Scraping javascript-generated data using Python

Question

I want to scrape some data of following url using Python. http://www.hankyung.com/stockplus/main.php?module=stock&mode=stock_analysis_infomation&itemcode=078340

It's about a summary of company information.

What I want to scrape is not shown on the first page. By clicking tab named "재무제표", you can access financial statement. And clicking tab named "현금흐름표', you can access "Cash Flow".

I want to scrape the "Cash Flow" data.

However, Cash flow data is generated by javascript across the url. The following link is that url which is hidden, http://stock.kisline.com/compinfo/financial/main.action?vhead=N&vfoot=N&vstay=&omit=&vwidth=

Cash flow data is generated by submitting some option value and cookie to this url.

As you perceived, itemcode=078340 in the first link means stock code and there are as many as 1680 stocks that I want gather cash flow data. I want make it a loop structure.

Is there good way to scrape cash flow data? I tried scrapy but scrapy is difficult to cope with my another scraping code already I'm using.

Solution 1

There's also dryscape (a library written by me, so the recommendation is a bit biased, obviously :) which uses a fast Webkit-based in-memory browser to navigate around. It understands Javascript, too, but is a lot more lightweight than Selenium.

Solution 2

If you need to scape the page content which is updated with AJAX and you are not in the control of this AJAX interface I would use Selenium browser automator for the task:

http://code.google.com/p/selenium/

• Selenium has Python bindings

• It launches a real browser instance so it can do and scrape 100% the same thing as you see with your own eyes

• Get HTML document content after AJAX updates thru Selenium API

• Use lxml + xpath / CSS selectors to parse out the relevant parts out of the document

Comments

[trigger]

Thanks a lot. I'm gonna try selenium.

[abbood]

can i substitute jquery with this lxml +xpath part at the end (and follow the rest of the steps)?

[Mikko Ohtamaa]

Selenium comes with its own CSS selector engine (which probably uses the underlying browser), so you don't need neither jQuery nor lxml anymore