kaminari / kaminari Goto Github PK

I have problems using paginate with an AJAX call. I want to replace the pagination div like this:

$('.pagination').replaceWith('<%= escape_javascript(paginate(collection, :params => paginate_params).to_s) %>');

I tried multiple versions, but I can't get it working. I get multiple JS errors when I want to update the view.

currently it is generated only from params.
there should be option to pass additional data to url params without global params modification

Hi;

Inside of my index action I have something like
@shipments = Shipment.order('date desc').page(params[:page])

If page is null, not passing the parameter or page == 1 then I got results and my @shipments variable get loaded, but if page >= 2 I got nothing, actually looking at Rails log no query is being issue.

My shipments table have hundred of records.

If I try the exact same line of code in Rails console it works, it issue the query

I'm using version 0.10.4 of kaminari.

Do you have any suggest on how to fix this?

Is there a way to get the same view helper output from will_paginate's page_entries_info helper in kaminari? I've looked at all the documentation and don't see this.

I'm sure I can generate it by passing the collection i specify from <%= paginate @collection %> as a locals hash, then deriving it, but it would be nice to have this built in if its somewhere.

Here's the will_paginate readme:
<%= page_entries_info @posts %> #-> Displaying posts 6 - 10 of 26 in total

Has anyone figured this out? If not, I suppose I can write it and submit it back :)

per_page などのように、設定で使用している変数については partial で使用できると表現の幅が広がるので色々と便利だと思います。

It would be great if the :page parameter could be localizable so links could be generated like '?pagina=1' instead of 'page=1'

hi there,

i want kaminari to use the following route :

match "/:category(/page/:page)" => "viewer#index", :as => "category_index"

instead of going to the old route

/viewer/index?category=fiction&page=2

would i have to modiify the kaminari views? cant find any posts or comments from any1 with relation to this

0.4.0 の view で paginate を繰り返し呼ぶと呼んだ数だけ pagination のリンクが追加されてしまいます。0.2.1では発生しなかったので 0.3.0 以降だと思います。

ページ全体をhamlで書いている場合に、= paginate @entries でpaginationを出力しようとした場合に、「undefined method `safe_concat' for nil:NilClass」というエラーになりました。
（0.9.9までは大丈夫だった書き方）

rails generate kaminari:views default -e haml　を実行しapp/views/kaminari以下にファイルを作ると、hamlの場合でもpaginationが出力できるようになり、特に問題はありません。

0.9.12で「Moved the whole pagination logic to the paginator partial」とあるので、そのためかもしれません。

ドキュメントのどこかに「hamlを使う場合は最初に rails generate kaminari:views default -e haml を実行しましょう」のようなことが書かれているほうが分かりやすいかもしれません。

以下、試した際にviewに出たエラーです。
バニラな環境で再現できます。

undefined method `safe_concat' for nil:NilClass

Extracted source (around line #9):

6: remote: data-remote
7: paginator: the paginator that renders the pagination tags inside
8: -%>
9: <%= paginator.render do -%>
10:

I've noticed kaminari's helper adds the page parameter to every link except to the link for the first page.

That usually would work fine, but I've got a use case where it doesn't work as expected.

In an application I'm working on there's an admin area where, if no page param is provided, the last visited page by the user is loaded.

The reasoning for this follows. In a traditional situation:

• admin makes it to page 17 of posts, where there are two posts to modify.
• admin modifies the first post and saves it.
• admin goes back to the posts index.
• admin has to make all the way to page 17 again to modify the second one.

My solution (which has its drawbacks, of course) is saving page 17 as last page visited, and directing the admin there when there's no page parameter.

Everything would work (and I can't think of situations that would be broken) if the first page link had the page param as well. Here's a small patch:

https://gist.github.com/846609

Unfortunately I've tried to provide a test case, but failed miserably with the template stubbing.

Thanks.

Hi

Nice job with this gem. However, adding it to my Gemfile on a Rails3 app make my development environment leak 3MB of memory per request (using Mongrel).

