Jackett: Tracker request - Il Corsaro Verde

[Jorman]

Always me!

Another great torrent tracker for my local language!

Il Corsaro Vede:
https://www.icv-crew.org

It's a private tracker without invite so free to register!

I already start at loking in it but there's always a kind of different problem!
The website is basically a forum with release, is quite good organized so each forum is one category.
I don't know if is better to make only search or I can map every forum to the relative category and extract the exact path, so when one search can search in the right forum

The other question/problem is the search itself, start here
https://www.icv-crew.org/forum/index.php?action=search
but end here
https://www.icv-crew.org/forum/index.php?action=search2

So I don't know how to "intercept" all the data "input filed" and forum selection, to build the search string, then I don't know how to "request" the search2 page, is a strange functionality that I never see out there, or I don't pay attention to it!

Any ideas?

[kaso17]

This one won't work due to limitations of the definition format.
It needs to be implemented as native (C#) indexer like other forum based trackers too (e.g. 7tor and rutracker)

[kaso17]

ok, it might actually work, the search request is a post request (method: post) but you can use standard get values too. E.g. https://www.icv-crew.org/forum/index.php?action=search2&search=simpsons&submit=Ricerca&advacned=0

[bonny1992]

Ok, after a way too much time I may have the time to get my head on this, if kaso is cool with this.

There are a couple of problems, btw.
There is a limitation of 60 seconds between every search, how can we surpass this?
Also, to get the magnet link (if I recall correctly, this is needed by Jackett, right?) there is a "Thank you" button on top of the post, can we click on it?

[Jorman]

thank you @bonny1992
I don't know how can you bypass the search problem, the results are in another page that is generated after the 1st search (if nothing has changed in between) and the results itself are a "list" of topics and inside every topic there's a list of files ... So all with cardigan definitions? I'm very noob so I'm curious

J

[kaso17]

As long as each topic contains only exactly one download link it should be possible to make this work as a cardigann definition.

[bonny1992]

As I could have seen, in every topic there is exactly one download link, but often is hidden behind a "thank you" mechanism, there is something I can refer to?

[garfield69]

@bonny1992 refer to wiki https://github.com/Jackett/Jackett/wiki/Definition-format#download.
Example in use: https://github.com/Jackett/Jackett/blob/44d14dc19c1007015c02bceecc5494dbd44a94f3/src/Jackett.Common/Definitions/downloadville.yml#L157-L166

[bonny1992]

@garfield69 thank you for pointing me to this.
I have one more question by the way.
Even though the links are often hidden behind the "thank you" mechanism, there is sometimes where instead the magnet is in plain view (there is still only one link per topic, as I could have seen). The "before" can be facultative?

[garfield69]

I suspect that on indexers written in yaml, the download before code block can not be set up as optional....if this is a requirement then the indexer would probably need to be written in c#.
@kaso17 can you confirm that what I've just said is correct? Thanks.

[bonny1992]

@garfield69
Oh, well, I may be wrong, but doing an useless POST request should not harm Jackett, even though it may bring some complications with the tracker after some empty "thank you" requests.

[kaso17]

If the thank you link can be generated from information available on the search result page implementing this can be with the current before mechanism. If it's only available from the details page then a small Cardigann extension would be necessary (I can take care of that).

[bonny1992]

I don't sincerely know because we need the msg number, the topic id and member id (all obtainable from search result page), but we also need the cookie for the connection to SMF.
Here's the CURL needed to do the request (which will return the text of the entire topic):
curl 'https://URL/forum/index.php?action=thank;msg=119461;member=15;topic=50311;refresh=1;ajax=1;xml=1' -H 'pragma: no-cache' -H 'cookie: displayCookieConsent=y; SMFCookie68=LONG_COOKIE; PHPSESSID=SESSID_PHP' -H 'accept-encoding: gzip, deflate, br' -H 'accept-language: it-IT,it;q=0.9,en-US;q=0.8,en;q=0.7,is;q=0.6' -H 'user-agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/68.0.3440.106 Safari/537.36' -H 'accept: */*' -H 'cache-control: no-cache' -H 'authority: URL' -H 'x-requested-with: XMLHttpRequest' -H 'referer: https://URL/forum/index.php?topic=50311.msg119461' --compressed

Also we have a search delay of 60 seconds, we can avoid that in some way?

