strings (for single valued headers) or lists (for multi-valued headers).

type headers:

dict

param cookies:

the request cookies. Example:

request_with_cookies = Request(url="http://www.example.com",
                               cookies={currency: 'USD', country: 'UY'})

When some site returns cookies (in a response) those are stored in the cookies for that domain and will be sent again in future requests. That's the typical behaviour of any regular web browser. However, if, for some reason, you want to avoid merging with existing cookies you can instruct Scrapy to do so by setting the dont_merge_cookies item in the :attr:`Request.meta`.

System Message: ERROR/3 (<string>, line 64); backlink

Unknown interpreted text role "attr".

Example of request without merging cookies:

request_with_cookies = Request(url="http://www.example.com",
                               cookies={currency: 'USD', country: 'UY'},
                               meta={'dont_merge_cookies': True})
type cookies:

dict

param encoding:

the encoding of this request (defaults to 'utf-8'). This encoding will be used to percent-encode the URL and to convert the body to str (if given as unicode).

type encoding:

string

param priority:

the priority of this request (defaults to 0.0). The priority is used by the scheduler to define the order used to return requests. It can also be used to feed priorities externally, for example, using an offline long-term scheduler.

type encoding:

int or float

param dont_filter:
 

indicates that this request should not be filtered by the scheduler. This is used when you want to perform an identical request multiple times, to ignore the duplicates filter. Use it with care, or you will get into crawling loops. Default to False.

type dont_filter:
 

boolean

param errback:

a function that will be called if any exception was raised while processing the request. This includes pages that failed with 404 HTTP errors and such. It receives a Twisted Failure instance as first parameter.

type errback:

callable

System Message: ERROR/3 (<string>, line 103)

Unknown directive type "attribute".

.. attribute:: Request.url
    A string containing the URL of this request. Keep in mind that this
    attribute contains the escaped URL, so it can differ from the URL passed in
    the constructor.

System Message: ERROR/3 (<string>, line 109)

Unknown directive type "attribute".

.. attribute:: Request.method
    A string representing the HTTP method in the request. This is guaranteed to
    be uppercase. Example: ``"GET"``, ``"POST"``, ``"PUT"``, etc

System Message: ERROR/3 (<string>, line 114)

Unknown directive type "attribute".

.. attribute:: Request.headers
    A dictionary-like object which contains the request headers.

System Message: ERROR/3 (<string>, line 118)

Unknown directive type "attribute".

.. attribute:: Request.body
    A str that contains the request body

System Message: ERROR/3 (<string>, line 122)

Unknown directive type "attribute".

.. attribute:: Request.meta
    A dict that contains arbitrary metadata for this request. This dict is
    empty for new Requests, and is usually  populated by different Scrapy
    components (extensions, middlewares, etc). So the data contained in this
    dict depends on the extensions you have enabled.
    This dict is `shallow copied`_ when the request is cloned using the
    ``copy()`` or ``replace()`` methods.

System Message: ERROR/3 (<string>, line 134)

Unknown directive type "method".

.. method:: Request.copy()
   Return a new Request which is a copy of this Request. See also:
   :ref:`topics-request-response-ref-request-callback-arguments`.

System Message: ERROR/3 (<string>, line 139)

Unknown directive type "method".

.. method:: Request.replace([url, callback, method, headers, body, cookies, meta, encoding, dont_filter])
   Return a Request object with the same members, except for those members
   given new values by whichever keyword arguments are specified. The
   attribute :attr:`Request.meta` is copied by default (unless a new value
   is given in the ``meta`` argument). See also
   :ref:`topics-request-response-ref-request-callback-arguments`.

Caveats with copying Requests and callbacks

When you copy a request using the :meth:`Request.copy` or :meth:`Request.replace` methods the callback of the request is not copied by default. This is because of legacy reasons along with limitations in the underlying network library, which doesn't allow sharing Twisted deferreds.

System Message: ERROR/3 (<string>, line 152); backlink

Unknown interpreted text role "meth".

System Message: ERROR/3 (<string>, line 152); backlink

Unknown interpreted text role "meth".

For example:

request = Request("http://www.example.com", callback=myfunc)
request2 = request.copy() # doesn't copy the callback
request3 = request.replace(callback=request.callback)

In the above example, request2 is a copy of request but it has no callback, while request3 is a copy of request and also contains the callback.

Passing arguments to callback functions

The callback of a request is a function that will be called when the response of that request is downloaded. The callback function will be called with the :class:`Response` object downloaded as its first argument.