Thibaut

Say I have a route:

match ':year(/:month(/:day))' => 'posts#index', :as => 'posts_index'

Kaminari will generate page links like so: /2011/?page=3&month=1&day=21

undefined method `offset' for #Class:0x1046b4ae0
rails 3.0.0

trace:

activerecord (3.0.0) lib/active_record/base.rb:1016:in method_missing' kaminari (0.9.2) lib/kaminari/active_record.rb:18:ininherited'
activerecord (3.0.0) lib/active_record/named_scope.rb:107:in call' activerecord (3.0.0) lib/active_record/named_scope.rb:107:inpage'
app/controllers/opportunities_controller.rb:11:in index' actionpack (3.0.0) lib/action_controller/metal/implicit_render.rb:4:insend_action'
actionpack (3.0.0) lib/action_controller/metal/implicit_render.rb:4:in send_action' actionpack (3.0.0) lib/abstract_controller/base.rb:150:inprocess_action'
actionpack (3.0.0) lib/action_controller/metal/rendering.rb:11:in process_action' actionpack (3.0.0) lib/abstract_controller/callbacks.rb:18:inprocess_action'
activesupport (3.0.0) lib/active_support/callbacks.rb:460:in _run__1183185049__process_action__199225275__callbacks' activesupport (3.0.0) lib/active_support/callbacks.rb:409:insend'
activesupport (3.0.0) lib/active_support/callbacks.rb:409:in _run_process_action_callbacks' activesupport (3.0.0) lib/active_support/callbacks.rb:93:insend'
activesupport (3.0.0) lib/active_support/callbacks.rb:93:in run_callbacks' actionpack (3.0.0) lib/abstract_controller/callbacks.rb:17:inprocess_action'
actionpack (3.0.0) lib/action_controller/metal/instrumentation.rb:30:in process_action' activesupport (3.0.0) lib/active_support/notifications.rb:52:ininstrument'
activesupport (3.0.0) lib/active_support/notifications/instrumenter.rb:21:in instrument' activesupport (3.0.0) lib/active_support/notifications.rb:52:ininstrument'
actionpack (3.0.0) lib/action_controller/metal/instrumentation.rb:29:in process_action' actionpack (3.0.0) lib/action_controller/metal/rescue.rb:17:inprocess_action'
newrelic_rpm (2.13.4) lib/new_relic/agent/instrumentation/rails3/action_controller.rb:34:in process_action' newrelic_rpm (2.13.4) lib/new_relic/agent/instrumentation/controller_instrumentation.rb:252:inperform_action_with_newrelic_trace'
newrelic_rpm (2.13.4) lib/new_relic/agent/method_tracer.rb:141:in trace_execution_scoped' newrelic_rpm (2.13.4) lib/new_relic/agent/instrumentation/controller_instrumentation.rb:247:inperform_action_with_newrelic_trace'
newrelic_rpm (2.13.4) lib/new_relic/agent/instrumentation/rails3/action_controller.rb:33:in process_action' actionpack (3.0.0) lib/abstract_controller/base.rb:119:inprocess'
actionpack (3.0.0) lib/abstract_controller/rendering.rb:40:in process' actionpack (3.0.0) lib/action_controller/metal.rb:133:indispatch'
actionpack (3.0.0) lib/action_controller/metal/rack_delegation.rb:14:in dispatch' actionpack (3.0.0) lib/action_controller/metal.rb:173:inaction'
actionpack (3.0.0) lib/action_dispatch/routing/route_set.rb:62:in call' actionpack (3.0.0) lib/action_dispatch/routing/route_set.rb:62:indispatch'
actionpack (3.0.0) lib/action_dispatch/routing/route_set.rb:27:in call' rack-mount (0.6.13) lib/rack/mount/route_set.rb:148:incall'
rack-mount (0.6.13) lib/rack/mount/code_generation.rb:93:in recognize' rack-mount (0.6.13) lib/rack/mount/code_generation.rb:68:inoptimized_each'
rack-mount (0.6.13) lib/rack/mount/code_generation.rb:92:in recognize' rack-mount (0.6.13) lib/rack/mount/route_set.rb:139:incall'
actionpack (3.0.0) lib/action_dispatch/routing/route_set.rb:492:in call' oa-core (0.1.6) lib/omniauth/strategy.rb:50:incall_app!'
oa-core (0.1.6) lib/omniauth/strategy.rb:32:in call!' oa-core (0.1.6) lib/omniauth/strategy.rb:19:incall'
oa-core (0.1.6) lib/omniauth/builder.rb:30:in call' newrelic_rpm (2.13.4) lib/new_relic/rack/developer_mode.rb:20:incall'
