scRUBYt! 0.4.1


#1

Hello all,

After more than a year, I’d like to announce a new release of scRUBYt! -
so “scRUBYt!”.is_vaporware? == false now!

==========
scRUBwhat?

scRUBYt! is a (hopefully) easy to use, yet powerful Web scraping
framework
based on Hpricot, mechanize and/or FireWatir. It’s purpose is to free
you
from the drudgery of web page crawling, looking up HTML tags,
attributes,
XPaths, form names and other typical low-level web scraping stuff by
figuring these out from your examples copy’n’pasted from the Web page or
Firebug.

===============
In this release

Finally it is possible to use FireWatir as the agent for navigation,
enabling AJAX/more robust scraping via Firefox.
Another big news is that the RubyInline, ParseTree and Ruby2Ruby
dependency was dropped since we couldn’t solve this problem for win32
for
one year.
Of course a lot of bugs were fixed as well!

=========
CHANGELOG

  • [NEW] possibility to use FireWatir as the agent for scraping (credit:
    Glenn G.)
  • [FIX] navigation doesn’t crash if a 404/500 is returned (credit: Glen
    Gillen)
  • [NEW] navigation action: click_by_xpath to click arbitrary elements
  • [MOD] dropped dependencies: RubyInline, ParseTree, Ruby2Ruby (hooray
    for
    win32 users)
  • [NEW] scraping through frames (e.g. google analytics)
  • [MOD] exporting temporarily doesn’t work - for now, generated XPaths
    are
    printed to the screen
  • [MOD] possibility to wait after clicking link/filling textfield (to be
    able to scrape inserted AJAX stuff)
  • [NEW] possibility to fetch from a string, by specifying nil as the url
    and the html string with the :html option
  • [FIX] firewatir slowness (credit: jak4)
  • [FIX] lot of bugfixes and stability fixes

===========
What’s next

The biggest news is that scRUBYt! is going to be rewritten from scratch

the work has already been started by Glenn G… scRUBYt! has grown
too
big for our taste, so we decided to start anew, aiming for 100% rSpec
coverage, refactored code, speed/performance optimization and leaving
cruft behind. So scRUBYt! 0.4.1, the last one based on the original
scRUBYt! will be supported until the new, rewritten one (0.5.0) comes
out
and takes it’s place.

We are working on a public scraper repository
(http://github.com/scrubber/scrubyt_examples/tree/master) where you can
post your scRUBYt! snippets and check out what are the others doing.

And there are other interesting stuff in the pipeline as well - stay
tuned!

Cheers,
Glenn && Peter


http://scrubyt.org
http://www.rubypond.com (Glenn)
http://www.rubyrailways.com (Peter)


#2

On Dec 11 2008, 8:43 am, removed_email_address@domain.invalid wrote:

based on Hpricot, mechanize and/or FireWatir. It’s purpose is to free you
enabling AJAX/more robust scraping via Firefox.
Glenn G.)

  • [NEW] possibility to fetch from a string, by specifying nil as the url
    big for our taste, so we decided to start anew, aiming for 100% rSpec

Cheers,
Glenn && Peter

___http://scrubyt.orghttp://www.rubypond.com(Glenn)http://www.rubyrailways.com(Peter)

I tried to update this gem from 0.4.06 to 0.4.11 but it doesn’t seem
to see it, even when I use the --source URL option. Have others
experienced problems in updating too?

I tried

$ gem update scrubty --source ‘http://rubyforge.org/projects/scrubty

and it gave me the error message

Updating installed gems
ERROR: While executing gem … (Zlib::GzipFile::Error)
not in gzip format