CafeTran

I'm not sure if anyone has suggested these already or not, but...

It would be nice if CT has regex rule checks or nontranslatable checks in QA, with the options like 'found in source but not in target', 'found in target but not in source', 'different number of occurrences between source and target' etc etc. 

QA - Word List seems to not work for this purpose, as it checks either source or target.

It would be nice if CT has regex find and replace that allows, in addition to numbers, replacements of group of words, like MS Word's wildcard.

If would be nice if Igor implementing them in the next update!!

Any idea?

Kwang

>It would be nice if CT has regex rule checks or nontranslatable checks in QA, with the options like 'found in source but not in target', 'found in target but not in source', 'different number of occurrences between source and target' etc etc.

I can see that this can come in handy!

>It would be nice if CT has regex find and replace that allows, in addition to numbers, replacements of group of words, like MS Word's wildcard.

Can you please give an example? Perhaps it's just a matter of smart regular expressions needed :).

Hi Hans,

For example, a patent consists of a lot of:

- 35% of propylene glycol, 25% of ethanol... etc.

The translation would be 'โพรพิลีนไกลคอลร้อยละ 35 เอทานอลร้อยละ 25...' 

Note that the percentage symbols disappear here.

What I will do is to find and replace all 'of propylene glycol' in target with 'โพรพิลีนไกลคอล' and do the same thing for 'of ethanol'.

Then I will find (\d+)% (\w+) and replace with \2ร้อยละ \1, where \1 refers to (\d+) and \2 refers to (\w+) or 'โพรพิลีนไกลคอล' that I have replaced before.

With this regex F&R, I can achieve the results simultaneously for both propylene glycol and ethanol. However, the drawback is that it is a bit greedy.

In fact, I can achieve almost the same things I want from regex F&R by using AA, but regex F&R is better (for my language pair) because I do not have to add spaces or Thai intonation marks that sometimes are missing during the process of AA if they are located at the end of a word.

Oh, and ability to search (or filter) source and target simultaneously would be nice too.

Kwang

>In fact, I can achieve almost the same things I want from regex F&R by using AA, but regex F&R is better (for my language pair) because I do not have to add spaces or Thai intonation marks that sometimes are missing during the process of AA if they are located at the end of a word.

You are aware of the fact that you can use regular expressions for AA?

>Oh, and ability to search (or filter) source and target simultaneously would be nice too.

Isn't this possible with F/R using Multiple filter?

>You are aware of the fact that you can use regular expressions for AA?

Yes sure. :D That's what I mean. I have a regex glossary for this purpose.

>Isn't this possible with F/R using Multiple filter?

Thank you for pointing me to this! 

I just tried it. But how can I filter source and target simultaneously?

Let's say I want to find segments where the source contains 'propylene glycol' and the target contains 'โพรพิลีนไกลคอล'

Kwang

>Then I will find (\d+)% (\w+) and replace with \2ร้อยละ \1, where \1 refers to (\d+) and \2 refers to (\w+)or 'โพรพิลีนไกลคอล' that I have replaced before.

Did you get \1 and \2 to work? I get them inserted (instead of the Find strings) :(.

Ah no. That \1 I mentioned is just an example of a replacement group if we use it. 

But I don't really know what we should replace with :(

There would be a lot of possibilities if this kind of replacement group exists.

Kwang

You have to use the $ character :).

See my regular expression for Michael Beijer, who didn't find time to thank me for investing time to help him ;).