| Subject: | Fixes and code |
| Date: | Thu, 5 Nov 2009 18:22:33 +0000 |
| To: | bug-WWW-Search-Scrape [...] rt.cpan.org |
| From: | Oliver Elder <oliver [...] wahanda.com> |
Hi
Thanks for providing the scraper code - it is much appreciated.
I know this is an early release, but there are a couple of bugs and
bits of functionality i required.
My perl is more than a little out of practice, so feel free to keep,
reject, etc, the code as you see fit.
If you do incorporate the changes, please mail me back so I can pick
up any new release...it will make my life easier.
BUG:
1. Installation: It failed because the build did not pick up on the
dependency to WWW::Mechanize
2. Working with Google.pm
- the $mech->dump_forms; was not commented out
- the number of results was ignoring the parameter and was hardwired
to 100
3. Working with Bing.pm
- it blew up in the case where there was no link found
ADDED Functionality
- I have written a Yahoo.pm module. It has a dependency on URI::Escape
- I added another flag to set the geo location for the google search.
This is different from the frontpage - the results change depending on
where google think you are from, so to see how well our site is doing
on google.com for a US customer, the gl flag has to be set. That way I
can check our performance on, google.co.uk, google.com for someone n
the uk, google.com for someone in the us, etc.
I have attached all the files I have changed.
I have also attached the front end perl script I will be using to pull
it together, just for reference.
Thanks again
Oliver
Message body is not shown because sender requested not to inline it.
Message body is not shown because sender requested not to inline it.
Message body is not shown because sender requested not to inline it.
Message body is not shown because sender requested not to inline it.
Message body is not shown because sender requested not to inline it.