T-Kit Creation

Single source file as input. Pipeline needs to create and pass along multilingual versions of TextUnits created via TM, MT or from other resources. Output is multiple T-Kits for each language.

Possible (but less efficient) that multiple pipelines could be created for each T-kit needed.

Merge or Create Target Content

Input is translated resource with (optional) original source document. Output is a translated version of the original document.

Alignment

Input is two or more documents. Documents can be pre-filtered and stored in serialized resource format. Output is TMX, XLIFF or other information about the alignment.

DIFF and Reuse

Input is old source and its translation with new source. Output is new source with translations taken from the old target using the English (and possibly skeleton) as alignment points.

Multilingual Input Formats TMX, XLIFF, PO

Input is a single file with multiple locales. Output is determined by pipeline.

Several steps, not all being filter-driven

One input document: myfile.properties

task 1: auto-translate it

task 2: convert it to US-ASCII

output: myfile.fr.properties

The task 1 requires the use of severalsteps: a filter step that feeds the auto-translate the writing of the translated file.

Then the task 2 taks the translated file as input and converts its encoding. This tasks does not use the filter.