System Message: ERROR/3 (<string>, line 174); backlink

Unknown interpreted text role "class".

Example:

def parse_page1(self, response):
    request = Request("http://www.example.com/some_page.html",
                      callback=self.parse_page2)
def parse_page2(self, response):
    # this would log http://www.example.com/some_page.html
    self.log("Visited %s" % response.url)

In some cases you may be interested in passing arguments to those callback functions so you can receive those arguments later, when the response is downloaded. There are two ways for doing this:

  1. using a lambda function (or any other function/callable)

  2. using the :attr:`Request.meta` attribute.

    System Message: ERROR/3 (<string>, line 194); backlink

    Unknown interpreted text role "attr".

Here's an example of logging the referer URL of each page using each mechanism. Keep in mind, however, that the referer URL could be accessed easier via response.request.url).

Using lambda function:

def parse_page1(self, response):
    myarg = response.url
    request = Request("http://www.example.com/some_page.html",
                      callback=lambda r: self.parse_page2(r, myarg))
def parse_page2(self, response, referer_url):
    self.log("Visited page %s from %s" % (response.url, referer_url))

Using Request.meta:

def parse_page1(self, response):
    request = Request("http://www.example.com/some_page.html",
                      callback=self.parse_page2)
    request.meta['referer_url'] = response.url
def parse_page2(self, response):
    referer_url = response.request.meta['referer_url']
    self.log("Visited page %s from %s" % (response.url, referer_url))

Request subclasses

Here is the list of built-in :class:`Request` subclasses. You can also subclass it to implement your own custom functionality.

System Message: ERROR/3 (<string>, line 226); backlink

Unknown interpreted text role "class".

FormRequest objects

The FormRequest class extends the base :class:`Request` with functionality for dealing with HTML forms. It uses the ClientForm library (bundled with Scrapy) to pre-populate form fields with form data from :class:`Response` objects.

System Message: ERROR/3 (<string>, line 232); backlink

Unknown interpreted text role "class".

System Message: ERROR/3 (<string>, line 232); backlink

Unknown interpreted text role "class".

System Message: ERROR/3 (<string>, line 239)

Invalid class attribute value for "class" directive: "FormRequest(url, [formdata, ...])".

.. class:: FormRequest(url, [formdata, ...])
    The :class:`FormRequest` class adds a new argument to the constructor. The
    remaining arguments are the same as for the :class:`Request` class and are
    not documented here.
    :param formdata: is a dictionary (or iterable of (key, value) tuples)
       containing HTML Form data which will be url-encoded and assigned to the
       body of the request.
    :type formdata: dict or iterable of tuples
    The :class:`FormRequest` objects support the following class method in
    addition to the standard :class:`Request` methods:
    .. classmethod:: FormRequest.from_response(response, [formnumber=0, formdata=None, clickdata=None, dont_click=False, ...])
       Returns a new :class:`FormRequest` object with its form field values
       pre-populated with those found in the HTML ``<form>`` element contained
       in the given response. For an example see
       :ref:`topics-request-response-ref-request-userlogin`.
       Keep in mind that this method is implemented using `ClientForm`_ whose
       policy is to automatically simulate a click, by default, on any form
       control that looks clickable, like a a ``<input type="submit">``.  Even
       though this is quite convenient, and often the desired behaviour,
       sometimes it can cause problems which could be hard to debug. For
       example, when working with forms that are filled and/or submitted using
       javascript, the default :meth:`from_response` (and `ClientForm`_)
       behaviour may not be the most appropiate. To disable this behaviour you
       can set the ``dont_click`` argument to ``True``. Also, if you want to
       change the control clicked (instead of disabling it) you can also use
       the ``clickdata`` argument.
       :param response: the response containing a HTML form which will be used
          to pre-populate the form fields
       :type response: :class:`Response` object
       :param formnumber: the number of form to use, when the response contains
          multiple forms. The first one (and also the default) is ``0``.
       :type formnumber: integer
       :param formdata: fields to override in the form data. If a field was
          already present in the response ``<form>`` element, its value is
          overridden by the one passed in this parameter.
       :type formdata: dict
       :param clickdata: Arguments to be passed directly to ClientForm
          ``click_request_data()`` method. See `ClientForm`_ homepage for
          more info.
       :type clickdata: dict
       :param dont_click: If True the form data will be sumbitted without
         clicking in any element.
       :type clickdata: boolean
       The other parameters of this class method are passed directly to the
       :class:`FormRequest` constructor.