warden (0.10.7) lib/warden/manager.rb:35:in call' warden (0.10.7) lib/warden/manager.rb:34:incatch'
warden (0.10.7) lib/warden/manager.rb:34:in call' actionpack (3.0.0) lib/action_dispatch/middleware/best_standards_support.rb:17:incall'
actionpack (3.0.0) lib/action_dispatch/middleware/head.rb:14:in call' rack (1.2.1) lib/rack/methodoverride.rb:24:incall'
actionpack (3.0.0) lib/action_dispatch/middleware/params_parser.rb:21:in call' actionpack (3.0.0) lib/action_dispatch/middleware/flash.rb:182:incall'
actionpack (3.0.0) lib/action_dispatch/middleware/session/abstract_store.rb:149:in call' actionpack (3.0.0) lib/action_dispatch/middleware/cookies.rb:287:incall'
activerecord (3.0.0) lib/active_record/query_cache.rb:32:in call' activerecord (3.0.0) lib/active_record/connection_adapters/abstract/query_cache.rb:28:incache'
activerecord (3.0.0) lib/active_record/query_cache.rb:12:in cache' activerecord (3.0.0) lib/active_record/query_cache.rb:31:incall'
activerecord (3.0.0) lib/active_record/connection_adapters/abstract/connection_pool.rb:355:in call' actionpack (3.0.0) lib/action_dispatch/middleware/callbacks.rb:46:incall'
activesupport (3.0.0) lib/active_support/callbacks.rb:415:in _run_call_callbacks' actionpack (3.0.0) lib/action_dispatch/middleware/callbacks.rb:44:incall'
rack (1.2.1) lib/rack/sendfile.rb:107:in call' actionpack (3.0.0) lib/action_dispatch/middleware/remote_ip.rb:48:incall'
actionpack (3.0.0) lib/action_dispatch/middleware/show_exceptions.rb:46:in call' railties (3.0.0) lib/rails/rack/logger.rb:13:incall'
rack (1.2.1) lib/rack/runtime.rb:17:in call' activesupport (3.0.0) lib/active_support/cache/strategy/local_cache.rb:72:incall'
rack (1.2.1) lib/rack/lock.rb:11:in call' rack (1.2.1) lib/rack/lock.rb:11:insynchronize'
rack (1.2.1) lib/rack/lock.rb:11:in call' actionpack (3.0.0) lib/action_dispatch/middleware/static.rb:30:incall'
railties (3.0.0) lib/rails/application.rb:168:in call' railties (3.0.0) lib/rails/application.rb:77:insend'
railties (3.0.0) lib/rails/application.rb:77:in method_missing' railties (3.0.0) lib/rails/rack/log_tailer.rb:14:incall'
rack (1.2.1) lib/rack/content_length.rb:13:in call' rack (1.2.1) lib/rack/handler/webrick.rb:52:inservice'
/System/Library/Frameworks/Ruby.framework/Versions/1.8/usr/lib/ruby/1.8/webrick/httpserver.rb:104:in service' /System/Library/Frameworks/Ruby.framework/Versions/1.8/usr/lib/ruby/1.8/webrick/httpserver.rb:65:inrun'
/System/Library/Frameworks/Ruby.framework/Versions/1.8/usr/lib/ruby/1.8/webrick/server.rb:173:in start_thread' /System/Library/Frameworks/Ruby.framework/Versions/1.8/usr/lib/ruby/1.8/webrick/server.rb:162:instart'
/System/Library/Frameworks/Ruby.framework/Versions/1.8/usr/lib/ruby/1.8/webrick/server.rb:162:in start_thread' /System/Library/Frameworks/Ruby.framework/Versions/1.8/usr/lib/ruby/1.8/webrick/server.rb:95:instart'
/System/Library/Frameworks/Ruby.framework/Versions/1.8/usr/lib/ruby/1.8/webrick/server.rb:92:in each' /System/Library/Frameworks/Ruby.framework/Versions/1.8/usr/lib/ruby/1.8/webrick/server.rb:92:instart'
/System/Library/Frameworks/Ruby.framework/Versions/1.8/usr/lib/ruby/1.8/webrick/server.rb:23:in start' /System/Library/Frameworks/Ruby.framework/Versions/1.8/usr/lib/ruby/1.8/webrick/server.rb:82:instart'
rack (1.2.1) lib/rack/handler/webrick.rb:13:in run' rack (1.2.1) lib/rack/server.rb:213:instart'
railties (3.0.0) lib/rails/commands/server.rb:65:in start' railties (3.0.0) lib/rails/commands.rb:30 railties (3.0.0) lib/rails/commands.rb:27:intap'
railties (3.0.0) lib/rails/commands.rb:27
script/rails:6:in `require'
script/rails:6

Is it possible that this become easily extensible with other ORM like mongoid. I saw there is the active_record.rb in Railties. Can we do some configuration and use other file like mongoid.rb to change the data store and still use this paginator.

