Hi Ann-Laure,

it looks like you have used TMs with different segmentation rules for the different language pairs

Use for the English - Spanish TM the same translation rules you have used for the English - French TM, than you should get the same segmentation in the XLIFF files

Stefano

Hi Ann-Laure,

it looks like you have used TMs with different segmentation rules for the different language pairs

Use for the English - Spanish TM the same translation rules you have used for the English - French TM, than you should get the same segmentation in the XLIFF files

Stefano

Hi stefano,

thanks for your answer , but i used the project English-french as model for the other languages. how can i have differents rules?

regards

Is your English-Spanish project actually provided with an English-Spanish TM?

If a project is without a TM, standard segmentation rules are applied (they may not be the same as your English-French TM)

If you have an English - French TM and an English - Spanish TM compare their respective segmentation rules, They must be the same.

Hi Stefano

i checked the both translation memories they have the same rules (standard)

My projects were with "without a TM".

After, I created new projects with TM and i obtain the same result (for US-french 443 segments and for US-SP 500 segments). i used prject US-Fr as model.....

regards

Hi Ann-Laure

can you clarify your workflow?

1) You start with a source ddl and use an app called App translator to convert it to XLF
Maybe the segmentation of the XLF file is generated by App translator?

2) You create a US-SP project based on the US-FR project.
Do you add to the US-SP project a US-SP memory?
What do you mean with My projects were with "without a TM"?

Hi stefano

We use APP translator to generate file with .xlf : one for french and for spanish

if we opened the file with Notepad++, we had for french

and for spanish

you can see for the ID216 we have the same source

for the point 2 : i created a US-SP project based on the US-FR project

i added US-SP memory (too small than french because i want to do translation) and i prepare with out TM as described in the image (i do the same for both project)

i tested with "prepare"only and the result is the same (443 segments for french and 500 for spanish)

Looking at the Notepad++ screens above, theere is a difference between the XLF file for French and the XLF file for English

The XLF for French is pre-translated (segmentation is already present), so Trados performs no segmentation

The XLF for Spanish is not pre-translated, so Trados performs segmentation (in a different manner compared to the French XLF file

To obtain the same segmentation, can you pre-translate the XLF for Spanish in APP translator?

Hi Stefano,

we tested to pre-translate the xlf file in spanish in Apptranslator, and export the files in order to use in SDL trados.

i created a new project based on US-FR project , and prepare the file (with TM project) , but the result is the same

Anne-Laure Galaud

The problem is because your source files are not the same.  As STEFANO DE ASTIS has already explained one is segmented already and the other is not.  For example this file is already segmented so Studio respects this:

While this one is not so Studio segments the content based on its own segmentation rules:

The solutions are these:

• you either segment all your files in the same way by preparing the files with the source and target elements in your own application, OR 
• you don't provide the target elements at all in all of your files and allow Studio to carry out this task, OR
• you create segmentation rules in your Studio TM for the files that are received in languages that have not been segmented to match the ones you use for the languages that have

Either way, this is not a Studio problem.  This is all about file preparation.