- Added new labelslookup.json constants
- Added labelstemplatelookup.json and idforlabels.py
- Added build-label target for Makefile release
- Added
zh-Hant
language code and updated tests - Added Marwari and Bhojpuri langauge codes
- Added
m4v
to the list of convertible video formats - Added constants for
h5p
content type
- Added additional formats for other caption file types that are convertible by
pycaption
:- DFXP
- SAMI
- SCC
- TTML
- Fixed
document_thumbnail.order
so previews default to the document instead of the thumbnail
- Added necessary constants for Slideshows:
- New content kind added:
slideshow
- New format presets added:
slideshow_image
: represents a node-file association of the individual slideshow images part of the slideshow content nodeslideshow_thumbnail
: the thumbnail for content nodes of kind=slideshowslideshow_manifest
(allowed_formats:.json
) : the json manifest file for the slideshow generated duringexportchannel
for Kolibri consumption. This file will contain json of slideshow data computer from the studio DB (contents=association infostudio.models.ContentNode
--fk--studio.models.SlideshowSlide
)
- New content kind added:
- Added the following language codes:
- ach = Acholi
- fuv = Fulfulde Mbororo
- xog = Soga
- tuv = Turkana
- lwg = Oluwanga
- nyn = Nyankore
- myx = Masaaba
- Duplicated python constants/ definitions as json files under resources/ (for easier integration with frontend)
- Added tests to make sure python constants are in sync with json constants
- Implemented first_native_name helper (used to set channel names in chefs)
- Added list-like
convertible_formats
attribute for all presets - Added
html5_dependency
andvideo_dependency
presets - Updated constant mapping with epub -- The EPUB preset is used with document content type
- Added the following language codes:
- und = Undefined
- mul = Multiple languages
- arq = Algerian; Darja
- lkt = Lakhota; Lakotiyapi; Teton
- hdy = Hadiyya; Hadiya; Adea; Adiya; Hadia
- sid = Sidamo; Sidaamu afii; Sidaama; Sidama
- wal = Wolaytta; Borodda; Uba; Ometo
- Added file format constants for ePub
- Added RTL language lookup table in le_utils.constants.languages.RTL_LANG_CODES
- Added the following language codes:
- be-tarask = Belarusian
- hrx = Hunsrik
- lua = Luba-Kasai
- pms = Piedmontese
- sco = Scots
- rue = Rusyn
- Handle non-standard code
iw
for Hebrew used by YouTube API - Better classification of traditional and simplified Chinese scripts
- Added basic README
- Define constants for all file types
- fix wrong default Spanish lookup by name
- Added Sepedi and missing Tshivenda as discussed in previous PR
- Implemented getlang_by_native_name for easier lookups
- Language lookup helper functions
- changed version numbering to use minor versions not patch
- Added proquint
- added language direction constants
- added humanhash support
- changes to a few mimetypes