CODEWITHSUNDEEP
pythonxpathweb-scrapinglxml
Duccio A
Duccio A

Reputation: 1562

Python Web scraping: extract one attribute with multiple tags

I am trying to scrape data from my account on an online bookmark service. The page with the bookmarks is organised as following:

<!DOCTYPE html>
<html lang="en">
<body>
<div id="item1" class="outer_block">
    <div class="title">Bookmark 1</div>
    <div class="link">
        <a href="https://bookmark1.com">https://bookmark1.com</a>
    </div>
    <div class="tags">
        <a href="http://mylink.com/tag1">tag1</a>
        <a href="http://mylink.com/tag2">tag2</a>
    </div>
</div>
<div id="item2" class="outer_block">
    <div class="title">Bookmark 2</div>
    <div class="link">
        <a href="https://bookmark2.com">https://bookmark2.com</a>
    </div>
    <div class="tags">
        <a href="http://mylink.com/tag1">tag1</a>
    </div>
</div>
<div id="item3" class="outer_block">
    <div class="title">Bookmark 3</div>
    <div class="link">
        <a href="https://bookmark3.com">https://bookmark3.com</a>
    </div>
    <div class="tags">
        <a href="http://mylink.com/tag3">tag3</a>
    </div>
</div>
</body>
</html>

For each block I would like to extract the title, the link and the tags. In Python 3.5, I do:

# Import modules
import requests
from lxml import html

# Read the html
# url = 'mylink'
# page = requests.get(url)
# tree = html.fromstring(page.content)
# This is the replicable example
tree = html.fromstring('<!DOCTYPE html><html lang="en"><body><div id="item1" class="outer_block"> <div class="title">Item 1</div> <div class="link"> <a href="https://bookmark1.com">https://bookmark1.com</a> </div> <div class="tags"> <a href="http://mylink.com/tag1">tag1</a> <a href="http://mylink.com/tag2">tag2</a> </div></div><div id="item2" class="outer_block"> <div class="title">Item 2</div> <div class="link"> <a href="https://bookmark2.com">https://bookmark2.com</a> </div> <div class="tags"> <a href="http://mylink.com/tag1">tag1</a> </div></div><div id="item3" class="outer_block"> <div class="title">Item 3</div> <div class="link"> <a href="https://bookmark3.com">https://bookmark3.com</a> </div> <div class="tags"> <a href="http://mylink.com/tag3">tag3</a> </div></div></body></html>')

I use xpath to extract patterns of strings, say the title:

titles = tree.xpath('//div[@class="title"]/text()')
print(titles)

['Bookmark 1', 'Bookmark 2', 'Bookmark 3']

In order to extract the tags, I use the same principle:

tags = tree.xpath('//div[@class="tags"]//a/text()')
print(tags)

['tag1', 'tag2', 'tag1', 'tag3']

The problem is that each link has various tags so I cannot associate the array titles with the array tags. I thought I could extract each block and then work on them separately:

blocks = tree.xpath('//div[@class="outer_block"]')
block1 = blocks[0]

What I don't understand is that when I extract the tags from block1, it still maintains all of the tags of the original html.

tags_block1 = block1.xpath('//div[@class="tags"]//a/text()'
print(tags_block1)

['tag1', 'tag2', 'tag1', 'tag3']

How do I extract the title and the corresponding tags, what is the best output format and is there any other package that could do the job more easily?

Upvotes: 1

Views: 1422

Answers (2)

Mohsen Fard
Mohsen Fard

Reputation: 598

You can use two property in two different brackets

description = tree.xpath("//div[@class='details-content'][@itemprop='description']/text()")

Upvotes: 0

Mike R
Mike R

Reputation: 302

You should think about using BeautifulSoup. Consider the code below (source is a string of the HTML):

from bs4 import BeautifulSoup 

soup = BeautifulSoup(source, "html.parser")
outer_blocks = soup.find_all("div", class_="outer_block")

for block in outer_blocks:
    title = block.find("div", class_="title").contents[0]
    link = block.find("a").contents[0]
    tags = [x.contents[0] for x in block.find("div", class_="tags").find_all("a")]
    print([title, link, tags])

The output is: 

['Bookmark 1', 'https://bookmark1.com', ['tag1', 'tag2']]
['Bookmark 2', 'https://bookmark2.com', ['tag1']]
['Bookmark 3', 'https://bookmark3.com', ['tag3']]

Upvotes: 1

Related Questions