Similarity Score Filter Not Working

I can double this.

Thanks,
David

I have similar issue. Seems the filter does not work. No matter I choose score>5 or score<5, the result is the same.

I have the same issue as well. Its a great module, it would be wort fixing it...

I wonder if any trick I need to know to use this module.
The module gives almost the same list (of similar entries) to any node. I accidentally have two duplicated nodes with exactly the same content and title (they are feed items generated by feeds module from an RSS feed), and they are not 'similar entry' to one another according to the module.
I generated some Chinese nodes, and later some English nodes; the similar entries to any node is a list of Chinese nodes.
The similarity scores are all 0, but if I set the rule as "score>5", those with a score of 0 still list.
All these happened after a successful Cron.
Besides those in the views, any other configuration?
---------------------
update: I used to map feed description to feed item "description" field. "Similar Entries" does not consider the "feed items" "description. Instead, it considers "body". That's why all the scores are 0. So the right way is to map feed description to feed item 'body', and after that I do get scores above 0.

But the 'feeds' problem does not belongs to this issue, and the score filter still does not work.

Same issue - would be awesome if this filter worked.

Same issue - have someone start any work on this? every code snippet or early patch would be great to take a look on this issue for trying work on this issue. thx from germany

I located this bug. Patch will come tomorrow!

It seems that the argument handler needs a check for the language setting for the content type.

Perhaps somebody could test the patch with a multi-language D7 installation with translateable content types because I have currently not tested this case.

The views filter doesn't seem to work here either. I use Drupal 7 and l18n. I modified the module (manually), cleared cache under performance tab, did and update and a cron run, but no effect.

Funnily enough I also have the original drupal 6 site still running that used the similar entries block. This allows me to look at the same content and see that the difference quite clearly between a working module and one that doesn't.

I also have D6 and D7 sites with the same content. D6 version works great, D7 version doesn't.

#8 works for me. Thanks!

I think #8 should be committed we tested it on 5 projects.

#8 also fixed the bug in several of my projects.

This looks like a duplicate of #2561577: Various fixes. The other patch included other stuff too, so I used it.

I think its not duplicate because without this patch similar dont find correct node scores for sites on Russian, Ukrainian and other non-latin languages.

Also I've found in views/similar_handler_argument_nid.inc:
$title = preg_replace('/[^a-z0-9 _-]+/i', '', $node->title);

This string cuts off all non-latin symbols so, for russian language it shoud be:
$title = preg_replace('/[^a-z0-9а-я _-]+/iu', '', $node->title);

So I think we should use other option to remove punctuation from title.

Is there a working solution for 7.x-2.0-beta6?

No, 7.x-2.0-beta6 doesn't work with Cyrillic articles.