Hi! Can someone explain why it's better than will_paginate? Some feature?
I'm using will_paginate and mongoid and it's working well right know. Tks!

http://freelancing-god.github.com/ts/en/searching.html

here is example pagination thinking_sphinx results with will_paginate gem

how i can do that with kaminari?

tnx

Can i specify renderer some place, for example, i have to use kaminari in my admin and site parts, how can in handle different themes?

An awesome gem.. but i want to know if theres any way to implement the endless page feature, that twitter and Facebook use?! As I hear offset query is a bit slower.. http://www.slideshare.net/Eweaver/efficient-pagination-using-mysql

I tested using this gem over will_paginate in some of my views/actions. However, when I compare the before and after render times of the partial including the paginate helper I notice a massive increase in render time. I have here a few examples:
with will_paginate:
135ms
212ms
with kaminari:
509ms
583ms
560ms
406ms

This was done in development mode. I suspect it has to do with the huge ammount of partial templates you are using. Im using rails 3.0.3 and ruby 1.9.2.

Have tested this against version 0.10.2 and Mongoid 2.0.0.rc.7

Given the following model:

class Foo
  include Mongoid::Document
  field :value, :type => Integer
  scope :below, lambda {|v| where( :value.lt => v )}
  scope :above, lambda {|v| where( :value.gt => v )}
end

a call to the :page scope will eliminate any previous criteria made in the call chain prior to the call. For example:

Foo.above( 5 ).below( 10 )
 => #<Mongoid::Criteria
  selector: {:value=>{"$gt"=>5, "$lt"=>10}},
  options:  {},
  class:    Foo,
  embedded: false>
Foo.above( 5 ).below( 10 ).page( 1 )
 => #<Mongoid::Criteria
  selector: {},
  options:  {:limit=>25, :skip=>0},
  class:    Foo,
  embedded: false>

However, calling :page first seems to work alright:

Foo.page( 1 ).below( 5 )
 => #<Mongoid::Criteria
  selector: {:value=>{"$lt"=>5}},
  options:  {:limit=>25, :skip=>0},
  class:    Foo,
  embedded: false>

The definition of the :page scope (in models/mongoid_extension.rb) seems to be correct enough; adding in a limit or offset criteria on a method call chain does not seem to be an issue:

Foo.below( 5 ).limit( 10 ).offset( 5 )
 => #<Mongoid::Criteria
  selector: {:value=>{"$lt"=>5}},
  options:  {:limit=>10, :skip=>5},
  class:    Foo,
  embedded: false>

To reiterate: this seems to completely wipe out the previous criteria stack made in the method call chain. I used scopes with :where, above, but the issue occurs with other criteria operations as well. For instance, :asc/:desc, used for ordering by specified fields, results in the same issue.

With the default settings, visiting page 8 will show "1 2 … 4 5 6 7 8 9" and so on.

While I see how this is consistent with the window sizes (the 3 falls right between the outer window and the inner window), it's not good end-user experience.

I suggest that the truncation should only appear if it covers more than one page.

Since kaminari will append an offset and limit to each query, when trying to use the default .count method AREL will try to do a count with the limit and offset, which doesn't work. Likewise using something like .any? I ran into this bug because I have some logic that checks to see if any records exist before displaying them, otherwise I show a "There is no data" placeholder message. I was doing the check with data.any? which doesn't work after the first page.

I see that you added a total_count method which specifically excludes limit and offset, and I can use that in my code, but I wouldn't have known without digging through the code. Is it possible to override count to do the same thing, or does that have the potential to break too may things?

