Pipeline to Download PDF or Save page as PDF for scrapy item

Installation

Install scrapy-save-as-pdf using pip:

pip install scrapy-save-as-pdf

Configuration

(Optionally) if you want to use WEBDRIVER_HUB_URL, you can use docker to setup one like this:

docker run -d -p 4444:4444 -v /dev/shm:/dev/shm selenium/standalone-chrome:4.0.0-alpha-7-20201119

then WEBDRIVER_HUB_URL value is http://docker_host_ip:4444/wd/hub and we often debug on local host, so we use http://127.0.0.1:4444/wd/hub

Add the settings.py of your Scrapy project like this:

PROXY = ""
CHROME_DRIVER_PATH ='/snap/bin/chromium.chromedriver'
PDF_SAVE_PATH = "./pdfs"
PDF_SAVE_AS_PDF = False
PDF_DOWNLOAD_TIMEOUT = 60
PDF_PRINT_OPTIONS = {
    'landscape': False,
    'displayHeaderFooter': False,
    'printBackground': True,
    'preferCSSPageSize': True,
}
WEBDRIVER_HUB_URL = 'http://127.0.0.1:4444/wd/hub'

If both WEBDRIVER_HUB_URL and CHROME_DRIVER_PATH are set, we use WEBDRIVER_HUB_URL.

Enable the pipeline by adding it to ITEM_PIPELINES in your settings.py file and changing priority:

ITEM_PIPELINES = {
    'scrapy_save_as_pdf.pipelines.SaveAsPdfPipeline': -1,
}

The order should before your persist pipeline such as save to database and after your preprocess pipeline.

In the demo scrapy project, I put the SaveToQiniuPipeline after this plugin to persist pdf to the cloud.

Usage

set the pdf_url and/or url field in your yielded item

import scrapy

class MySpider(scrapy.Spider):
    start_urls = [
        "http://example.com",
    ]

    def start_requests(self):
        for url in self.start_urls:
            yield scrapy.Request(url, self.parse)

    def parse(self, response):
        yield {
            "url": "http://example.com/cate1/page1.html",
            "pdf_url": "http://example.com/cate1/page1.pdf",
        }
        yield {
            "url": "http://example.com/cate1/page2.html",
            "pdf_url": "http://example.com/cate1/page2.pdf",
        }

the pdf_url field will be populated with the downloaded pdf file location, if pdf_url field has old value then move it to origin_pdf_url field, you can handle them in your next pipeline.

Getting help

Please use github issue

Contributing