[bonny1992]

Oh now I understood, maybe.
The cookie must be set from the "settings" page of the indexer, right?

EDIT: Also, sadly, I'm quite rusty with the writing of the definitions, is there a command to "test" the definition without having to reboot jackett everytime?

[Jorman]

Hi @bonny1992
Sorry that I can't help in anyway or maybe I can but I don't know c++

But when I have to test all, I make a copy of jackett, then I edit the configuration to change the port, then I remove all indexer except the one I'm testing ... but at the end I've to reload every time jackett :D

[Jorman]

Hi @garfield69
I just noticed the todo in progress label for this.
Can I ask you if you have any kind of eta for this? After tnt suspended activity, I think this's the most available for our country.
I only ask some information about, my only sorrow is that I can't help in any way

[garfield69]

waiting on a c# volunteer to code.

[Jorman]

Hi, any news or update with this tracker?

[garfield69]

as it happens, I am currently working on TheOdyssey #8054, a similar Italian tracker that is using the same web engine, SMF 2.0.*

If I can get my yaml indexer for this site to work, then I can also do the same for verde and I will post here that news.

But if I discover that the odyssey indexer cannot be written with the current implementation of Cardigann's yaml processor, then this too will need rewriting in C#, and this will be in the same boat as verde, which is currently pending a C# volunteer to devote time to coding.

When any such volunteer begins work on verde we will update this post. until then, there will be no further news.

[Jorman]

Thanks for your support!

[garfield69]

is this site filtering IP addresses only?
https://www.icv-crew.org/ returns me ERR_NAME_NOT_RESOLVED

[buzzqw]

the server is under maintenace. few days and will be back

[garfield69]

thanks for the update.

[buzzqw]

New url https://www.icv-crew.com/forum/

[YfheA3r]

I'm working on a C# indexer that will work both for this and for #8054 since they use a similar system and i have a few question:
-How can i return multiple magnets with the Download function? (both site have multiple magnets per thread and i can't split them in different ReleaseInfo since that would require sending the "thank you" to all the search result)
-Is there a way to cache a index with all the (Title, ThreadUrl) pair that you can find in the auto-generated threads lists? (i don't want to use SMF integrated search function since it work horribly and almost never find what you're looking for)

Examples of thread list:
https://www.icv-crew.com/forum/index.php?topic=45255.0
http://theodyssey.cc/index.php?topic=12.0

[garfield69]

@cadatoiva or @ngosang a C# question https://github.com/Jackett/Jackett/issues/3189#issuecomment-623123130

[ngosang]

I can't see the download links so my answer can be inaccurate.

-How can i return multiple magnets with the Download function? (both site have multiple magnets per thread and i can't split them in different ReleaseInfo since that would require sending the "thank you" to all the search result)

You can't. Each torrent/release can have 1 .torrent link or/and 1 magnet. In those cases we generate several ReleaseInfo (one per magnet). Usually each magnet has different qualities or uploader so the titles are different. We are not sending "thank you" or interacting with the website in other trackers. There are some feature requests related but I think we have enough keeping the search/download feature. Maybe we can help you when we see the code.

-Is there a way to cache a index with all the (Title, ThreadUrl) pair that you can find in the auto-generated threads lists? (i don't want to use SMF integrated search function since it work horribly and almost never find what you're looking for)

Yes, there is a cache but is not working well so I don't recommend it. Please, try to make as few requests as possible, even if the results are worse. Normally each search should make only 1 request to the site. We have thousands of users and we don't want to get banned or DDoS the site.

[cadatoiva]

There is a download function that will let you do a thank you for the link someone wants to download specifically within separate release-info links. Not sure how it works in C# since I've only seen it in the Cardigann system, but it's there underneath Cardigann for all C# indexers to use.

[Jorman]

Hi, any news/update on this tracker?
I think remain the best alternative to tntvillage

[ngosang]

No. We don't have a C# developer to work on this. PRs are welcome.