On a table I have of users (total 3). I set the per_page to 2 so I can test out pagination. The first two users get displayed but when I go to the next page, the laster user doesn't show up. Looking in the SQL code, I see you are doing the following:

SQL (0.3ms) SELECT COUNT() FROM users WHERE users.active = 1 LIMIT 2 OFFSET 0
User Load (0.2ms) SELECT users. FROM users WHERE users.active = 1 ORDER BY id ASC LIMIT 2 OFFSET 0

This works to produce the first two users. When I click on the 2n'd page, no users get loaded. The SQL code is:

SQL (0.3ms) SELECT COUNT(*) FROM users WHERE users.active = 1 LIMIT 2 OFFSET 2

When I test out this SQ code myslef, it shows that:

SQL (0.3ms) SELECT COUNT(*) FROM users WHERE users.active = 1 LIMIT 2 OFFSET 0

it returned 3 (I would think it would return 2)

SQL (0.3ms) SELECT COUNT(*) FROM users WHERE users.active = 1 LIMIT 2 OFFSET 2

returns 0 (I would think it would return 1)

I'm using MySql 5.1.51 on MacBook Pro running Rail 3.0.4.

My User controller is issuing this command:

@users = User.order('id DESC').page(params[:page]).per(2)

I'm not sure what is wrong with the SQL code you are generating but it isn't working right for me. Am I doing something wrong?

I was looking into some more details of will_paginate's behavior related to :outer_window, and it occurs to me that there's a more basic problem with the current versions of both will_paginate and kaminari: neither seems to properly support "first page" and "last page" links in the default configuration.

Will_paginate-style outer windows are in fact uncommon in the wild, and they only really exist to cover the lack of a proper "first" and "last". In this light, will_paginate's definition of the outer windows as "extra pages around the first and last pages" is more understandable. But, it's also an opportunity for kaminari to do things right by reflecting the predominant styles used on the web in its defaults.

Take a look at this nice set of example paginators from Smashing Magazine (via the will_paginate docs). Although the specific elements that are present varies, the order of presentation is invariant:

[first] [previous] 1 2 ... 5 6 [7] 8 9 ... 12 13 [next] [last]

Of particular note, a "first page" link, if present, should always precede the "previous page" link, and vice-versa for last/next. The outer windows cannot properly substitute for these links because by definition they appear interior to "previous" and "next". Furthermore, "first" and "last" links should appear in addition to the page numbers of the first and last pages if you are using an outer window, as shown above. (If you were on page 12, "13", "next", and "last" might all point to the same page; that's perfectly fine.)

A complete paginator specification requires answers to all of the following questions:

1. Should a "first page" link be displayed?
2. If so, should it be omitted when you are on the first page (vs. displayed but disabled)?
3. Should a "previous page" link be displayed?
4. If so, should it be omitted when you are on the first page (vs. displayed but disabled)?
5. Should the first N pages always be displayed? How many?
6. How many pages surrounding the current page should be displayed?
7. Should the last N pages always be displayed? How many?
8. Should a "next page" link be displayed?
9. If so, should it be omitted when you are on the last page (vs. displayed but disabled)?
10. Should a "last page" link be displayed?
11. If so, should it be omitted when you are on the last page (vs. displayed but disabled)?

To cover the full range of options, I propose the following modifications to paginate():

• :window, :outer_window, :left, :right would be interpreted similarly to how they are now, except the supplied N would be the number of pages you actually get. 0 means disable for the outer windows.
• Outer windows would be disabled by default in favor of first/last links.
• Currently, "next" and "previous" links are always generated and their labels come from the translation database. I suggest making :next and :previous params to paginate(), with the same default values; if either is nil, don't generate the corresponding span.
• Add :first and :last params to paginate() with suitable (present) defaults. If set to nil, these suppress the first and last links.
• Currently, "next" and "previous" are rendered as empty spans (no link, no text) when they point to the current page. This is hard to style and inconsistent. Better: include a "disabled" or "inactive" class when appropriate, and always include the label text (but no link tag when current page). Users that want these tags to disappear completely when inapplicable can style them to display: none. (Similarly for "first" and "last".)
• The double angle quotes currently used in the default labels for previous and next are potentially confusing because they bind more naturally to "first" and "last" than to "previous" and "next". (Consider a standard set of video or audio transport controls: << < stop > >>) Better: "Prev" and "Next" or "< Prev" and "Next >"

Under this regime, the defaults would be:

:window       => 3 or 4 (currently 4)
:outer_window => 0 unless :first == nil or :last == nil; in that case, 2
:next         => raw(t 'views.pagination.next'), with default translation "Next >"
:previous     => raw(t 'views.pagination.previous'), with default translation "< Prev"
:first        => raw(t 'views.pagination.first'), with default translation "<< First"
:last         => raw(t 'views.pagination.last'), with default translation "Last >>"

Some examples:

