RE: [xliff] XLIFF 2.0 + ICU messages in Angular

["Rodolfo M. Raya" <rmraya@maxprograms.com>]

Hi Yoshito,

My comment on the "state" attribute is based on this extract from the Processing Requirements in section 4.3.1.31

• Writers MUST NOT set the state attribute values to other than the default initial if and only if the <segment> element where the attribute is set doesn't have the <target> child.

Attribute "state" is optional and there is no need to set it if you want to populate <target> with a copy of <source>.

Regards,

Rodolfo

--
Rodolfo M. Raya
Maxprograms http://www.maxprograms.com

-------- Original Message --------
Subject: RE: [xliff] XLIFF 2.0 + ICU messages in Angular
From: "Yoshito Umaoka" <yoshito_umaoka@us.ibm.com>
Date: Tue, April 13, 2021 6:26 pm
To: "Rodolfo M. Raya" <rmraya@maxprograms.com>
Cc: "xliff@lists.oasis-open.org" <xliff@lists.oasis-open.org>

Hi Rodolfo,


> When I mentioned contacting the Angular group, I was thinking about
> the validity issues I found in the sample file I received.
>
> I can't provide any advice on ICU Message Format, but I can mention
> that the "state" attribute should not be set to "initial" when
> <segment> contains a <target> element, for example.
>

I see. Thank you for clarification.

BTW, my interpretation of the spec is that "state" value can be "initial" even when non-empty <target> is supplied. I personally would not do it in my own tool, but I can imagine some people may prefer to produce <target> for all segment and fill in the same value with <source> as the start point. In this case, the state of segment is not really "translated".

If the intent of the spec is to invalidate such use case, I think we need more explanation. (Or, am I missing something?)

-Yoshito