Releases: mysociety/sayit
Version 1.5
Version 1.4
- Drop support for Django before 1.8.
- Add a web interface for importing Akoma Ntoso files. #435
- Show speaker's speeches most recent first. #482
- Hebrew and Slovak translations.
- Improved testing using tox.
- Use regular arrows in back/forward buttons. #509
- Akoma Ntoso import
- Add improved skip/replace/merge import options. #456 #434
- On directory import, import files in sorted order.
- Raise exception on empty 'showAs' attributes. #442
- Warn if docDate missing attribute or doesn't parse. #492
- Print out statistics in command line import. #460
- Print traceback for import exceptions. #475
- Make SSL certificate verification optional. #476
- Bugfixes
- Add translation of metadata. #497
- Fix issue with importing empty docDate/source_url. #484
- Compare section and speech keys with same type. #472
- Fix redisplaying of speech date in Django 1.7+.
- Give sections with no headings a title of 'Untitled'.
- Fix flake8 part of pre-commit hook, and colour code output.
- Fix display of speaker name when used in code.
Thanks to James McKinney, Duncan Parkes, Jaroslav Semančík and Matthew Somerville for their work on this release.
Version 1.3.2
Version 1.3.1
- Fix error in speaker colour display in Python 3.
- Akoma Ntoso import
- Better generic error handling, and of missing speaker TLCPersons
Thanks to Jaroslav Semančík and johnfelipe for reporting these issues.
Version 1.3
- Delete/ merge speaker front end interface. #407
- PopIt and Popolo JSON import management command and front end. #231 #310
- OpenGraph data output. #364 #416
- Dots in debate subsection timeline, fix section divider display. #372
- Spanish translation, bundled .mo files. #356
- Nicer section and speaker forms. #350
- Update the slug if you edit speaker or section. #433
- Show more editing buttons on home/ speakers pages. #417
- Improve speaker sorting, including ability to provide sort name.
- Use /section instead of /sections for editing URL endpoints.
- Akoma Ntoso import
- Development
Thanks to Duncan Parkes, James McKinney, and Matthew Somerville for their work on this release.
Version 1.2.2
- Fix a bug in South migrations introduced in v1.2 (80e3f0b) where migrating could give an error when it tried to rename the content type.
- Upgrade
django-subdomain-instances
to get its migrations. - Don't show keyboard instructions when no next/previous section.
- Fix typo in link in
README
.
Thanks to Andrew Langhorn, Duncan Parkes, Matthew Somerville, and Martin Wright for their work on this release.
Version 1.2.1
- Don't error if speaker image caching fails
- Add new-style migrations for Django 1.7
- Allow loading of remote Akoma Ntoso files
- Make it easier to override the
ImportCommand
behavior that is filesystem-specific - Remove requirement to run
collectstatic
before tests
Thanks to James McKinney and Matthew Somerville for their work on this release.
Version 1.2
- Cache speaker image from URL, using face recognition for thumbnail
- Improve display of subsections on section pages
- Unicode slug support
- Paginate /speeches
- Link truncated speech text to speech
- Tweaks to speech templates, including showing speech titles
- Add a date picker to speech/ section forms
- Unhide unimportant form sections which contain errors
- More translatable strings, including form labels
- Reduce database queries in get_ancestors and search indexing
- Model additions:
- Speech 'type'
- Section start date and time
- Section/Speech num, heading (renamed from title) and subheading
- Akoma Ntoso import
- Allow anonymous speeches
- Add support for more preface elements
- Development
- Easier local development settings override
- No requirement to have auth middleware installed
Thanks to James McKinney, Duncan Parkes, Matthew Somerville, Martin Wright, and Zarino Zappia for their work on this release.
Version 1.1.1
This is a small change to v1.1, switching to using a hosted version of django-bleach to remove a manual step when using Python 3.
Version 1.1
- Display improvements, including to the default front page, the speech detail page, nicer empty content messages and various tweaks
- Editing improvements, including a much nicer speech editing form, with inline creation of new speakers and sections
- django-popolo is now used to store people data
- Python 3 compatibility
- Removal of unneeded dependencies
- Included git pre-commit hook for CSS compilation
- Realtime search processer enabled in the example project
- Akoma Ntoso import
- Support coverPage and docTitle
- Log rather than die on error
- Namespace handling fix
Thanks to Richard Lewis, James McKinney, Duncan Parkes, Matthew Somerville, and Zarino Zappia for their work on this release.