能否用Selenium和Nokogiri根据邻近标签定位元素并提取内容？

Your pseudocode is spot-on—this kind of text-based element targeting is totally feasible with both tools. Let’s break it down into working code, covering both static HTML parsing (Nokogiri alone) and dynamic page scenarios (combining Selenium + Nokogiri).

Nokogiri-Only Solution (For Static HTML) #

If you already have the parsed HTML document (like your sample snippet), here’s how to turn your pseudocode into functional Ruby:

require 'nokogiri'

# Replace this with your actual parsed document source
html = <<~HTML
<div> 
  <h2>Details</h2> 
  <ul> 
    <li><b>Height:</b>6 ft</li> 
    <li><b>Weight:</b>6 kg</li> 
    <li><b>Age:</b>6</li> 
  </ul> 
</div>
HTML

doc = Nokogiri::HTML(html)

# Locate the div containing the "Details" heading
div_of_interest = doc.at('div:has(h2:text("Details"))')

# Find the list item that includes "Weight:"
weight_li = div_of_interest.at('li:contains("Weight:")')

# Extract just the value after "Weight:"
selected_text = weight_li.text.sub(/Weight:\s*/, '').strip

puts selected_text # Outputs: "6 kg"

How this works: #

• div:has(h2:text("Details")): Uses Nokogiri’s extended CSS selectors to zero in on the div with an <h2> child containing exactly "Details".
• li:contains("Weight:"): Targets the specific list item holding the weight data.
• The regex sub(/Weight:\s*/, '') cleans up the text by removing "Weight:" and any trailing whitespace, leaving only the value.

Combining with Selenium (For Dynamic Pages) #

If your content loads dynamically (e.g., via JavaScript), use Selenium to fetch the page first, then pass the source to Nokogiri:

require 'selenium-webdriver'
require 'nokogiri'

# Initialize Selenium (use :firefox, :edge, etc., as needed)
driver = Selenium::WebDriver.for :chrome

begin
  # Navigate to your target page
  driver.get('https://your-target-url.com')

  # Parse the loaded page source with Nokogiri
  doc = Nokogiri::HTML(driver.page_source)

  # Reuse the same logic as the static solution
  div_of_interest = doc.at('div:has(h2:text("Details"))')
  weight_li = div_of_interest.at('li:contains("Weight:")')
  selected_text = weight_li.text.sub(/Weight:\s*/, '').strip

  puts selected_text # Outputs: "6 kg"
ensure
  driver.quit # Always close the driver to avoid resource leaks
end

Alternative: Use Selenium Directly (No Nokogiri Needed) #

You can also locate the element and extract the value using Selenium’s built-in methods, with XPath:

require 'selenium-webdriver'

driver = Selenium::WebDriver.for :chrome

begin
  driver.get('https://your-target-url.com')

  # Use XPath to find the exact li element
  weight_element = driver.find_element(
    xpath: '//div[h2[text()="Details"]]/ul/li[contains(text(), "Weight:")]'
  )
  
  # Clean up the text to get just the weight value
  selected_text = weight_element.text.gsub(/Weight:\s*/, '').strip

  puts selected_text # Outputs: "6 kg"
ensure
  driver.quit
end

Pro Tips: #

• For even cleaner parsing (no regex), target the <b> tag directly and grab its next sibling text: weight_li.at('b').next_sibling.text.strip—this works if the HTML structure is consistent.
• If text might have case variations or extra spaces, adjust the regex to be more flexible (e.g., sub(/Weight:\s*/i, '') for case-insensitive matching).

Content of the question originates from Stack Exchange, question author Joe Morano