= paginate @things
[<< First] [< Prev] ... 6 7 8 [9] 10 11 12 ... [Next >] [Last >>]

= paginate @things :first => nil, :last => nil
[< Prev] 1 2 ... 6 7 8 [9] 10 11 12 ... 16 17 [Next >]

= paginate @things, :previous => nil, :next => nil
[<< First] ... 6 7 8 [9] 10 11 12 ... [Last >>]

= paginate @things, :first = nil, :left => 2, :window => 2
[< Prev] 1 2 ... 7 8 [9] 10 11 ... [Next >] [Last >>]

= paginate @things, :first => "First Page", :previous => "Previous", :next = "Next", :last => "Last Page"
 [First Page] [Prev] ... 6 7 8 [9] 10 11 12 ... [Next] [Last Page]

Hi, I'm trying to use kaminari in a project that uses jquery-rails.
After installing the gem and adding the appropriate code to the controller and index.html.erb
I get the following error:
Missing partial exercises/exercise with {:handlers=>[:erb, :rjs, :builder, :rhtml, :rxml], :formats=>[:html], :locale=>[:en, :en]}

Any ideas?

Regards,

Jens

I have following class in my project:

class User < AR::Base
  ...
  class Admin < User
  end

  class God < User
  end
  ...
end

Before creating a new user, you can selected the preferred type via a select-box. This select is filled by collecting all subclasses of User (User.descendants.collect(&:name)). Everything works fine unless kaminari is installed as a gem. After installing and using it, the descendants method always returns an empty array.

Entry.accepted.size # => 4
Entry.accepted.page(1).per(2).size # => 4

The returned array only has 2 elements as you would except, but size reports 4.

This may not be a kaminari issue, since you are just delegating to ActiveRecord scopes.

The default paginator has separate partials for the current page, the first page, and the last page. However, these are not mutually exclusive sets. To facilitate styling, the paginator should emit page spans with both the current and first/last classes when this is appropriate.

Perhaps there should just be one page partial: include "first" class if page.first?, include "last" class if page.last?, include "current" class if page.current?, and link the page number if not page.current?.

I think there should be a way to set the default_per_page nicer than now. The constant DEFAULT_PER_PAGE can be set, but in a Rails application, there is no way to set it before the gem loads, so you're always overriding it, resulting in a warning.

In Rails 3.0.5, any subclassed model that is loaded, ever, will produce the following warning in the Rails server log:

Creating scope :page. Overwriting existing method <model_name>.page.

I believe a proper fix here would be to have Kaminari make sure that it has not created the page (and others?) scope before doing so. This way, when a subclass initializes it would not recreate/override the scope that was already defined on the parent class. It's not necessary and it produces this warning message (see: https://rails.lighthouseapp.com/projects/8994/tickets/4083-patch-named-scopes-should-be-allowed-to-overwrite-an-existing-method).

If you'd like to see a sample app where this happens I can upload one to github.

Thanks!

When you specify :outer_window => 1 (as in the default), you actually get two pages in each outer window, e.g.:

1 2 ... <more pages> ... 9 10

More generally, you always get one more page than you ask for. This may be done for will_paginate compatibility, but it seems a little strange and perhaps a case in which reasonable behavior would be more valuable than compatibility.

Additionally, it would be helpful to be able to disable the outer windows entirely by setting :outer_window to 0. Currently, this gives you one-page outer windows.

Rails 3.0.4, slim, ruby 1.9.2, osx

== pagination @users

render only

version 0.9.10 work fine

Great work! Thanks so much for providing a rails3-aware pagination solution.

currently having trouble in 0.10.2, as follows...

controller-code:
@Clients = Client.page(params[:page]).per(50)

works ok, as long as page is nil or 1 or 0

however, as soon as page is > 1, @Clients will be empty unless i do this:

puts @Clients

so if i don't do this, view will render empty page although there are results for the query.
Maybe an issue with Arel lazy loading. Will try to provide more details...
Odd thing is: it works for most of my models, but not for Client

検索機能などと併用して使うときにpaginateした結果が"1"とだけ表示されてしまうので表示しないようにして欲しいです。

This is really a trivial thing, but I was wondering if there was a way to do this... Take the following url as an example

/categories/2/articles

This is actually loading the articles controller, with the category_id parameter. Now, pagination works just fine, but if you click a pagination link, it moves you to a url like the following:

/articles/?category_id=2&page=2

This isn't a functional issue, as it works as intended, and loads the proper records, and page. What I would like to know is if you can specify a way to keep the url structure in tact, without having to duplicate code into a second controller.

