Start a new topic

Detect source language

CafeTran set adapt the target language in a DOCX, provided that a source language is set in the source DOCX.

How about setting the source language in the Project Configuration dialogue box too, for DOCX?

I mean: how about letting CafeTran set the source language for a new project from a dragged docx automatically, based on a language code present in the dragged docx.

It would already be nice to type a letter into the corresponding field (e.g. "C") and have CT jump to the possible combinations in the list (with "C", this would be "Catalan") . 

This would speed up a change between different language combinations a lot.

Jawohl Torsten. All content of selection lists should be sorted. Preferable alphabetically ;).

No, not sorted. My comment was about navigating the language list.

Imagine that for a new project, you need to select another source language (e.g. Spanish instead of English).

Actually, in CT, you need to open the drop down menu and browse with your mouse/trackpad. This can be – depending on the language – a bit cumbersome. 

If it was possible to enter a letter ("S") and the whole list would jump to the first entry beginning with "S" (in this case "Sanskrit"), this would be very nice and accelerate the navigation through the whole list. This is also the standard behaviour for most drop down lists in the internet and in most Windows and Mac programs.

If entering the second letter ("p") would make the list jump to "Spanish" instead to the first entry of "P" ("Polish"), even better. 

This automatic recognition Hans proposes makes only sense if the source language is set in the source doc. Some of my clients don't do this always, mostly because of ignorance, but I do not know how many docs are set to the wrong source language. 

>No, not sorted. My comment was about navigating the language list.

Yep. That list is sorted. But other lists, like the file formats in the Project Config dialogue box aren't sorted. So no letter picking possible. And yes, of course, second pick letter should be supported too.

About the missing or wrong SL: Well, now it's always wrong (except when you previously translated with the same SL). The code detection would give an accuracy between 33% and 100%.

Login to post a comment