Request usage examples

Using FormRequest to send data via HTTP POST

If you want to simulate a HTML Form POST in your spider, and send a couple of key-value fields you could return a :class:`FormRequest` object (from your spider) like this:

System Message: ERROR/3 (<string>, line 303); backlink

Unknown interpreted text role "class".
return [FormRequest(url="http://www.example.com/post/action",
                    formdata={'name': 'John Doe', age: '27'},
                    callback=self.after_post)]

Using FormRequest.from_response() to simulate a user login

It is usual for web sites to provide pre-populated form fields through <input type="hidden"> elements, such as session related data or authentication tokens (for login pages). When scraping, you'll want these fields to be automatically pre-populated and only override a couple of them, such as the user name and password. You can use the :meth:`FormRequest.from_response` method for this job. Here's an example spider which uses it:

System Message: ERROR/3 (<string>, line 316); backlink

Unknown interpreted text role "meth".
class LoginSpider(BaseSpider):
    domain_name = 'example.com'
    start_urls = ['http://www.example.com/users/login.php']
    def parse(self, response):
        return [FormRequest.from_response(response,
                    formdata={'username': 'john', 'password': 'secret'},
                    callback=self.after_login)]
    def after_login(self, response):
        # check login succeed before going on
        if "authentication failed" in response.body:
            self.log("Login failed", level=log.ERROR)
            return
        # continue scraping with authenticated session...

Response objects

A :class:`Response` object represents an HTTP response, which is usually downloaded (by the Downloader) and fed to the Spiders for processing.

System Message: ERROR/3 (<string>, line 346); backlink

Unknown interpreted text role "class".
param url:

the URL of this response

type url:

string

param headers:

the headers of this response. The dict values can be strings (for single valued headers) or lists (for multi-valued headers).

type headers:

dict

param status:

the HTTP status of the response. Defaults to 200.

type status:

integer

param body:

the response body. It must be str, not unicode, unless you're using a encoding-aware :ref:`Response subclass <topics-request-response-ref-response-subclasses>`, such as :class:`TextResponse`.

System Message: ERROR/3 (<string>, line 359); backlink

Unknown interpreted text role "ref".

System Message: ERROR/3 (<string>, line 359); backlink

Unknown interpreted text role "class".

type body:

str

param meta:

the initial values for the :attr:`Response.meta` attribute. If given, the dict will be shallow copied.

System Message: ERROR/3 (<string>, line 365); backlink

Unknown interpreted text role "attr".

type meta:

dict

param flags:

is a list containing the initial values for the :attr:`Response.flags` attribute. If given, the list will be shallow copied.

System Message: ERROR/3 (<string>, line 369); backlink

Unknown interpreted text role "attr".

type flags:

list

System Message: ERROR/3 (<string>, line 374)

Unknown directive type "attribute".

.. attribute:: Response.url
    A string containing the URL of the response.

System Message: ERROR/3 (<string>, line 378)

Unknown directive type "attribute".

.. attribute:: Response.status
    An integer representing the HTTP status of the response. Example: ``200``,
    ``404``.

System Message: ERROR/3 (<string>, line 383)

Unknown directive type "attribute".

.. attribute:: Response.headers
    A dictionary-like object which contains the response headers.

System Message: ERROR/3 (<string>, line 387)

Unknown directive type "attribute".

.. attribute:: Response.body
    A str containing the body of this Response. Keep in mind that Reponse.body
    is always a str. If you want the unicode version use
    :meth:`TextResponse.body_as_unicode` (only available in
    :class:`TextResponse` and subclasses).

System Message: ERROR/3 (<string>, line 394)

Unknown directive type "attribute".

.. attribute:: Response.request
    The :class:`Request` object that generated this response. This attribute is
    assigned in the Scrapy engine, after the response and request has passed
    through all :ref:`Downloader Middlewares <topics-downloader-middleware>`.
    In particular, this means that:
    - HTTP redirections will cause the original request (to the URL before
      redirection) to be assigned to the redirected response (with the final
      URL after redirection).
    - Response.request.url doesn't always equals Response.url
    - This attribute is only available in the spider code, and in the
      :ref:`Spider Middlewares <topics-spider-middleware>`, but not in
      Downloader Middlewares (although you have the Request available there by
      other means) and handlers of the :signal:`response_downloaded` signal.

System Message: ERROR/3 (<string>, line 412)

Unknown directive type "attribute".

