OLD: Bit Che 2.0 - Latest Version -- Build 60 (April 11, 2013)

[loninappleton]

Chip,

  here is a sample of the error text which I get only on certain torrent requests.  Some information ****'ed out for privacy.

A sample of a long torrent name appearing in Tixati from a BitChe request is:

[verbatim]

C:\DOCUME~\***~1.***\APPLIC~1\Convivea\Bit_Che\2\scripts\temp\temp_7494***.torrent

A path name in my system is being applied to the named torrent requested.  The ~ which indicate truncations are part of the "named" .torrent

[chip!]

Exoskeletor,  thanks, i see what you are talking about and can fix that. I also made it so that the filter will stay active for new searches until you clear it, but i'm not sure if i like it defaulting that way, or being cleared on every new search (and then needing to type in the filter again to apply)..

loninappleton,  can you take some screen shots? I'm confused now as to what is going on.  also, can you reproduce this everytime with certain 'results' ?  if so, please PM me how to reproduce that. for example, tell me the site searched and the result clicked.

[AnthonyB]

Using 2.0 Build 35 and have updated scripts.

I'm seeing a time out with any torrent I try to open from BitSnoop.com
In the torrent details window I see the following in the status bar:

Code: [Select]

Connecting to BitSnoop.com
Connecting to zoink.it
Error occured with: http://zoink.it/torrent/<<torrent data removed for this post>>
Is there an issue with the Bitsnoop.com script?

[chip!]

Is there an issue with the Bitsnoop.com script?

the script itself is fine... but we were forcing the use of the zoink mirror as it was more reliable in the past. right now, zoink appears to be down so the script has been modified to just use the default source/mirror.

[AnthonyB]

Is there an issue with the Bitsnoop.com script?

the script itself is fine... but we were forcing the use of the zoink mirror as it was more reliable in the past. right now, zoink appears to be down so the script has been modified to just use the default source/mirror.

Thanks Chip!

[CuF]

Found a possible bug.

When using $seeds=array(@q,4) on a site, if there are no seeds the array has: - - -  (which appears as --- on the results page)

The debugger properly shows that 0 seeds are found, but the main Bit Che results window returns the number found on the previous loop.
The same happens for 0 leeches.

For now it's easy enough to do a replace of - - - with 0 in the script, but it should probably be fixed.


Also, is Bit Che+ suppose to be able to handle .torrent links that redirect?  Having an issue with that as well.

The link is properly found by BC, but clicking download gives an error.  Copying the URL to the clipboard and pasting in the browser will download the .torrent file.  Not sure if that's a bug, or what the proper handling should be in the script.

I found both these issues on my AsianDVDClub script.

[chip!]

Found a possible bug.

When using $seeds=array(@q,4) on a site, if there are no seeds the array has: - - -  (which appears as --- on the results page)

The debugger properly shows that 0 seeds are found, but the main Bit Che results window returns the number found on the previous loop.
The same happens for 0 leeches.

For now it's easy enough to do a replace of - - - with 0 in the script, but it should probably be fixed.


Also, is Bit Che+ suppose to be able to handle .torrent links that redirect?  Having an issue with that as well.

The link is properly found by BC, but clicking download gives an error.  Copying the URL to the clipboard and pasting in the browser will download the .torrent file.  Not sure if that's a bug, or what the proper handling should be in the script.

I found both these issues on my AsianDVDClub script.

thanks CuF...  I was able to fix the 0 seeds/leech issue without using your posted script.  

i would like to take a look at the redirect issue you are having too but I will need some more info.   I did some google searching and found what the download URL is typically.. and unfortunately, that wasnt any help since it just redirects to the login page...

SO... I'll need access to the site in order to test with Bit Che to see if there is a redirect bug or not.    we can do 2 things.. one, you can email an invite to me and i'll setup an account to test with.. or two, you can change your pw to something temporary and PM me so I can login and test.. (then you can change your pw back). either way, I actually have to debug the actual http connection... if you dont want to do either of those options... last resort would be to send me a HTTP log while you are using BC to try and download a .torrent... you can create an HTTP log using something like:  http://www.nirsoft.net/utils/http_network_sniffer.html

EDIT: updated

[chip!]

