| Commit message (Collapse) | Author | Age | Lines |
|
|
|
|
|
|
|
|
| |
Problems can have an associated defect type, that can be assigned during
an inspection. Include an admin interface for managing these types, that
can also be assigned on a per-category basis, currently available to the
Oxfordshire cobrand.
(Also include 'TM' in traffic management Exor RDI output.)
|
| |
|
| |
|
|\ |
|
| |
| |
| |
| |
| |
| |
| | |
The Open311 JSON output was wrapping each request parameter in a list.
This was presumably to prevent the XML output including the parameters
as attributes rather than children. A better way to fix this is to use
the NoAttr parameter to XMLout which forces them to be children.
|
|\ \
| |/
|/| |
|
| |
| |
| |
| |
| | |
If we're chugging through a list of reports, this save on a lot of
pointless statting of the upload directory which we know is there.
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
I can't put it better than
https://roland.codes/blog/legacy-flexbox-invisible-float-bug/ :
"Webkit’s implementation of legacy Flexbox has an strange bug that makes
flex items become invisible when they’re floated. What makes it strange
is that the flex items will still take up space in the DOM, as if they
have visibility: hidden applied.
Usually you wouldn’t use float on flex items (the spec even says that
float has no effect on them). But it’s useful for providing a fallback
layout when Flexbox isn’t supported.
Luckily I chanced on a hack that fixes the bug. All you have to do is
give the floated flex items a position, like position: relative, and
they’ll (strangely) become visible again. Another option is to use
Modernizr to only apply the float when Flexbox isn’t supported.
This bug affects legacy Flexbox in all Webkit browsers. Which includes
Chrome 20-, Safari 6-, iOS 6.1- and Android 4.3-."
|
|/
|
|
| |
This has been running for a year, added in e17eb20.
|
|\ |
|
| |
| |
| |
| | |
This works better if e.g. a user stops being associated with a body.
|
| | |
|
| |
| |
| |
| |
| |
| | |
To deal with a recipient mail server not allowing inbound email
using the same domain as an internal domain, e.g.
https://community.mimecast.com/docs/DOC-1419
|
|/ |
|
|
|
|
| |
As with 1115f106, a default might not be everything.
|
| |
|
|
|
|
| |
Uses the same environment as gettext-extract.
|
| |
|
|\ |
|
| |
| |
| |
| |
| | |
A cobrand may have a blank default that isn't everything, so the
"All reports" option must specify all the possible options.
|
| | |
|
|\ \ |
|
| | | |
|
| |/ |
|
| |
| |
| |
| | |
Also slightly tidy up display.
|
| | |
|
| |
| |
| |
| |
| | |
IE11 caches this Ajax call, not calling out to the server at all if it
is made again (e.g. after an update has been made). Fixes #1638.
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Avoid doing cobrand-specific logic within SendReport::Open311.
The hooks defined are:
open311_config: set extra fields
open311_pre_send: set endpoints and munge data
These are tested in the new t/app/sendreport/open311.t, which does
not send any actual reports, but just checks that all the data is
munged into the expected format. (tests under t/open311* should
already test the actual sending)
Remove unused Open311 special cases.
|
|/ |
|
| |
|
|\ |
|
| | |
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
* Tweak wording to make it clearer that the form is unsuitable
for residents wanting to report street problems.
* Record HTTP Referer when form is submitted, so we can see
which pages are leading people to the FMS4C page.
* Tidy up capitalisation elsewhere in headings on the page.
Fixes #1553.
|
|/
|
|
|
|
|
|
| |
This improves usability on mobile devices, allowing users to pan around
the problem and see the area in more detail.
I've also changed the CSS around a bit, so we can use the
`map-fullscreen` class in other places.
|
| |
|
|
|
|
| |
The interception of 404s etc is too much of an inconvenience.
|
|\ |
|
| |
| |
| |
| |
| |
| |
| | |
The list code was checking the photo ID given in the database existed on
disk – this shouldn't ever not be the case, and if it was, all that
would happen would be a broken image when the full photo code looks for
it (this is only for e.g. lists of reports).
|
| |
| |
| |
| |
| | |
The versioning code wasn't remembering when it had found that a file did
not exist, so was checking for them each time.
|
| | |
|
|\ \
| |/
|/| |
|
|/
|
|
| |
Fixes mysociety/fixmystreetforcouncils#98
|
| |
|
|\ |
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Reports in the generated Exor RDI file need to have an external_id so they can
be linked to existing enquiries (PEMs) when the RDI is imported into Exor.
There is a small window of opportunity for a report to be created, inspected and
downloaded in an RDI before it's been sent to Exor via Open311 and had an
external_id assigned. This commit closes that window by excluding reports from
the RDI that don't have an external_id.
Fixes mysociety/fixmystreetforcouncils#168
|
|/ |
|
|
|
|
| |
Fixes mysociety/fixmystreetforcouncils#166
|
|
|
|
|
|
|
|
|
|
| |
The timestamps returned from the database are in server-local time, not UTC.
Because FMS expects timestamps in Open311 responses to be UTC, the timestamps
of updates from OCC appear one hour ahead when DST is in effect.
This commit converts the timestamps to UTC before outputting them.
Fixes mysociety/FixMyStreet-Commercial#499
|
|\ |
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
The RDI file format encapsulates information about inspections that have taken
place, and can be uploaded into Exor to create defects in bulk.
This commit adds a page to the Oxfordshire cobrand's admin allowing RDI files
to be generated and downloaded from FMS.
For mysociety/fixmystreetforcouncils#127
|