adapt-node-transaltion

Adapt translation extract

this repo aims to ease translation of Adapt' s modules. It was developed as a quick and dirty extractor

currently these are notes for the development of this module

usage

for the moment this is a 2 step process

extraction

• cd to the script directory
• copy the content of the course folder of the ADAPT course to a folder named src (create it in the root dir of this script
• in the root of the script dir run npm run extract

this will treats :

• article
• config.json
• course.json
• contentObjects.json
• components.json

and create a extracted folder at the root of the script's directory containing MODIFIED version of the json files with reference to the translatable fields and a trad.json file.

KEEP THESE FILES TOGHTHER !!! order of the translated fields IS important so a trad.json is linked to the ADAPT's json files, and modification to the main files may end up genrating different ids

write trads

node wrTrad

will take the content of extracted and generate an output in the dest directory

modification of the trad.json file is left to the script user since the translation workflow is specific to each projects

goals/task

this tool should:

• [x]read the needed source files
• [x]extract needed fields
  • [x]store the field
  • [x]create field object containing
    • [x]a uId
    • [x]the original field key
    • [x]the original field content
• [x] replace extracted fields content with a template sting containing the field uId
• [x] write the extracted fields as a json object
• [x] read / write csv <=> json for the translators
• [x] write the source file with modified fields
• write a full model for the input, output and translations, for the next step (at work in the objectify branch)
• write objects to a database (level db)
• translation via object rather than lodash templates ?
• more exports formats (at work, XLS support)
• build an ui for translation input
• make config more clear and centralised
• write clean documentation
• harmonize trad object outputs
• automate process
• [x] have some tests

SHORT TERMS TODO

• finishing writing tests
• [x] treat special case of the hotgraphic and textinput (see git repo examples)
• finish all types of components
• [x]add specific cases for files other than component (currently treated by generic case)
• switch to streams
• write schema with all translated properties specified
• switch all debugging to use debug module
• create a better naming scheme for the translation keys
• switch translation writting from file based to object based
• write a script to reorder components
• make a better build process
  • use excel file instead of csv
  • steamline axa's csv production
  • integrate web vtt production to the build (and output files in the dest files as well)

notes

this system is useless now because of adapt system, but need improvement for legacy project, no need to go to far tho data extractions functions are in the lib/transformers.js folder
cases for each type of objects are located in the lib/cases folder

this is an internal tool, that is not yet in a polished state, use at your own risks!!!
A id based key is at work but for now this require stable json files since adding item may offset extracted translations keys.