Help: Setup segmentation rule for Chinese

Jessicaliu wrote:
Hi! I find there is no segmentation rule for Chinese in OmegaT. Therefore, my source text is segmented in paragraph.

I tried to setup my own segmentation rule as follows.

Intention: set a segment after "。" (Chinese period)
Before: 。
After:

But, it's not working. All the sentence are still stick together. Could anyone tell me what to do? Many thanks.

Jessicaliu wrote:

Thank you for reply. I check the box.

The version I use is 2.2.0_2.

I also tried several source texts and move up the rule to the first one on the list. But, it seems that OmegaT does not recoginze my segmentation rule.

Jessicaliu wrote:
I checked the sentence-level segmenting.

But, your reply remind me that I forget to change "language pattern" from default to a specific Chinese language code that OmegaT is able to recoginize. I changed it to ZH-HK. And, it works!

Pierret Adrien wrote:

I am trying to set up some more segmentation rules for Chinese, and this is what I'd like to get :

No segmentation exception after a period 。is followed by a closing quotation mark ” (when said period is set up to be segmented in all other instances).

Example : “我想此句不要分开。”译员说。Shouldn't be segmented.
Under current rules, it gets segmented as follows :
“我想此句不要分开。[segment]”译员说。[segment]

My set up :
Break/Exception : unchecked
Pattern Before : [。？！]
Pattern After : [’”"]

But it doesn not seem to work, when all other rules set up for this language are all working fine. Any clue ?

Pierret Adrien wrote:
By the way, is there a place where we can consult a comprehensive list of signs used in the making of segmentation rules ?

I understood that [] means "any one of those signs", but what about +, or {} ? I couldn't get the grasp of it, and documentation doesn't mention it as far as I checked.

Pierret Adrien wrote:

Thank you, your link is actually helpful enough so I can improve Chinese segmentation a bit (I'll give you my feedback should you be interested in including it in a future release).