/categories/2/articles?page=2

Here's a gist of my code.
https://gist.github.com/0b1d2f6891b02e630f01

try to use :param_name => "search[page]"
you will end with {:search => {:page => old_page}, "search[page]" => nil}. I guess you have to convert such a param_name to proper hash before merge, and merge it recursively .

eg.

Creating scope :page. Overwriting existing method Invitation.page.

How can I show the total count of the found set. So if I have 900 customers, showing 100 per page, I would like to show "100 of 900 Customers" Thanks!

It would be nice if there was a paginate method similar to will_paginates ( paginate({ :page => 1, :per_page => 10 }) ) so that when coming from will_paginate to kaminari we don't have to change any code.

I tried:

<%= paginate @comments, :class => 'pagination ajax_pagination' %>

Which I would like to render as:

<nav class="pagination ajax_pagination">...</nav>

Is such a thing possible?

Hi,

No matter what I do the page parameter name won't change. Running version (0.10.1)

paginate @users, :param_name => :pagina

Any ideas?

I got a weird issue with an app rails3/mongoid

In my index controller, when calling .page(params[:page]), I got:

undefined method `conditions' for nil:NilClass

.bundle/ruby/1.9.1/gems/activesupport-3.0.5/lib/active_support/whiny_nil.rb:48:in method_missing' (eval):3:inpage'
.bundle/ruby/1.9.1/gems/kaminari-0.10.4/lib/kaminari/models/mongoid_extension.rb:10:in page' app/controllers/admin/images_controller.rb:7:inindex'

As the backtrace is not very verbose, I debugged and targetted the error here:
kaminari-0.10.4/lib/kaminari/models/mongoid_extension.rb:10
self.klass.page(*args).criteria.merge(self)
(rdb:2) pp self.klass.page
NoMethodError Exception: undefined method `conditions' for nil:NilClass

Any thoughts on this?

After adding kaminari to my Gemfile, pickle steps are no longer recognised by cucumber. Commenting out the gem brings them back. Anyone else noticed this?

Do you have any plans to support not ORM solutions like sunspot ?

I've got pagination in a few places in my application. For all of them I'd like to display the same number of results, and with the same pagination options.

So far, I'm setting the number of results in an initializer:
Kaminari::DEFAULT_PER_PAGE = 15 # Doesn't feel right

And creating a helper for the pagination links:
def custom_paginate(records)
paginate records, :window => 3
end

Then from the other places I call the helper:
= custom_paginate(@pages)

Perhaps there could be a simpler way to set some these options in order to overried Kaminari's defaults?

Thanks!

Hi,

When I try to paginate on meta_search results it just tells me that there is no method like "page"

def index
  @search = Article.search(params[:search])
  @articles = @search.page params[:page]
  respond_with @articles
 end

Error is:
undefined method `page' for #MetaSearch::Searches::User:0xb55d3440

Example from meta_search gem which is working flawlessly with will_paginate is:
def index
@search = Article.search(params[:search])
@articles = @search.paginate(:page => params[:page])
respond_with @articles
end

Not sure if this is a bug with Rails, Kaminari or what, but:

DEV 12:25:48 >> Item.count
SQL (1.5ms)  SELECT COUNT(*) FROM `items`
=> 574
DEV 12:24:33 >> Item.page(2).present?
SQL (1.6ms)  SELECT COUNT(*) FROM `items` LIMIT 25 OFFSET 25
=> false
DEV 12:25:43 >> Item.page(2).length
Item Load (0.9ms)  SELECT `items`.* FROM `items` LIMIT 25 OFFSET 25
=> 25

If you load the collection before using present?, it works:

DEV 12:27:08 >> items = Item.page(2); items.all; items.present?
Item Load (13.2ms)  SELECT `items`.* FROM `items` LIMIT 25 OFFSET 25
=> true

But remove the items.all; bit and it will fail like above.

I use MySQL.

Hello,

because I have an association called page, I'm getting lots of these error messages:
Creating scope :page. Overwriting existing method Model.page
Is there a workaround for that?

Not sure how to give a simple example, but if you have an AREL query that includes a group(term) clause, getting @result.total_count returns a hash of the group terms and their counts, not a count for all the items found.

