Database.xml doesn't allow the same search capabilities of website

[LightDestory]

I tried the new searching "feature" and it seems a bit unreliable and incomplete.

I tried to perform a search based on the keyboard: "brain" and I get different results:

• Website returns 18 entries
• database returns 12 entries

The search & filter logic of my script is the following: check for the existence of the keyword as substring of both title and description.

The database.xml doesn't contain all the entries because I checked and the missing 6 entries are not listed inside the file.

Moreover information such as leechers and seeders is not provided.

Is this a bug?

[ieee8023]

I tried to perform a search based on the keyboard: "brain" and I get different results:

It is possible that the website search includes extra entries because is searching other fields such as the file name to provide more results.

The database.xml doesn't contain all the entries because I checked and the missing 6 entries are not listed inside the file.

I just verified that from that search all 18 entries are present in the database.xml file.

Moreover information such as leechers and seeders is not provided.

That info is currently not provided but may be in the future. It is not something that is quick to add right now.

[LightDestory]

The database.xml doesn't contain all the entries because I checked and the missing 6 entries are not listed inside the file.

I just verified that from that search all 18 entries are present in the database.xml file.

I am sorry, yesterday I was tired and on my fast check I just thought that such entries were missing.

I found the cause: the missing 6 entries don't have the description field filled. Maybe your system just trim or don't allow very long description.

Right now with the provided database.xml is impossible to provide the same search experience of the website unfortunately.

Regards.

[ieee8023]

I found the cause: the missing 6 entries don't have the description field filled. Maybe your system just trim or don't allow very long description.

Can you tell me a specific infohash to look at?

[LightDestory]

b469546454e43f22804bedbb8fe9e43be2a8d95c 5e70f4f907aa02a4a6308b85cff9762f2e2a036b ce11d11d267500308ec39e2f747edd9721efe773 7197ed06604fcc7791d321afc229efe7c24dc472 eaa25083f5ef8b56ec203b0ba38c42842adaa47d 515e2989eedc853a8e256424de112f6f48f10d80

What is considered description? Abstract? Terms?

[ieee8023]

Those entries don't have descriptions. The pages just have "terms" which are not considered descriptions because they are often just some legal text but it is searched when using the web based search.

What is considered description?

The "abstract" key of the bibtex.

[LightDestory]

I don't know how is designed the back-end but I think that your search system checks Terms too because in 5e70f4f907aa02a4a6308b85cff9762f2e2a036b there is no brain keyword on torrent's name or files name, there is one "brain" inside terms.

To be clear, my issue is that by using the database.xml I am not able to provide the same results that your back-end provides. By "I am not able" I mean that I don't have enough elements to implement a more detailed filtering.

I am going to change the issue title as "Database.xml doesn't allow the same search capabilities of website", is it ok for you?

[ieee8023]

I am going to change the issue title as "Database.xml doesn't allow the same search capabilities of website", is it ok for you?

Yes, that is the case. But I don't think the differences are significant to prevent someone from finding what they are looking for.