Did you know the entire web was made of data? You probably did. Scrapekit helps you get that data with simple Python scripts. Based on requests, the library will handles caching, threading and logging.
See the full documentation.
from scrapekit import Scraper
scraper = Scraper('example')
@scraper.task
def get_index():
url = 'http://databin.pudo.org/t/b2d9cf'
doc = scraper.get(url).html()
for row in doc.findall('.//tr'):
yield row
@scraper.task
def get_row(row):
columns = row.findall('./td')
print(columns)
pipeline = get_index | get_row
if __name__ == '__main__':
pipeline.run()
Scrapekit doesn't aim to provide all functionality necessary for scraping. Specifically, it doesn't address HTML parsing, data storage and data validation. For these needs, check the following libraries:
- lxml for HTML/XML parsing; much faster and more flexible than BeautifulSoup.
- dataset is a sister library of scrapekit that simplifies storing semi-structured data in SQL databases.
- Scrapy is a much more mature and comprehensive framework for developing scrapers. On the other hand, it requires you to develop scrapers within its class system. This can be too heavyweight for a simple script to grab data off a web site.
- scrapelib is a thin wrapper around requests that does throttling, retries and caching.
- MechanicalSoup binds BeautifulSoup and requests into an imperative, stateful API.
Scrapekit is licensed under the terms of the MIT license, which is also included in LICENSE. It was developed through projects of ICFJ, ANCIR and ICIJ.