.. attribute:: Response.meta
    A dict that contains arbitrary metadata for this response, similar to the
    :attr:`Request.meta` attribute. See the :attr:`Request.meta` attribute for
    more info.

System Message: ERROR/3 (<string>, line 418)

Unknown directive type "attribute".

.. attribute:: Response.flags
    A list that contains flags for this response. Flags are labels used for
    tagging Responses. For example: `'cached'`, `'redirected`', etc. And
    they're shown on the string representation of the Response (`__str__`
    method) which is used by the engine for logging.

System Message: ERROR/3 (<string>, line 425)

Unknown directive type "method".

.. method:: Response.copy()
   Return a new Response which is a copy of this Response.

System Message: ERROR/3 (<string>, line 429)

Unknown directive type "method".

.. method:: Response.replace([url, status, headers, body, meta, flags, cls])
   Return a Response object with the same members, except for those members
   given new values by whichever keyword arguments are specified. The
   attribute :attr:`Response.meta` is copied by default (unless a new value
   is given in the ``meta`` argument).

Response subclasses

Here is the list of available built-in Response subclasses. You can also subclass the Response class to implement your own functionality.

TextResponse objects

System Message: ERROR/3 (<string>, line 447)

Invalid class attribute value for "class" directive: "TextResponse(url, [encoding[, ...]])".

.. class:: TextResponse(url, [encoding[, ...]])
    :class:`TextResponse` objects adds encoding capabilities to the base
    :class:`Response` class, which is meant to be used only for binary data,
    such as images, sounds or any media file.
    :class:`TextResponse` objects support a new constructor arguments, in
    addition to the base :class:`Response` objects. The remaining functionality
    is the same as for the :class:`Response` class and is not documented here.
    :param encoding: is a string which contains the encoding to use for this
       response. If you create a :class:`TextResponse` object with a unicode
       body it will be encoded using this encoding (remember the body attribute
       is always a string). If ``encoding`` is ``None`` (default value), the
       encoding will be looked up in the response headers anb body instead.
    :type encoding: string
    :class:`TextResponse` objects support the following attributes in addition
    to the standard :class:`Response` ones:
    .. attribute:: TextResponse.encoding
       A string with the encoding of this response. The encoding is resolved by
       trying the following mechanisms, in order:
       1. the encoding passed in the constructor `encoding` argument
       2. the encoding declared in the Content-Type HTTP header. If this
          encoding is not valid (ie. unknown), it is ignored and the next
          resolution mechanism is tried.
       3. the encoding declared in the response body. The TextResponse class
          doesn't provide any special functionality for this. However, the
          :class:`HtmlResponse` and :class:`XmlResponse` classes do.
       4. the encoding inferred by looking at the response body. This is the more
          fragile method but also the last one tried.
    :class:`TextResponse` objects support the following methods in addition to
    the standard :class:`Response` ones:
    .. method:: TextResponse.headers_encoding()
        Returns a string with the encoding declared in the headers (ie. the
        Content-Type HTTP header).
    .. method:: TextResponse.body_encoding()
        Returns a string with the encoding of the body, either declared or inferred
        from its contents. The body encoding declaration is implemented in
        :class:`TextResponse` subclasses such as: :class:`HtmlResponse` or
        :class:`XmlResponse`.
    .. method:: TextResponse.body_as_unicode()
        Returns the body of the response as unicode. This is equivalent to::
            response.body.encode(response.encoding)
        But **not** equivalent to::
            unicode(response.body)
        Since, in the latter case, you would be using you system default encoding
        (typically `ascii`) to convert the body to uniode, instead of the response
        encoding.

HtmlResponse objects

System Message: ERROR/3 (<string>, line 517)

Invalid class attribute value for "class" directive: "HtmlResponse(url[, ...])".

.. class:: HtmlResponse(url[, ...])
    The :class:`HtmlResponse` class is a subclass of :class:`TextResponse`
    which adds encoding auto-discovering support by looking into the HTML `meta
    http-equiv`_ attribute.  See :attr:`TextResponse.encoding`.

XmlResponse objects

System Message: ERROR/3 (<string>, line 528)

Invalid class attribute value for "class" directive: "XmlResponse(url[, ...])".

.. class:: XmlResponse(url[, ...])
    The :class:`XmlResponse` class is a subclass of :class:`TextResponse` which
    adds encoding auto-discovering support by looking into the XML declaration
    line.  See :attr:`TextResponse.encoding`.
Note: See TracBrowser for help on using the browser.