RE: [xliff] Simplified XLIFF element tree

[Yves Savourel <ysavourel@translate.com>]

Hi Asgeir, all,

A last note before closing the laptop for a full week off:


> after translation:
> <trans-unit>
>  <source>
>    <seg id='seg1'>This is the first sentence.</seg>
>    <seg id='seg2'>This is the second sentence.</seg>
>  </source>
>  <target>
>    <seg id='seg1'>Første setning.</seg>
>    <seg id='seg2'>Andre setning.</seg>
>  </target>
> </trans-unit>

I find that approach would be simple and easy to work with.

One possible slight change could be to always have at least one <seg>. Maybe it would be more clear to express this using different names to avoid associating our knowledge of the current <trans-unit>, etc. to those elements. For example:

Un-segmented:
<unit>
 <source>
  <part id='seg1'>First sentence. Second sentence.</part>
 </source>
</unit>

Segmented and translated (same as Asgeir):
<unit>
 <source segmented='yes'>
  <part id='seg1'>First sentence. </part>
  <part id='seg2'>Second sentence.</part>
 </source>
 <source>
 <target segmented='yes'>
  <part id='seg1'>Første setning. </part>
  <part id='seg2'>Andre setning.</part>
 </target>
</unit>


> However, there probably need to be some ability to manage 
> state and workflow on the segment level, but that could 
> in a similar way live under the <trans-unit> element 
> and reference the segment ids.

And, for attributes, by simply moving the relevant ones to the <part>.

It would be easy to map xml:tm to this: <tm:te> to <unit> and <tm:tu> to <part>.

It would also allow Rodolfo to always have its XLIFF pre-segmented, and any tool could easily access the un-segmented content when needed: the sum of all <part> elements (and possibly the content in between if we would decide to allow content in-between).

And we would have no duplication of content.


Cheers,
-yves