crwlr-crawlervsdude
This library provides kind of a framework and a lot of ready to use, so-called steps, that you can use as building blocks, to build your own crawlers and scrapers with.
Some features: - Crawler Politeness innocent (respecting robots.txt, throttling,...) - Load URLs using - a (PSR-18) HTTP client (default is of course Guzzle) - or a headless browser (chrome) to get source after Javascript execution - Get absolute links from HTML documents link - Get sitemaps from robots.txt and get all URLs from those sitemaps - Crawl (load) all pages of a website spider - Use cookies (or don't) cookie - Use any HTTP methods (GET, POST,...) and send any headers or body - Iterate over paginated list pages repeat - Extract data from: - HTML and also XML (using CSS selectors or XPath queries) - JSON (using dot notation) - CSV (map columns) - Extract schema.org structured data in JSON-LD format from HTML documents - Keep memory usage low by using PHP Generators muscle - Cache HTTP responses during development, so you don't have to load pages again and again after every code change - Get logs about what your crawler is doing (accepts any PSR-3 LoggerInterface)
Dude (dude uncomplicated data extraction) is a very simple framework for writing web scrapers using Python decorators. The design, inspired by Flask, was to easily build a web scraper in just a few lines of code. Dude has an easy-to-learn syntax.
The simplest web scraper will look like this:
from dude import select
@select(css="a")
def get_link(element):
return {"url": element.get_attribute("href")}
dude supports multiple parser backends:
- playwright
- lxml
- parsel
- beautifulsoup
- pyppeteer
- selenium
Example Use
<?php
require_once 'vendor/autoload.php';
use Crwlr\Crawler;
$crawler = new Crawler();
$crawler->get('https://example.com', ['User-Agent' => 'webscraping.fyi']);
// more links can be followed:
$crawler->followLinks();
// and current page can be parsed:
$response = $crawler->response();
$title = $crawler->filter('title')->text();
echo $response->getContent();
</div>
<div class="lib-example" markdown>
```python
from dude import select
"""
This example demonstrates how to use Parsel + async HTTPX
To access an attribute, use:
selector.attrib["href"]
You can also access an attribute using the ::attr(name) pseudo-element, for example "a::attr(href)", then:
selector.get()
To get the text, use ::text pseudo-element, then:
selector.get()
"""
@select(css="a.url", priority=2)
async def result_url(selector):
return {"url": selector.attrib["href"]}
# Option to get url using ::attr(name) pseudo-element
@select(css="a.url::attr(href)", priority=2)
async def result_url2(selector):
return {"url2": selector.get()}
@select(css=".title::text", priority=1)
async def result_title(selector):
return {"title": selector.get()}
@select(css=".description::text", priority=0)
async def result_description(selector):
return {"description": selector.get()}
if __name__ == "__main__":
import dude
dude.run(urls=["https://dude.ron.sh"], parser="parsel")