How to scrape url in Python using scrapy

Question

I want to extract URLs from a particular website using scrapy in python which has the following HTML structure

<div class="comic-table">
<div id="comic">
		<img src="http://demowebsite.com/uploads/image1" alt="" title="">
		<img src="http://demowebsite.com/uploads/image2" alt="" title="">
</div>
</div>

here is the scrapy code I have written:

import scrapy
from scrapy.contrib.spiders import Rule, CrawlSpider
from scrapy.contrib.linkextractors import LinkExtractor
from Pencils.items import PencilsItem

class Spider(CrawlSpider):
name = 'pencil'
allowed_domains = ['demowebsite.com']
start_urls = ['http://demowebsite.com']
rules = [Rule(LinkExtractor(allow=['/uploads/.*']), 'parse_pencil')]

def parse_pencil(self, response):

    image = PencilsItem()
    rel = response.xpath("WHAT_SHOULD_I_PUT_HERE").extract()
    image['image_urls'] = ['http:'+rel[0]]
    return image

what Should I put in the response.xpath field.

P.S I'm a beginner in HTML and Python

Answer 1

Try this:

    '//div[@id="comic"]/img'

//   =>  search the whole html page
@    =>  attribute 

That xpath looks for all <div> tags which have an attribute named id which is equal to "comic" (there should only be one <div> tag with the attribute id="comic" because an id should be unique), and extracts the <img> tags therein.

With scrapy you can do something like the following to get all the <img> tags:

import scrapy

class TestSpider(scrapy.Spider):
    name = "my_spider"

    start_urls = [
        "file:///Users/7stud/python_programs/scrapy_stuff/html_files/html.html"
    ]

    def parse(self, response):
        for selector in response.xpath('//div[@id="comic"]/img'):
            src = selector.xpath('@src').extract()
            print src[0]



--output:--
(scrapy_env)~/python_programs/scrapy_stuff$ scrapy crawl my_spider
2016-03-29 02:19:09 [scrapy] INFO: Scrapy 1.0.5 started (bot: scrapy_stuff)
2016-03-29 02:19:09 [scrapy] INFO: Optional features available: ssl, http11
2016-03-29 02:19:09 [scrapy] INFO: Overridden settings: {'NEWSPIDER_MODULE': 'scrapy_stuff.spiders', 'SPIDER_MODULES': ['scrapy_stuff.spiders'], 'BOT_NAME': 'scrapy_stuff'}
2016-03-29 02:19:09 [scrapy] INFO: Enabled extensions: CloseSpider, TelnetConsole, LogStats, CoreStats, SpiderState
2016-03-29 02:19:09 [scrapy] INFO: Enabled downloader middlewares: HttpAuthMiddleware, DownloadTimeoutMiddleware, UserAgentMiddleware, RetryMiddleware, DefaultHeadersMiddleware, MetaRefreshMiddleware, HttpCompressionMiddleware, RedirectMiddleware, CookiesMiddleware, ChunkedTransferMiddleware, DownloaderStats
2016-03-29 02:19:09 [scrapy] INFO: Enabled spider middlewares: HttpErrorMiddleware, OffsiteMiddleware, RefererMiddleware, UrlLengthMiddleware, DepthMiddleware
2016-03-29 02:19:09 [scrapy] INFO: Enabled item pipelines: 
2016-03-29 02:19:09 [scrapy] INFO: Spider opened
2016-03-29 02:19:09 [scrapy] INFO: Crawled 0 pages (at 0 pages/min), scraped 0 items (at 0 items/min)
2016-03-29 02:19:09 [scrapy] DEBUG: Telnet console listening on 127.0.0.1:6023
2016-03-29 02:19:09 [scrapy] DEBUG: Crawled (200) <GET file:///Users/7stud/python_programs/scrapy_stuff/html_files/html.html> (referer: None)
http://demowebsite.com/uploads/image1
http://demowebsite.com/uploads/image2
2016-03-29 02:19:09 [scrapy] INFO: Closing spider (finished)
2016-03-29 02:19:09 [scrapy] INFO: Dumping Scrapy stats:
{'downloader/request_bytes': 263,
 'downloader/request_count': 1,
 'downloader/request_method_count/GET': 1,
 'downloader/response_bytes': 243,
 'downloader/response_count': 1,
 'downloader/response_status_count/200': 1,
 'finish_reason': 'finished',
 'finish_time': datetime.datetime(2016, 3, 29, 8, 19, 9, 251971),
 'log_count/DEBUG': 2,
 'log_count/INFO': 7,
 'response_received_count': 1,
 'scheduler/dequeued': 1,
 'scheduler/dequeued/memory': 1,
 'scheduler/enqueued': 1,
 'scheduler/enqueued/memory': 1,
 'start_time': datetime.datetime(2016, 3, 29, 8, 19, 9, 139531)}
2016-03-29 02:19:09 [scrapy] INFO: Spider closed (finished)
(scrapy_env)~/python_programs/scrapy_stuff$ 

And in fact, if all you want is the src attribute from the <img> tags, you can get the src attributes directly using the following xpath:

def parse(self, response):
    for selector in response.xpath('//div[@id="comic"]/img/@src'):
        print selector.extract()

--output:--
...
2016-03-29 02:33:56 [scrapy] DEBUG: Crawled (200) <GET file:///Users/7stud/python_programs/scrapy_stuff/html_files/html.html> (referer: None)
http://demowebsite.com/uploads/image1
http://demowebsite.com/uploads/image2
2016-03-29 02:33:57 [scrapy] INFO: Closing spider (finished)
...

P.S I'm a beginner in HTML and Python

What about xml and xpath? The subject you really need to explore is xpath. But, I would suggest that as a beginner to html and xpath you should start with BeautifulSoup for scraping web pages.

Answer 2

In order to get all links you should use

response.xpath("//div[@id='comic']/img/@src").extract()

and you code will look like

import scrapy
from scrapy.contrib.spiders import Rule, CrawlSpider
from scrapy.contrib.linkextractors import LinkExtractor
from stackoverflow.items import PencilsItem

class Spider(CrawlSpider):
    name = 'pencil'
    allowed_domains = ['demowebsite.com']
    start_urls = ['http://demowebsite.com']
    rules = [Rule(LinkExtractor(allow=['/uploads/.*']), 'parse_pencil')]

    def parse_pencil(self, response):
        item = PencilsItem()
        item['image_urls'] = response.xpath("//div[@id='comic']/img/@src").extract()
        yield item

use this code if img src doesn't contain domain

from urlparse import urlparse
parsed_uri = urlparse(response.url)
domain = '{uri.scheme}://{uri.netloc}/'.format(uri=parsed_uri)
links = [domain+link for link in response.xpath("//div[@id='comic']/img/@src").extract()]