#1256 Add tool for creating team projects

[Aaron Madlon-Kay]

It's a bit of a pain to set up a new team project, as there is a minimal set of required files that aren't really documented.

It would be nice to have a tool that would create the required files and perform any other required initialization for you, in the vein of "init" or "setup" CLI commands like git init.

Example:

$ mkdir teamproj
$ cd teamproj
$ git init
Initialized empty Git repository in .../teamproj/.git/
$ omegat team init en ja
Initialized en-ja team project. Place source files in ./source and supplemental
material in ./tm, ./glossary, and ./dictionary as required, then commit
everything.

This would automatically generate ./omegat.project with default settings (the Project Properties dialog GUI could optionally be shown to allow customization), create an empty ./omegat/project_save.tmx and ./glossary/glossary.txt, and create the other project folders such as ./tm, etc.

It could be useful to also detect the repo type (git or svn) and automatically stage (add) the generated files.

[Aaron Madlon-Kay]

Implemented in trunk. Usage is as above, or see team -h.

The option to show the Project Properties dialog is not implemented because the dialog implementation is too tightly coupled to the Core, Preferences, and Plugins systems at the moment.

[Jean-Christophe Helary]

How is that feature working ? I see a "Repository mapping" in the Create Project dialog, is that what it's about ?

[Aaron Madlon-Kay]

It's a CLI tool. Usage is as shown in the sample terminal session above (where omegat is java -jar OmegaT.jar if you're not using a launch script or alias).

[Jean-Christophe Helary]

Ok, I misunderstood the "it would be nice to have..." for what the implementation could look like. Thank you.

[Didier Briel]

Implemented in the released version 4.0 of OmegaT.

Didier

[Florian Aurich]

I just tried this and noticed that the generated project does not have the correct tokenizer set:
After creating the project with omegat team init en de, both the source and target tokenizer are set to DefaultTokenizer. It's the same for en-us de-de and EN-US DE-DE. When I choose these source and target languages via the GUI dialog, the tokenizers are automatically set to LuceneEnglishTokenizer and LuceneGermanTokenizer.

[Aaron Madlon-Kay]

Thanks for the report. Fixed in [bugs:#895].