Question

I have some text like this:

This is some text Z书. This is Zsome more text Z计算机.
This is yet some more Z电脑 text Z.

I need to delete all cases matching the pattern Z+(CJK), where (CJK) is any number of continuous CJK characters. The file above would become:

This is some text . This is Zsome more text .
This is yet some more  text Z.

How can I delete all CJK text matching this pattern?

Answer 1

You can using GNU sed to check the codes of non-ASCII characters:

sed -n l0 file.txt

Results:

This is some text Z\344\271\246. This is Zsome more text Z\350\256\241\347\256\227\346\234\272.$
This is yet some more Z\347\224\265\350\204\221 text Z.$

---

Then you can use GNU sed to do the replacement you desire. In my testing I had to set my locale to POSIX:

LC_ALL="POSIX" sed -r 's/Z[\o200-\o377]+//g' file.txt

Results:

This is some text . This is Zsome more text .
This is yet some more  text Z.