Scrapy Sticky Meta

Handy tools to maintain persistent meta values between requests in Scrapy spiders. Available as spider middleware and spider callback decorators.

Installation

pip install stickymeta

Usage

As spider middleware

Add middleware to settings.py:

SPIDER_MIDDLEWARES = {
    ...
    'stickymeta.StickyMetaMiddleware': 543,
    ...
}

and sticky_meta attribute containing persistent meta keys to spider:

class TheSpider(scrapy.Spider):
    name = 'thespider'
    sticky_meta = ('cookiejar', 'foo', 'bar')

All values for the corresponding keys will be kept persistent between all the requests and responses.

As decorators

@stick_meta

from stickymeta import stick_meta

Keep persistent values for keys passed as decorator parameters:

@stick_meta('a', 'b', 'c')
def parse(self, response):
    ...
    yield scrapy.Request(url)

is equivalent to:

def parse(self, response)
    ...
    meta = {
        'a': response.meta['a'],
        'b': response.meta['b'],
        'c': response.meta['c'],
    }
    yield scrapy.Request(url, meta=meta}

@stick_cj

from stickymeta import stick_cj

Shortcut for stick_meta handling cookiejar as default argument value, so

@stick_cj('a', 'b', 'c')
def parse(self,response):
    ...

is equivalent to

@stick_meta('cookiejar', 'a', 'b', 'c')
def parse(self,response):
    ...

Spider attribute sticky_meta also affects to decorators, next two pieces of code will handle meta in the same way:

class TheSpider(scrapy.Spider):
    name = 'thespider'
    sticky_meta = ('a', 'b', 'c')

    @stick_meta()
    def parse(self, response):
        ...
        yield Request(url)

vs

class TheSpider(scrapy.Spider):
    name = 'thespider'

    @stick_meta('a', 'b', 'c')
    def parse(self, response):
        ...
        yield Request(url)