| Commit message (Collapse) | Author | Age | Lines |
|---|
| | |
|
| | |
|
| | |
|
| | |
|
| |
|
|
|
|
|
| |
- Redefined User#public_banned? to User#banned?
- Add specs for User#banned?
- Deprecate User#public_banned?
- Replace use of User#public_banned? with User#banned?
|
| |\ |
|
| | |
| |
| |
| |
| | |
Existing behaviour seems to be that when a tag is specified, only
bodies that already have that tag will be updated.
|
| | | |
|
| | |
| |
| |
| |
| | |
As I understand it, the expectation is to replace any existing tags
with both the tag from the UI and the tags from the csv.
|
| | | |
|
| | | |
|
| | | |
|
| | |
| |
| |
| |
| |
| |
| |
| |
| | |
Previously the call to public_body.send would return the value for
the default locale if no value was set in the current locale,
meaning that translations for attributes that were the same as
the attribute values in the default locale were not being loaded.
Fixes #2134.
|
| | |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
The code here was hard to follow. Now we have a cleaner path to decide
whether to show the Make Request button or not.
- Always show any notes associated with the body
- If a request can be made to the body
- Show any tag-specific notes that may affect the request
- Show the Make Request button
- If a request cannot be made to the body
- Show the most relevant reason why a request cannot be made
Added PublicBodyHelper#public_body_not_requestable_reasons to extract
the logic of finding the reason a user can’t request out of the template
|
| | | |
|
| | | |
|
| | | |
|
| | | |
|
| | | |
|
| |\ \ |
|
| | | |
| | |
| | |
| | |
| | | |
We want the HTML to be encoding in the default encoding so that
it can be combined with other strings.
|
| | |\ \
| | | |
| | | |
| | | |
| | | | |
Conflicts:
config/initializers/alaveteli.rb
|
| |\ \ \ \ |
|
| | | | | | |
|
| | | | | | |
|
| |\ \ \ \ \
| | |_|/ /
| |/| | |
| | | | |
| | | | | |
Conflicts:
config/initializers/alaveteli.rb
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
The pages for these search results are extremely slow to load. This
is not an ideal solution by any means. Really we want to dig into why
high offsets are so slow, and whether there's anything we can do to
fix that.
|
| |\ \ \ \ \
| |_|/ / /
|/| | / /
| | |/ /
| |/| | |
Conflicts:
spec/models/public_body_spec.rb
|
| | | | |
| | | |
| | | |
| | | |
| | | | |
Removes logic from views and obsoletes
AdminPublicBodyHelper#public_body_form_object
|
| | | | | |
|
| | | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
Covers core functionality.
Could do with extracting multiple assertions in to individual specs.
Strange that you need to call `reload` on the PublicBody instance when
testing “updates an existing translation and adds a third translation”
|
| | | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
Also fixes #empty_translation? to check for String and Symbol keys named
'locale'.
Specs use a pre-change check to assert difference.
For some reason rspec change matcher fails with 'nil is not a symbol'.
|
| | | | | |
|
| | |/ / |
|
| | | | |
|
| | | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
String#capitalize downcases remaining letters, so:
> 'heLLo WorLd'.capitalize
# => "Hello world"
Our version only works on the first character of the String, preserving
the case of the rest of the String:
> 'heLLo WorLd'.sub(/\S/) { |m| Unicode.upcase(m) }
# => 'HeLLo WorLd'
Also handle unicode.
|
| | |/
|/| |
|
| |\ \ |
|
| | | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
Store the search parameters in the flash when a search is made via the
select_authority_path page. Stores the parameters for both POST and
AJAX typeahead searches.
The presence of stored search params renders the link on the
PublicBodyController#show template.
“keep”s the search params in PublicBodyController#show so that if the
user clicks the browser’s back button the “Back to search results” link
can still be rendered on the next search result they click.
“keep”s all flash keys in ServicesController#other_country_message
as it’s called through AJAX and ends up sweeping the flash. [1]
[1] More details about this:
http://mikenaberezny.com/2007/09/08/keep-the-flash-and-test-it-too/
|
| |\ \ \ |
|
| | | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
Dropped the law_used_full interpolation:
So I think there are two significant bits of context here. One is that the
law_used_full attribute is derived from law_used, which in turn is set on
InfoRequest initialisation based on the tags applied to the public body. So
different requests in a batch could have different values for law_used_full
- some of them might be to bodies that only accept environmental
information requests. So using the value from the batch template is not
really a good proxy for the whole collection.
The second is that, in any case, the distinction between the two types of
request is a UK-specific feature which should be moved to the UK theme
(#2085).
Given these two factors, I think the cleanest thing might be to drop
law_used_full from this descriptive text, and just have it say "Your
requests will be sent shortly", without specifying what law will be used.
– Louise Crow (@crowbot)
|
| | |/ / |
|
| |\ \ \
| |/ /
|/| | |
|
| | | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
Fills in the first “Your e-mail:” field it finds, which could be
the sign _up_ form if the view is overridden to swap the order
of the sign in and sign up forms.
Also compacts a simple conditional on to one line.
|
| |\ \ \
| | |/
| |/| |
|
| | | | |
|
| | | | |
|
| | | | |
|
| | | |
| | |
| | |
| | |
| | | |
Some fields have moved to different models, or exist in translation
tables.
|
| | | | |
|
