scrapy-rss 1.1.0

Basic usage

Declare your item directly as RssItem():

import scrapy_rss

item1 = scrapy_rss.RssItem()

Or use predefined item class RssedItem with RSS field named as rss that’s instance of RssItem:

import scrapy
import scrapy_rss

class MyItem(scrapy_rss.RssedItem):
    field1 = scrapy.Field()
    field2 = scrapy.Field()
    # ...

item2 = MyItem()

Set/get item fields. Case sensitive attributes of RssItem() are appropriate to RSS elements. Attributes of RSS elements are case sensitive too. If the editor allows autocompletion then it suggests attributes for instances of RssedItem and RssItem. It’s allowed to set any subset of RSS elements (e.g. title only). For example:

from datetime import datetime

item1.title = 'RSS item title'  # set value of <title> element
title = item1.title.value  # get value of <title> element
item1.description = 'description'

item1.guid = 'item identifier'
item1.guid.isPermaLink = False  # set value of attribute isPermalink of <guid> element,
                                # isPermaLink is True by default
is_permalink = item1.guid.isPermaLink  # get value of attribute isPermalink of <guid> element
guid = item1.guid.value  # get value of element <guid>

item1.category = 'single category'
category = item1.category
item1.category = ['first category', 'second category']
first_category = item1.category[0].value # get value of the element <category> with multiple values
all_categories = [cat.value for cat in item1.category]

# direct attributes setting
item1.enclosure.url = 'http://example.com/file'
item1.enclosure.length = 0
item1.enclosure.type = 'text/plain'

# or dict based attributes setting
item1.enclosure = {'url': 'http://example.com/file', 'length': 0, 'type': 'text/plain'}
item1.guid = {'value': 'item identifier', 'isPermaLink': True}

item1.pubDate = datetime.now()  # correctly works with Python' datetimes


item2.rss.title = 'Item title'
item2.rss.guid = 'identifier'
item2.rss.enclosure = {'url': 'http://example.com/file', 'length': 0, 'type': 'text/plain'}

All allowed elements are listed in the scrapy_rss/items.py. All allowed attributes of each element with constraints and default values are listed in the scrapy_rss/elements.py. Also you can read RSS specification for more details.

RssItem derivation and namespaces

You can extend RssItem to add new XML fields that can be namespaced or not. You can specify namespaces in an attribute and/or an element constructors. Namespace prefix can be specified in the attribute/element name using double underscores as delimiter (prefix__name) or in the attribute/element constructor using ns_prefix argument. Namespace URI can be specified using ns_uri argument of the constructor.

from scrapy_rss.meta import ElementAttribute, Element
from scrapy_rss.items import RssItem

class Element0(Element):
    # attributes without special namespace
    attr0 = ElementAttribute(is_content=True, required=True)
    attr1 = ElementAttribute()

class Element1(Element):
    # attribute "prefix2:attr2" with namespace xmlns:prefix2="id2"
    attr2 = ElementAttribute(ns_prefix="prefix2", ns_uri="id2")

    # attribute "prefix3:attr3" with namespace xmlns:prefix3="id3"
    prefix3__attr3 = ElementAttribute(ns_uri="id3")

    # attribute "prefix4:attr4" with namespace xmlns:prefix4="id4"
    fake_prefix__attr4 = ElementAttribute(ns_prefix="prefix4", ns_uri="id4")

    # attribute "attr5" with default namespace xmlns="id5"
    attr5 = ElementAttribute(ns_uri="id5")

class MyXMLItem(RssItem):
    # element <elem1> without namespace
    elem1 = Element0()

    # element <elem_prefix2:elem2> with namespace xmlns:elem_prefix2="id2e"
    elem2 = Element0(ns_prefix="elem_prefix2", ns_uri="id2e")

    # element <elem_prefix3:elem3> with namespace xmlns:elem_prefix3="id3e"
    elem_prefix3__elem3 = Element1(ns_uri="id3e")

    # yet another element <elem_prefix4:elem3> with namespace xmlns:elem_prefix4="id4e"
    # (does not conflict with previous one)
    fake_prefix__elem3 = Element0(ns_prefix="elem_prefix4", ns_uri="id4e")

    # element <elem5> with default namespace xmlns="id5e"
    elem5 = Element0(ns_uri="id5e")

Access to elements and its attributes is the same as with simple items:

item = MyXMLItem()
item.title = 'Some title'
item.elem1.attr0 = 'Required content value'
item.elem1 = 'Another way to set content value'
item.elem1.attr1 = 'Some attribute value'
item.elem_prefix3__elem3.prefix3__attr3 = 'Yet another attribute value'
item.elem_prefix3__elem3.fake_prefix__attr4 = '' # non-None value is interpreted as assigned
item.fake_prefix__elem3.attr1 = 42

Several optional settings are allowed for namespaced items:

FEED_NAMESPACES

list of tuples [(prefix, URI), ...] or dictionary {prefix: URI, ...} of namespaces that must be defined in the root XML element

FEED_ITEM_CLASS or FEED_ITEM_CLS

main class of feed items (class object MyXMLItem or path to class "path.to.MyXMLItem"). Default value: RssItem. It’s used in order to extract all possible namespaces that will be declared in the root XML element.

Feed items do NOT have to be instances of this class or its subclass.

If these settings are not defined or only part of namespaces are defined then other used namespaces will be declared either in the <item> element or in its subelements when these namespaces are not unique. Each <item> element and its sublements always contains only namespace declarations of non-None attributes (including ones that are interpreted as element content).