XBench cannot detect mistakes in the target text if there are multiple source terms with at least one correct target term present.
Normally speaking, if there is one term appear twice in the source text, the translator shall appear twice in the same target sentence.
However, if the translator translates the term incorrectly in the first position, and suddenly translate it correctly in the second position, the Xbench can NOT tell the issue even there is a mistranslation in the same sentence.
Please refer to the picture below for better understanding:
四月 = April (highlighted in blue)
六月 = June (highlighted in yellow)
一月 = January (highlighted in green)
You may see, the first 六月 is mistranslated (marked in RED with yellow background), and the second 六月 is translated correctly (marked in BLUE with yellow background).
I tried to write an expression to search, and here is my expression:
Source:
“六月”
Target :
-(June)
Search mode: Regular Expressions
PowerSearch: On.
Unfortunately, it doesn’t work perfectly.
May I know is there any better way to fix the expression in order to pinpoint such a sentence, please?
Your search looks for any segment containing a word “六月” in source and June is missing in target.
If the Chinese terms appears more than once but the target contains at least one instance of the translated term, those segments are not reported.
Is it possible that one of your double quotes are curly quotes? If you simply copied and pasted omartin’s expression, you might have entered the second double quotes in source as curly double quotes instead of straight double quotes.
Also please note that the expression in the checklist applies specifically to the case of 2 instances of June (it does not cover 1 instance or 3 instances).