@ontology_terms = OntologyTerm.where("").group('term_id').page(1).per(10).order('name')

[#<OntologyTerm id: 2019, ontology_id: 6, term_id: "1150|RS:0001685", ncbo_id: "1150", name: " F344-Tg(Cyp1a1-Ren2)10", annotations_count: 0, created_at: "2009-11-09 10:00:00", updated_at: "2009-11-09 10:00:00">, #<OntologyTerm id: 4635, ontology_id: 6, term_id: "1150|RS:0001731", ncbo_id: "1150", name: "(F344.OLETF-(D14Rat23-D14Rat12)(D14Rat8-D14Rat26)/2...", annotations_count: 0, created_at: "2009-11-09 10:00:00", updated_at: "2009-11-09 10:00:00">, #<OntologyTerm id: 4637, ontology_id: 6, term_id: "1150|RS:0001733", ncbo_id: "1150", name: "(F344.OLETF-(D14Rat23-D14Rat12)(D14Rat8-D14Rat26)/2...", annotations_count: 0, created_at: "2009-11-09 10:00:00", updated_at: "2009-11-09 10:00:00">, #<OntologyTerm id: 4640, ontology_id: 6, term_id: "1150|RS:0001736", ncbo_id: "1150", name: "(F344.OLETF-(D7Mgh16-D7Mgh20)(D14Rat23-D14Rat12)/2T...", annotations_count: 0, created_at: "2009-11-09 10:00:00", updated_at: "2009-11-09 10:00:00">, #<OntologyTerm id: 4641, ontology_id: 6, term_id: "1150|RS:0001737", ncbo_id: "1150", name: "(F344.OLETF-(D7Mgh16-D7Mgh20)/Tj X F344.OLETF-(D8Ra...", annotations_count: 0, created_at: "2009-11-09 10:00:00", updated_at: "2009-11-09 10:00:00">, #<OntologyTerm id: 4643, ontology_id: 6, term_id: "1150|RS:0001739", ncbo_id: "1150", name: "(F344.OLETF-(D8Rat54-D8Mgh17)/2Tj x F344.Cg-Leprfa(...", annotations_count: 0, created_at: "2009-11-09 10:00:00", updated_at: "2009-11-09 10:00:00">, #<OntologyTerm id: 4644, ontology_id: 6, term_id: "1150|RS:0001740", ncbo_id: "1150", name: "(F344.OLETF-(D8Rat54-D8Mgh17)/2Tj x F344.Cg-Leprfa(...", annotations_count: 0, created_at: "2009-11-09 10:00:00", updated_at: "2009-11-09 10:00:00">, #<OntologyTerm id: 4645, ontology_id: 6, term_id: "1150|RS:0001741", ncbo_id: "1150", name: "(F344.OLETF-(D8Rat54-D8Mgh17)/2Tj x F344.Cg-Leprfa(...", annotations_count: 0, created_at: "2009-11-09 10:00:00", updated_at: "2009-11-09 10:00:00">, #<OntologyTerm id: 471, ontology_id: 6, term_id: "1150|RS:0000003", ncbo_id: "1150", name: "13M", annotations_count: 0, created_at: "2009-11-09 10:00:00", updated_at: "2009-11-09 10:00:00">, #<OntologyTerm id: 423, ontology_id: 6, term_id: "1150|RS:0000270", ncbo_id: "1150", name: "A2", annotations_count: 2, created_at: "2009-11-09 10:00:00", updated_at: "2009-11-09 10:00:00">] 

@ontology_terms.total_count

{"1150|RS:0001685"=>1, "1150|RS:0001731"=>1, "1150|RS:0001733"=>1, "1150|RS:0001736"=>1, "1150|RS:0001737"=>1, "1150|RS:0001739"=>1}...snipped

If you are using both kaminari and the meta_where gems in a project, you may get the columns in the order clause or the resulting SQL being repeated.

Full details, including how to reproduce are here: http://gist.github.com/885335

In summary, the problem occurs when you use lines like this in a controller:

@articles = Article.order(:title.desc).page(1)

Note the usage of :title.desc, a feature provided by the meta_where gem (this yields a MetaWhere::Column instance). This will result in the following SQL being generated:

SELECT "articles".* FROM "articles" ORDER BY "articles"."title" DESC, "articles"."title" DESC LIMIT 25 OFFSET 0

Notice the repetition in the ORDER BY clause. I'm not sure if this is a meta_where or a kaminari issue.