CuF:  new idea... lets try using Bit Che to tell me the Headers response (what I think I need mostly) when you request the .torrent:

1.  attached is a slightly modified version of your script. put it in your Bit Che script directory.
2.  do a search in Bit Che with the script so you have some results
3.  open the Script Debugger
4.  then double click one of your results (which then triggers the 2nd code section)
5.  in the Script Debugger, hit the 'Headers' button
6.  copy and paste everything in your debugger to me.. either post it here, PM me, or email me.

hopefully that gives me enough insight into the problem

[chip!]

Posted Build 36 - January 16, 2013
-fixed: critical internal param handling and cookie bug
-fixed: script engine [IF THEN ELSEIF and ELSE]
-fixed: malformed 'seeders' number reporting (thanks CuF)
-improved: stability tweaks
-improved: language translations support and form sizes (thanks Exo)
-added: Debugger [IF THEN ELSEIF and ELSE] support (able to step through and back, plus improved display)
-added: right click on Results columns to select/unselect columns to be hidden (thanks sunstrip)
-added: new option to add Script Packs via bc://scriptpack?name=Name&update=URL for one-click installation

attached to this post:  http://convivea.com/forums/index.php?topic=2131.msg19117#msg19117

[CuF]

Sorry, to be slow to get back to you, I was stuck with jury duty the last few days.

I'll test the new version and see if that fixes what's been ailing my scripts as soon as I can.
Does the malformed seeder fix also effect malformed leeches?  It occurred for both.

[chip!]

Sorry, to be slow to get back to you, I was stuck with jury duty the last few days.

I'll test the new version and see if that fixes what's been ailing my scripts as soon as I can.
Does the malformed seeder fix also effect malformed leeches?  It occurred for both.

aye! not fun... but no worries. that would be great to test the new version because the param handling/cookie bug could have affected more than it seemed. i'm wondering if that was the source of many problems with sites that relied on cookies (all login sites)... but if there is still a redirect bug, i need to squash it.   i have an entire function that "validates" the URL redirection, so its perfectly possible I've got a bug there too!

also, yes, the fix applied for both seeds and leeches (forces the value to be numeric)

thanks!

[CuF]

Bug in Build 36:

Under some conditions if you use t=array(@q,1) and that array is blank, the variable is deleted and doesn't appear the next time =array(@q,n) is used.
This is causing the next loop's results to be removed.

For example:
Using my BakaBT script (attached) that works with Build 35 but not 36 and searching "Ghost Shell Innocence", there should be 4 results.

Loop 1: Finds 1080p version
Loop 2: Finds only the text "Alternate Versions:" and loops
Loop 3: t=array(@q,0) is checked but empty (removing the variable 't' in the debugger window), t=array(@q,1) does not get the data from @q(1)

This works properly in Build 35.

Found in Build 35 and 36:
If you click in Bit Che's results area, it will occasionally anchor a select region box.  Can't be more specific, sorry.

[Exoskeletor]

for me it still opens 2 links when i click on forums button or the open source page button

[chip!]

Bug in Build 36:

Under some conditions if you use t=array(@q,1) and that array is blank, the variable is deleted and doesn't appear the next time =array(@q,n) is used.
This is causing the next loop's results to be removed.

For example:
Using my BakaBT script (attached) that works with Build 35 but not 36 and searching "Ghost Shell Innocence", there should be 4 results.

Loop 1: Finds 1080p version
Loop 2: Finds only the text "Alternate Versions:" and loops
Loop 3: t=array(@q,0) is checked but empty (removing the variable 't' in the debugger window), t=array(@q,1) does not get the data from @q(1)

This works properly in Build 35.

Found in Build 35 and 36:
If you click in Bit Che's results area, it will occasionally anchor a select region box.  Can't be more specific, sorry.

thanks CuF.. was able to track this down and fix it! :/


Exoskeletor.. what browser are you using?  the code only executes ShellExec API once, and I dont think anyone else has mentioned this problem, so I'm not sure where to begin..

[loninappleton]

Unrelated to the present discussion above but this seems like the only place to ask:

Will Mega be added to the user- updated scripts like KAT and so forth?  I do not know how to add indexes but I read this thread regularly.