incith:google :)

[De Kus]

Is it possible to skip subresults?
example:
http://www.google.de/search?q=incith
the script prints the subresult as the second result instead of the true second result.
btw. I am using v1.8.4.

[hello]

Thanks speechless on the update of my question can't wait until u get some time to play with it on adding that.

Thank madwoota on the update will be nice to see google 2.0 once you are done with it and ready to release it.

[speechles]

@De Kus,

Code: Select all

      # fetch the html
      set html [fetch_html $input 1]
      # dunno how efficient this is, but all sibling results are within <blockquote's>
      # so this might be all you need. who knows. i barely know tcl, but luv regsub's.. heh
      regsub -all -- {<blockquote.+?>(.+?)<.+?blockquote>} $html {} html

more of a quick pre-parsing hack than an actual fix within the google regexp/regsub's, but it _should_ work

@hello,
i'm improving some of my procs in an effort to clean them up to port them over to madwoota's 2.0 update (when it's released that is).. I'll link something soon when I finally get it working like I want it to..

[De Kus]

its a nasty hack, but I can confirm it fixing the issue.

[incith]

Regular searches seem to be broken tonight, they can be fixed via finding the # regular searches line in proc google (sorry it's not open right now and I'm posting this in a rush), and changing the regexp <p to regexp <div, and the regsub a few lines below it also from the <p to <div

Regards~

[madwoota]

EMERGENCY BUGFIX RELEASE! for v1.8 series - now at v1.8.5

- fixed main google results broke ! Google changed <p> -> <div>
- fixed weather
- fixed area code map results (eg: !g 90210 or !g beverly hills, ca)
- added new setting (subresults) to DEFAULT to NOT return secondary results - seems like blockquotes are gone speechless

Looks like there is a bunch more stuff thats broken too ... but I'll get to them soon(ish).

http://www.woota.net/~madwoota/incith-g ... ent.tar.gz

-madwoota

[v00j00]

I was wondering why my egg was acting up after I rebooted, I thought I broke it.

Thanks madwoota!

[De Kus]

the script missing URL character conversion...

!google 13:48 <+ultra> Hi everybody

leads to

Code: Select all

[12:50:12] #De_Kus# set errorInfo
Currently: Illegal characters in URL path
Currently:     while executing
Currently: "::http::geturl $query -timeout [expr 1000 * 10]"
Currently:     (procedure "fetch_html" line 50)
Currently:     invoked from within
Currently: "fetch_html $input 1"
Currently:     (procedure "google" line 16)
Currently:     invoked from within
Currently: "google $search"
Currently:     (procedure "send_output" line 39)
Currently:     invoked from within
Currently: "send_output "$input" "$chan" "$nick" "$uhand""
Currently:     (procedure "incith::google::public_message" line 16)
Currently:     invoked from within
Currently: "incith::google::public_message $_pubm1 $_pubm2 $_pubm3 $_pubm4 $_pubm5"

Edit:

...
running eggdrop v1.6.18:
...
OS: Linux 2.6.16-test-ovz
Tcl library: /usr/local/lib/tcl8.4
Tcl version: 8.4.13 (header version 8.4.13)

.tcl package require http 2.4
Tcl: 2.5.2

[madwoota]

the script missing URL character conversion...

!google 13:48 <+ultra> Hi everybody

leads to

Code: Select all

[12:50:12] #De_Kus# set errorInfo
Currently: Illegal characters in URL path
Currently:     while executing 

For me, it leads to this:

Code: Select all

<madwoota> !google 13:48 <+ultra> Hi everybody
<|EggDrop|> New TyTn / M3100 Owners buying memory ca @ http://forum.xda-developers.com/archive/index.php/t-274008.html | Seeking-delays-from @ http://www.fibul.com/group-1339-13.html | Ultra Gypsy's topics - tribe.net @ http://ultragypsy.tribe.net/threads/rss

Although I do agree with your point, I should probably be sanitising the input string before it gets submitted to the url fetcher.

[h44i]

I'm running the latest Eggdrop on Debian Etch and I just installed the latest version of this script. It works flawless but there's one bug I guess. It only outputs 1 search result. The TCL has 3 results configured though.

Any ideas? Did not change anything to the TCL, it's like it's provided a few posts above.

[incith]

Seems to be working good for us still.

Download the latest version @ http://xrl.us/incithgoogle

[Cold Shady]

I'm using incith:google v1.8.6 and it seems that the url's don't come through properly.

Code: Select all

[14:21:57] <Cold_Shady> !google woot
[14:21:57] <Gammle> Woot : One Day, One Deal @ http://www.woot.com/%20class=l
[14:21:59] <Gammle> Woot : One Day, One Deal @ http://www.woot.com/Default.aspx%20class=l
[14:22:00] <Gammle> Woot (retailer) - Wikipedia, the free @ http://en.wikipedia.org/wiki/Woot.com%20class=l

I'm trying to fix it, but haven't got any luck so far.
Any suggestions?

[incith]

Yes, I fixed this earlier when a user mentioned it to me.. go down to the line:

Code: Select all

        # regular search

about line 445. Look for..

Code: Select all

regexp -- {<div.+?class=g.*?<a.+?href=(.+?)</a>} $html - link desc

about 6 or 7 lines down, above if {[info exists link] == 1}, and change the regexp to this:

Code: Select all

regexp -- {<div.+?class=g.*?<a.+?href=(.+?)\s+?.*?>(.+?)</a>} $html - link desc

Regards,

[Cold Shady]

Thank you!

[incith]

Download: incith-google.tcl

Temporary fixed version.